This PDF file contains the front matter associated with SPIE Proceedings Volume 13291, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

The corrugated conical horn has been developed for a long time as an excellent feed with high gain and low cross-polarization. However, as the antenna system develops in the direction of dual band or even multi-band, the feed horn also needs to work in dual-band or multi-band at the same time as well. This paper starts with the design idea of ordinary corrugated conical horn, uses a simple parametric modeling design method to improve the design, and extends it to the design of dual depth corrugated conical horn. Finally, a 23.8GHz/31.4GHz dual-depth corrugated conical horn is designed, which can achieve VSWR less than 1.1 and cross-polarization peak below -20dB in 22.6GHz~25GHz and 30.2GHz~32.6GHz two discrete frequency bands. -12dB beam width is below about 25 degrees, so it can be used as an excellent far-field irradiation feed for the optimal design of quasi-optical optical path.

To achieve high quality industrial Anomaly Detection (AD) is difficult as there are no anomalous samples in the training set and one need to treat the defective region in the test images as outliers. Recently, some researchers even propose to perform the AD in a zero-shot manner. The algorithm is required to predict the defective state of each test image without the training phase. Surprisingly, the pioneering work has achieved relatively high accuracy in this challenging scenario thanks to the smart low density assumption of the anomalous patches in the feature space. In this paper, we propose to improve the successful zero-shot method by introducing the dynamic density estimation of the patch features. The experiment results illustrate a significant performance gain in all the involved 4 metrics for anomaly detection or localization.

Double port self-expanding millibubble resonators based on interior thermal air expansion was fabricated. Using the fabricated self-expanding bubble resonator and tapered fiber, whispering gallery modes (WGMs) coupling was realized. Furthermore, the fabricated double port bubble resonator was employed to conduct remote temperature sensing based on remote pressure variation caused by temperature change. Airflow caused by remote pressure variation reduced the inner pressure of bubble resonator and blue mode shift in transmission spectra was generated. Under 50°C temperature change, 39pm blue mode shift was detected and the corresponding sensitivity is about 0.8pm/°C. The feasibility of all-optical sensing based on remote temperature change is confirmed, which is anticipated to find applications in remote fire warning, dangerous chemical storage warning, etc.

The high-voltage electrical safety of electric vehicles plays a crucial role in their reliability development process, insulation fault is an important indicator to measure the high-voltage electrical safety of electric vehicles. Currently, there is relatively little systematic research on insulation fault diagnosis and localization, in view of this, we have modeled various high-voltage modules of electric vehicles and constructed a system simulation platform using Simulink. Based on the simulation platform, an insulation fault data-set was constructed, and the effectiveness of the system was verified through machine learning methods, followed by diagnosis and localization of insulation faults.

Gold metarsurfaces are widely used in sensing due to their ability to enhance localized light fields. To reduce the manufacturing cost of gold metasurfaces, we proposed a reusable metasurface biosensor based on parylene C with microfluidic technology. The electric field distribution of the gold metasurface was simulated using the finite-difference time-domain (FDTD) method. The polydimethylsiloxane (PDMS) microfluidic channels were bounded on the surface of the biosensor to provide specific reaction area to detect different concentrations of NaCl solution. For the purpose of reusability, the growth and removal of parylene C on the metasurface were realized. Besides, the parylene C film acted as the connection layer to enable the one-step physical modification of antigen and antibody, which made highly efficient biosensing possible. The feasibility of this approach is experimentally validated and has potential to be widely used in other metasurface biosensors.

Carrier phase-shift (CPS) modulation is a common modulation method for cascade H-bridge (CHB) multilevel inverters, while the traditional CPS modulation method uses a fixed phase shift angle. The disadvantage is that the output harmonic content is high, and the volume and weight of the filter are large. In order to further reduce harmonics and filter volume, a chaotic control strategy combining switching carrier phase angle method and carrier phase shift sinusoidal pulse width modulation (CPS-SPWM) is proposed. In this scheme, the carrier wave is jitter irregularly and the total number of switching times is unchanged, and the frequency spectrum concentrated on the fundamental wave is dispersed to the high frequency part which is easy to filter out, so the harmonic content is reduced and the power output quality is improved. A 7-level parallel CHB simulation model and experimental prototype were built for verification. The total harmonic distortion degree (THD) was reduced by 47.5% compared with the traditional method, which verified the effectiveness of the scheme.

On-load tap changers produce mechanical signals with unstable and nonlinear characteristics, making it difficult to accurately identify faults from vibration signals. To improve the accuracy of fault diagnosis of loaded tap changers, this paper proposes a fault diagnosis method based on variational mode decomposition (VMD) and multidimensional Transformer. Firstly, the dung beetle optimization (DBO) is used to optimize the penalty coefficient and the number of decomposed modes in VMD. Then, the decomposed modal components are input into the multidimensional Transformer encoder for signal feature learning. Finally, the fault of the on-load tap changers are realized by intelligent classification. The experimental results show that the recognition rate of the proposed method for the faults of loose contacts, loose springs, and broken springs of on-load tap changers reaches more than 96%. Compared with the Transformer and BP neural network algorithm, the recognition effect is significantly improved.

To further enhance the prediction accuracy and mitigate issues such as large current ripples, subpar steady-state precision, and excessive computational load in the Model Predictive Control (MPC) of Permanent Magnet Synchronous Motors (PMSM), this paper proposes a novel control strategy termed Discrete Space Grid-based Virtual Voltage Vector Model Predictive Current Control (DSG-VV-MPC). This methodology augments the prediction accuracy by discretizing the control space, which leads to an increased number of potential virtual vectors that more closely approximate the actual voltage vectors. Furthermore, the implementation of a preprocessing technique reduces the number of candidate vectors and the associated optimization iterations, consequently diminishing the computational burden. The findings from both simulation and experimental analyses indicate that the DSG-VV-MPC method significantly outperforms the conventional three-sector VV-MPC approach, resulting in improved steady-state performance and validating the effectiveness and feasibility of the proposed approach.

Accurately predicting the temperature of the stator winding of the traction motor and preventing the accelerated deterioration of the insulation system caused by high temperature plays an important role in prolonging the service time of the motor and reducing the operating cost of the EMU. The condition monitoring data of the traction motor has both temporal and spatial characteristics. However, most of the existing temperature prediction methods do not carefully consider the specific characteristics of time and space in the condition monitoring data of traction motors, resulting in the prediction effect not meeting expectations. Therefore, this paper proposes a prediction model based on the Deep Temporal Convolution Network (DTCN) and Convolution Block Attention Mechanism (CBAM), considering long-term time dependence and global spatial characteristics. First, DTCN is designed to extract multi-dimensional long-term time features in motor parameters. Secondly, CBAM is used to extract spatial features, and the attention weight distribution of multiple channels and spaces is realized at the same time. Finally, the temperature of the motor is predicted by fusing all the features of the nonlinear fully connected layer. Experiments on the motor temperature public dataset on Kaggle verify the superiority of the algorithm.

In the case of asymmetric faults in the power grid, it is of great significance to allocate the inverter output reasonably to achieve voltage support, which helps reduce the risk of generator tripping and promote the stable operation of the system. This paper proposes a voltage support strategy considering the maximum support level under asymmetric voltage drops. Firstly, the voltage support equation considering zero-sequence voltage is established using the vector method, and the active power deficit point determination formula is proposed. Secondly, considering the voltage drop depth and inverter output limitations, the voltage support level coefficient is used to divide the scenarios, achieving the three control objectives of voltage support, maximum active power output, and reducing voltage imbalance. Finally, the effectiveness and superiority of the support equation and control strategy proposed in this paper are verified through simulations.

Mimic defense is an important means of cyberspace defense. It defends against unknown attacks by constructing multiple heterogeneous executors to make the system redundant. However, existing mimetic arbiters often suffer from low efficiency because they judge the response data of executors. To address this issue, a mimetic decision rule based on heterogeneous executors path comparison is proposed to improve the voting efficiency. Based on the judgment of paths of executors, key positions in each heterogeneous executor are instrumented with route markers so that each executor automatically outputs markers on the actual execution path during runtime. The markers output by each executor are grouped based on similarity, then the algorithm of decision and scheduling rules based on historical performance and base address distance are set according to the grouping. Experimental results show that this method can improve the efficiency of mimetic decision-making while ensuring security.

Power silicon carbide (SiC) MOSFETs are gradually replacing silicon-based power devices in industrial production due to their higher switching speed, lower switching loss, and higher operating temperature capabilities, especially in high power density and high-efficiency power electronic applications. However, the higher switching speed also brings about issues such as voltage and current overshoots, oscillations. This paper proposes an on-chip integrated active gate driver for SiC MOSFETs, the principle of this circuit is to turn off some PMOS/NMOS during the switching process, which has a significant impact on the above problems, to reduce the driving current, in the stage where the impact on the above problems is insignificant, more PMOS/NMOS are turned on to increase the driving current, this approach aims to mitigate overshoots, oscillations, while maximizing the switching speed and minimizing power consumption. Compared to PCB-level active gate drivers, the on-chip integrated active gate driver have advantages in terms of reducing size, decreasing delay, and lowering application complexity. Finally, simulation results of the proposed driver under various conditions are provided.

Interrupted-Sampling Repeater Jamming (ISRJ), as a typical selfscreening jamming pattern, achieves effective interference against radar devices by generating multiple false target echo signals to mask the true target echo signals that need protection. According to the principle of jamming generation, it is known that the sampling duration and the repeater number of sampling segment are key parameters that determine the effectiveness of the jamming. In this study, through theoretical derivation and simulation verification, the impact of changes in the sampling duration and the repeater number of sampling segment on the interference effect was analyzed in-depth by adopting pulse compression processing on the signal. The results show that the sampling duration can affect the distribution of false target groups, and the repeater number of sampling segment can influence the spacing between primary and secondary false targets within the groups. By setting appropriate values, the false targets generated through Interrupted-Sampling Repeater Jamming can protect the true targets.

To achieve low power consumption and miniaturization design goals, this paper focuses on the research of a primary-side regulation quasi-resonant flyback converter structure. The converter operates in boundary conduction mode under heavy load, reducing the complexity of system control and avoiding excessive peak currents; it switches to discontinuous conduction mode under light load to improve conversion efficiency. The converter uses a clamp circuit to sample the output voltage and convert it into a current signal. The sampling is completed by a sample and hold circuit at the knee point, simplifying the output voltage sampling and eliminating the need for auxiliary windings. Circuit design and layout implementation were completed based on a standard 0.18μm BCD process. Simulation results show that the converter achieves a peak conversion efficiency of 88.27% under input voltage of 3-32V and output voltage of 5V, and it also has a conversion efficiency of over 70% under light load. The chip area is 1.48mm × 2.5mm.

Under the premise of ensuring network performance, ResNet proposes shortcut connections as the basic structure of the network to self-adjust the depth of the network to converge to the identity map. This resolves the performance degradation that occurs after the number of network layers becomes deeper. Field Programmable Gate Array (FPGA) is a highly flexible hardware platform, which is more convenient for the deployment of different neural network structures. Therefore, this paper accelerates ResNet object detection algorithm based on the FPGA platform. Firstly, the ping-pong buffer mechanism ensures data synchronization and correctness in parallel computing, so that the intermediate feature maps are alternately cached within the clock cycle. Secondly, it implements loop unrolling execution of computing tasks at each stage in the pipeline, which made full use of hardware resources and improved computing throughput. In addition, due to the limited resources of FPGA, operator fusion was combined to reduce the power consumption by reducing the number of reads and writes to the cache by the network. Although the accuracy of ResNet acceleration based on the FPGA platform is reduced, the detection speed is increased, and there are certain advantages in performance.

The current methods for anomaly detection in water level data are often limited in accuracy and application scope, failing to fully consider the non-linearity and non-stationarity of cascade hydropower station water level data. This paper proposes an improved water level data anomaly detection method based on Variational Autoencoders (VAE). This method integrates Bidirectional Long Short-Term Memory networks (Bi-LSTM) with the foundational VAE framework to learn the forward and backward temporal correlations in water level time series, enabling high-precision data reconstruction. To adapt to the varying water levels in different periods at cascade hydropower stations and enhance the sensitivity of anomaly detection, dynamic setting of anomaly thresholds is achieved by combining Exponential Weighted Moving Average (EWMA) with a sliding window technique. Compared to methods like ARIMA, Isolation Forest, and VAE-LSTM, this proposed approach demonstrates more stable detection performance with higher overall recall and F1 scores when dealing with various types of water level anomalies.

Wind turbine (WT) bearing health significantly impacts operational efficiency, but existing diagnosis methods struggle with noise and adaptability. This paper proposes a novel model to tackle these challenges. By framing the problem as a Markov decision process, the proposed approach automatically extracts features and learns optimal fault identification, eliminating manual effort. Tailored state/action spaces, reward functions, and exploration strategies enable the model to handle complexities in WT bearing signals effectively. The proposed intelligent diagnostic model was validated using the experimental data of WT bearing faults. Experiments demonstrate the superiority of the proposed approach over traditional methods, achieving significantly higher performance across diverse fault types. This paves the way for automated, intelligent, and universal WT bearing diagnosis, improving wind power reliability, safety, and cost-effectiveness. The study highlights the potential of the proposed method for tackling noisy, non-stationary data in complex industrial settings.

This paper proposes a fully integrated low dropout (LDO) regulator with gate-coupled flipped voltage follower(GC-FVF). The proposed GC-FVF addresses the limited output swing issue of conventional PMOS FVF in LDOs while maintaining a low output impedance. Besides, this LDO introduces a cascode compensation loop, which, along with the low output impedance of GC-FVF, pushes the output pole far away from the unity-gain bandwidth under both light and heavy load conditions. Consequently, the LDO becomes a stable two-pole system, supporting a high loop gain of up to 100dB and significantly enhancing the load and line regulation. Key specifications include a preset output voltage of 1.8V, a minimum unregulated input voltage of 2V, a maximum output current of 100mA, a ground current of 32μA, and an active chip area of 260μm×180μm.

In response to the problem of overheating and unstable temperature fluctuations in the electronic expansion valve in the car thermal management system, a control model based on fuzzy PID was developed by analyzing the working principle of the electronic expansion valve. The model was then used for thermal simulation in the entire vehicle thermal management system. The simulation results show that, at an ambient temperature of 25°C and a battery heating power of 8000W, when the cooling condition is activated and the compressor starts at a speed of 1000 rpm, the electronic expansion valve is controlled by the algorithm, resulting in the overheating starting from 0°C and reaching a maximum of 13°C, and eventually stabilizing at around 5°C with an overshoot of less than 5%. At the same time, the temperature of the battery pack is maintained at 15°C, meeting the requirements for the overall vehicle thermal management performance.

With the increase of distributed generation penetration, the inertia of the new power system decreases significantly. In order to reduce the effects of network disturbance, variable coupling and model mismatch, a linear active disturbance rejection control strategy based on improved bandwidth method is proposed. First, a first-order linear active disturbance rejection controller is designed for the LCL grid-connected inverter. Secondly, the bandwidth method is improved to solve problems such as the difficulty in selecting the initial value of the linear active disturbance rejection controller and the imperfect parameter tuning process. The regulation rules of bandwidth ratio and bandwidth center are introduced. At the same time, combined with the D division method of S domain, the appropriate parameters are selected according to the stable region, which improves the design efficiency and rationality of the parameters of the linear active disturbance rejection controller. Finally, the simulation and experiment verify the effectiveness of the method.

Using SOGI-PLL algorithm to control the inverter for pre-synchronous tracking of grid voltage can effectively ensure its safety and reduce the impact on the grid. Considering the phase delay caused by digital control and hardware circuitry, this paper carries out preliminary tracking of the grid voltage based on the SOGI-PLL phase-locking principle, and then extracts the grid phase and the inverter output voltage phase by using the switching phase identification algorithm, and then compensates the phase difference between the two by combining the PI link and the trigonometric formula. After verifying the effectiveness of the compensation method through simulation, a test platform with F28335 chip as the core is built, and the results show that the algorithm can compensate the phase delay of the output voltage of the grid-connected inverter, and the average value of the phase difference is 2.2° before compensation, and the phase difference is less than 0.005° after compensation, which effectively improves the phase tracking accuracy.

In response to the increasing demand for semiconductor power devices and highly reliable electronic products in China's military industry, a highly reliable sample was developed based on the chip structure and chip fabrication process of power VDMOS devices. Chip packaging and full parameter testing were conducted, and the sample was fully evaluated through application verification experiments. The experimental results show that the device has excellent performance, and all test results meet the design requirements, which can provide reference for the subsequent research of highly reliable VDMOS devices.

With the development of modernization, the demand for high-to-low voltage conversion is becoming increasingly important. To meet the requirements of power conversion, a new type of modular DC-DC step-down converter has been designed. First, a novel modular step-down circuit is designed based on the Buck-Boost circuit structure. Subsequently, a flyback power supply module is introduced, which, together with the new modular step-down circuit, forms a two-stage DC-DC step-down structure, significantly increasing the step-down ratio. The use of the flyback power supply module also solves the self-consumption problem of the step-down module, enhancing the overall stability of the step-down converter.

With the widespread popularity of electric vehicles, the development and application of V2G technology has become one of the hot spots of current research. However, the V2G system generates harmonics and unbalanced pollution during charging and discharging, which brings challenges to the operation and stability of the power system. In this paper, we use MATLAB/Simulink software to build a simulation model of V2G system, and propose a harmonic liability assessment index based on IEEE Std 1459-2010 power theory for the harmonic and unbalanced pollution problems generated by V2G system during charging and discharging. By measuring the voltage and current components on both sides of the V2G system during charging and discharging, the evaluation index is calculated and verified. Compared with the traditional harmonic responsibility assessment index, the proposed harmonic pollution evaluation indexes are more applicable for the V2G system. This provides a basis for accurately assessing the harmonic pollution responsibility share of electric vehicles and power grids in V2G systems and formulating corresponding harmonic control strategies, which helps to protect the safe and stable operation of power systems and the interests of both sides of power grids, and provide new ideas and methods for the research and application of V2G systems.

Large-scale new energy access to power system brings great challenge under double-carbon background, and demand-side resource (DSR) provides an important solution for power system regulation. However, a large number of distributed DSRs are limited to participating power system regulation because of their market size and the ability to participate in the scheduling, and thus lead to the lack of market competitive and low actual response of these DSRs at present. An optimal regulation method of DSR aggregation with the participation of load aggregators (LAs) is proposed, considering carbon emission and multi-type demand response. Firstly, the multi-level management mode of DSRs is introduced. Secondly, aiming at the lowest economic cost and the lowest carbon emission, a DSR aggregation optimization regulation model was established considering carbon emission and multi-type demand response. Finally, the validity of the optimal scheduling model is verified by simulation solution. The simulation results show that the model can effectively reduce the economic cost and carbon emissions of the resource subject on the demand side, reduce the energy cost of distributed DSRs by 5%, improve the level of resource management on the demand side, and provide an effective means for the economic optimization and regulation of the power grid under the requirements of low carbon and toughness.

This paper makes a profound study of flexible printing plate-making technology, and successfully developed an efficient, green, and stable flexible printing plate-making system. Through the presentation of imaging results, it is evident that the system achieves significant advantages in print quality and accuracy. The open design of this product provides robust support for the flexibility and maintainability of this system. Ultimately, through the design of manual plate loading and automatic plate advancement, this system achieves a balance between operational convenience and efficiency, making flexible printing plate-making work more efficient and convenient. This research provides strong support for the development of flexible printing plate-making technology, and that introduces Internet of Things technology which could achieve intelligent interconnection between devices, remote monitoring, data collection and analysis, and provide users with a more intelligent and convenient printing experience.

The reclaimer, an auxiliary equipment, is important in thermal power plants, it affects the safety and economy of thermal power plants. Therefore, the bearing fault detection of the reclaimer is studied in this paper. For the acquisition of 1D bearing vibration data, empirical mode decomposition (EMD) is employed for signal decomposition. And the residual network is used to extract local features from the data. Further, Bi-directional long short-term memory (Bi-LSTM) is used to extract global features of the data. In the experiments, the numbers of EMD decomposition layers and iterations are analyzed to improve the accuracy. And, compared to convolutional neural network (CNN), the proposed model has high accuracy.

Data-driven methods based on machine learning have been successfully used in image processing, data dimensionality reduction and other fields. At present, there is still no theory in the field of composite materials applications that can give highly accurate prediction results under all circumstances. This paper designed biaxial tests of composite materials to obtain multi-dimensional data such as DIC digital speckle data, testing machine load displacement data, and strain data. Based on the fusion algorithm, a multi-dimensional data set that can be used for machine learning was constructed, which will expand the training data-driven model. Data availability. If rich constitutive information is combined with physical models, it can lay the foundation for subsequent data-driven research on failure criteria and mechanisms of composite materials.

The primary focus of this research is the investigation of the cycloidal pin wheel drive system, specifically emphasizing the derivation of the tooth profile equation for the cycloidal wheel. By applying Hertzian contact theory, the spacing between two wheels during initial interlocking, along with the contemplation of contact distortion throughout the meshing procedure, are taken into account. This enables accurate determination of the interlocking position and derivation of the contact load and stress during meshing. Using the prototype as the analysis object, MATLAB was utilized for iterative calculations, yielding the curve of contact load and stress relative to the meshing phase angle. It provides theoretical support and experimental data for the analysis and optimization of cycloidal pin wheel drive, and enhances understanding of the drive system's mechanism and performance.

In order to solve the vibration problem of cycloidal wheel and planetary frame in the process of RV reducer transmission, finite element analysis software ANSYS Workbench was used to carry out constraint modal analysis. By analyzing the first 12 order frequency and mode nephogram of cycloidal wheel and assembly, it is found that the weak links are located in the tooth output axis of cycloidal wheel respectively. The meshing frequency of the gear and the rotation frequency of the planetary frame are calculated and compared with the first order frequency to verify that the cycloidal wheel and the assembly do not resonate.

A virtual power plant has the capability to aggregate decentralized energy resources, including generation, storage, and demand, thereby enhancing the flexibility of the power system and providing support for the integration of new energy sources and the development of innovative power systems. This paper focuses on a distributed industrial and commercial park with thermal and cold energy storage facilities, constructing a microgrid structure for the Virtual Power Plant (VPP) park. Unit models for energy production, conversion, and storage devices within the park are established. The objective function of minimizing operational costs for the park is considered, taking into account constraints related to energy balance within the park, grid interactions, and device operations. An optimization scheduling model for the distributed combined cooling, heating, and power (CCHP) VPP park, incorporating thermal and cold energy storage, is developed. The inclusion of energy conversion equipment such as gas-fired boilers and electric refrigeration, along with energy storage facilities like thermal and cold storage, effectively enhances the flexibility of energy supply scheduling within the park. This allows for the decoupling of heating, cooling, and electricity scheduling and enables the time-shifting of energy consumption, leading to a noticeable increase in the potential and profitability of the distributed VPP park's participation in power demand response. Finally, a case study is conducted to validate the proposed model, using a real VPP park in Shanghai, China, participating in peak shaving demand response. The results indicate that the distributed CCHP VPP park, equipped with thermal and cold energy storage facilities, can provide peak shaving and load-filling capabilities to the power system. Furthermore, the effective integration of energy storage enhances the overall flexibility of the VPP park, and appropriate demand response compensation mechanisms effectively incentivize its participation in the demand response market.

Taking a 1000 MW ultra-supercritical boiler heating surface pipe as the research object, the microstructure and mechanical properties of T92/TP310HCbN secondary reheater pipe and its welded joints after its service for about 60, 000 h were analyzed. The deterioration degree of the two materials and welded joints after long-term service was studied. The results show that the microstructure of welded joints and base materials of T92/TP310HCbN steel is normal and no obvious aging phenomenon is observed after high-temperature service. In terms of mechanical properties, the yield strength and tensile strength of the two kinds of steel pipe welds meet the standard requirements, but the elongation of TP310HCbN steel pipe after fracture decreases obviously, indicating that its brittleness increases. In addition, the hardness value of the coarse crystal zone on the side of T92 steel is high, and the metal supervision in this area should be strengthened to ensure the safe operation of dissimilar steel welded joints under long-term high temperatures and high pressures.

The PN junction temperature (T_cell) is a crucial factor in the lifetime of photovoltaic (PV) modules, which is a significant parameter that impacts the output power of PV stations. However, the current model to estimate the T_cell of PV modules in real operating conditions suffers from low precision and limited application scenes. This paper proposes a method to estimate the T_cell by incorporating multiple environmental factors. Firstly, an optimization algorithm is employed in obtaining the T_cell by fitting the output power characteristics of PV modules under various operating conditions. Subsequently, the ambient temperature and other meteorological data are utilized as inputs in training a neural network, with the T_cell to be the output. Finally, the trained neural network is employed to calculate T_cell for the specific operating conditions. The resulting T_cell is subsequently plugged into the transformation equation to yield a more precise estimation of the PV module’s output characteristics.

To address the challenge of the lack of aging characterization methods for outdoor silicone rubber used in electrical field, a non-destructive quantitative method for evaluating the aging process of silicone rubber has been proposed. A long afterglow phosphor is selected and introduced into the silicone rubber as an indicator. Samples are studied after artificially accelerated aging. The results show that the average ratio of luminescence intensity is positively correlated with the hydrophobicity of silicone rubber (an indication index of the aging process). It has been demonstrated that the method is feasible for evaluating the aging process of room temperature vulcanized (RTV) silicone rubber with a high accuracy represented by the narrow distribution of the indication index. Further, it has been proved the addition of the indicator may not degrade the performances of silicone rubber.

To meet the ever-increasing power supply reliability requirements of the power system, this paper proposes an XGBoost fault warning algorithm based on normal operation data for switchgear. The method effectively analyses a large amount of data by establishing a temperature prediction model using the XGBoost algorithm, which is then combined with a dynamic threshold algorithm for fault determination. In real engineering cases, through the validation of historical data, we verify the significant advantages of the method in terms of prediction accuracy and real-time performance. This method, which integrates multiple technical tools, is expected to provide accurate fault warnings for power operation and maintenance, thus ensuring the safe operation of the power system.

In the parallel operation system of the low-voltage microgrid inverters, due to the difference of line impedance and other factors, the equivalent output impedance of inverters is mostly resistive or resistive-inductive, and the use of traditional sag control will make uneven reactive power distribution and power circulation in the parallel inverters. Based on the traditional sag control strategy, a sag control strategy based on adaptive virtual impedance is proposed to adjust the inverter equivalent output impedance to be inductive, which improves the accuracy of reactive power allocation and suppresses the power circulating current; meanwhile, the adaptive virtual impedance follows the changes of the system voltage and current, and continuously adjusts the impedance value on-line to make up for the system voltage landing. The inverter parallel operation simulation model is built in MATLAB/Simulink, and the simulation results verify the effectiveness of the methods.

Aiming at the continuous improvement of aerospace DC/DC conversion efficiency requirements and the relatively poor electrical characteristics of available aerospace power semiconductor devices, the research on half-bridge resonant soft switch topology for efficient conversion was carried out. The working process of the half-bridge resonant converter under PFM and PWM control modes is analyzed. A half-closed loop hybrid control method suitable for output voltage interval control and synchronous rectification control method based on multiple side current sampling is proposed. For different output voltage control sections, different control schemes and synchronous rectification schemes are adopted to achieve the balance between converter efficiency improvement and power supply security.

Due to the time-varying admittance matrix of the detailed electromagnetic transient model of VSC, it is inefficient to simulate large power grids with large-scale new energy stations. To solve this problem, this paper studies the mixed integration method, and proposes a decoupling model based on this method, which has simple decoupling circuit, constant admittance matrix, and can account for the internal characteristics of the converter. Firstly, based on the state equation of VSC converter, the Mixed integral model of two-level VSC converter is constructed by using matrix splitting and delay techniques to realize the decoupling of AC and DC. Then, the calculation of parameters of the decoupling model is simplified by taking advantage of the mutual exclusion of the longitudinal commutation of the upper and lower VSC bridge arms, so that the admittance matrix of the system is constant and the calculation speed is further improved. The numerical results show that the proposed method can improve the calculation speed while maintaining high precision.

It is difficult to accurately calibrate the traveling wave head and calculate the wave velocity by using the single-ended traveling wave location method. In this paper, a dynamic first traveling wave energy that varies with the fault distance by deeply analyzing the attenuation characteristics of the line transmission function on each first traveling wave frequency component. The maximum first traveling wave energy, namely the main frequency component, monotonically decays with the increase of fault distance, showing a one-to-one mapping relationship with the fault distance. On this basis, this paper proposes a single-ended fault location method for flexible HVDC transmission lines using the linear interpolation. The positioning results are calculated using the linear interpolation method on the distance-initial traveling wave main frequency component curve. The simulation results show that the proposed positioning method eliminates the dependence of traditional single-ended traveling wave positioning methods on the wave head calibration and the wave velocity calculation. The absolute positioning error is less than 0.6km, and the positioning result is not affected by transition resistance.

With the development of science and technology, the stability and reliability of power systems are required to be higher. Substations, as an important part of power systems, directly affect the stability of power systems and user experiences. In recent years, the access of new energy and the development of power electronic technology have made power quality issues increasingly prominent, and the assessment of power quality in substations is particularly important. This paper proposes judging the operation condition of substation equipment based on the output power quality of substations. Secondly, the combination weight is obtained through the Lagrange multiplier method, and then the power quality of the substation is evaluated by the good and bad distance method. Power indices are visually displayed in the form of a chart to judge the operation status of the substation and whether maintenance is needed. By evaluating the real operation data of three substations, the feasibility and effectiveness of the proposed evaluation model in the power quality assessment of substations are verified.

This article focuses on the modeling of aggregated interference in cognitive radio-based Wireless Regional Area Network (WRAN). We develop two distribution models to represent the spatial distribution of primary and secondary users, examining various factors that influence overall interference levels. The first model describes the aggregated interference experienced by primary user receivers due to secondary users within the network, assuming that the primary user's location is independent of the secondary user network and their receiver information is known. The second model considers the interference generated by secondary users within the network on primary user receivers, with both the primary user's location and receiver information being unknown. We derive the cumulative distribution function (CDF) for both models, employing a logarithmic normal distribution approximation to reduce computational complexity. Experimental results confirm the consistency between our analytical model and Monte Carlo simulations, validating the accuracy of our approach. Finally, we analyze the impact of key parameters such as secondary user density, transmitter power, and distance from primary to secondary users on cognitive network interference. These findings offer valuable insights for practical deployment of cognitive networks in real-world scenarios.

Silicon Photomultipliers (SiPM) is a low light detection sensor with a wide range of applications. The readout circuits of SiPM need to be optimized for time jitter in applications that require high time performance. This paper reports three readout circuits for SiPM (Hamamatsu S14160-4050HS), using radio frequency (RF) amplification method, transimpedance (TIA) amplification method with voltage and current feedback operational amplifiers. The effect of different sampling resisters or feedback resisters on the signal rise time was tested. The increase of rise time when the sampling resistance is very small in the test results is simulated and analyzed, which is considered to be the influence of parasitic parameters. The time jitter of the SiPM at different overbias voltages was tested using three circuits. Among them, the time performance of RF circuit and current feedback TIA is better, and the time jitter of RF circuit can reach about 100 ps FWHM.

To validate the conversion performance of ADC at different temperatures, the demand for components in wide-temperature application systems is addressed by selecting high-reliability system components. This ensures stable operation of the analog-to-digital converter module across various temperatures, particularly in high-temperature environments. To achieve high-precision analog-to-digital conversion in wide-temperature environments, the ADC peripheral driving circuit is designed and optimized to guarantee signal acquisition accuracy, stability, and immunity to interference. In the experimental section, a high-temperature ADC performance testing system is constructed based on this analog-to-digital conversion module. The dynamic performance of the 12-bit ADC is evaluated using FFT spectrum analysis. Experimental results demonstrate that the effective number of bits of the ADC exceeds 10 bit within the temperature range of -40°C to 250°C, indicating excellent conversion performance of the ADC and high reliability of the peripheral circuit. Implementation of the high-temperature ADC performance testing system also provides valuable reference for constructing signal acquisition systems based on high-temperature ADC in practical high-temperature application scenarios.

The rapid development of electronic warfare greatly promotes the importance of the radar cross section (RCS) reduction of array antenna mounted on the combat aircraft. In this paper, the influence of position error of antenna elements on array RCS is studied statistically through mathematical deduction. The random position error submitted to normal distribution is taken as an example to deduce the relation between RCS variation and the tolerance of antenna elements, which provides an effective and direct way to assess the influence of the position error of antenna elements on the stealth performance of the combat aircraft.

The accurate identification and diagnosis of single-phase line faults in distribution networks based on primary and secondary fusion column switches have long been challenging due to the lack of effective detection and protection algorithms, directly impacting the safety of distribution lines. To address this issue, this study investigates the variations in sequence current and line voltage signals on the power source side and load side of single-phase line faults based on composite sequence networks. By leveraging the easily observable characteristics of the magnitude jump of positive and negative sequence currents of the faulted phase, the fault type is indexed. Additionally, the fault location is determined based on the characteristics of line voltage magnitude at different positions along the faulted line. Through Matlab simulation and experimental testing on a test bench, the results demonstrate that the fault detection algorithm, which integrates the variations in sequence current and line voltage magnitudes, can successfully achieve fault localization, laying the foundation for the design of subsequent fault detection schemes with superior performance.

With the gradual maturity and popularization of DC distribution network technology, there is an increasing demand for DC distribution networks with multi-voltage level configurations in various power system applications. Traditional DC converter technology is limited to providing only one output voltage level, making it challenging to satisfy the intricate needs of DC distribution networks that require multiple voltage levels. In this paper, a new multi-port photovoltaic-energy storage DC distribution network topology for multi-voltage levels is proposed, i.e., using multi-winding transformers and two AC power input ports to construct AC power buses with multiple voltage levels, and forming DC buses with different voltage levels of 0V, 750V, 1500V, and 2250V on the DC side through AC/DC converter units. Simultaneously, the suggested structure incorporates energy storage batteries and solar panels, which are connected to the DC bus using DC/DC converter technology. This setup offers dynamic energy support and the ability to regulate the distribution network.

In this article, tensor technology is applied to address the issue of slow speed in target positioning for near-surface finite conductors with a transient electromagnetic (TEM) system. Both simulation and experiment have been conducted based on a portable TEM system. Simulation results show that the tensor localization is more accurate when the baseline distance is smaller and the target depth is larger. Experiment results show that with a baseline distance of 40 cm and a target depth of 50 cm, the horizontal positioning error does not exceed 7 cm and the depth error does not exceed 6 cm after compensation when taking the average of positioning results of 25 measurement points. As a result, tensor localization can achieve accurate target localization for TEM detection.

To address the challenge of complex degradation in solenoid valves with increasing operation cycles, this study proposes a method for analyzing solenoid valve performance degradation based on multiple physical characteristics. Initially, by scrutinizing the operational mechanism of the solenoid valve, physical performance indices such as the first peak time of current, spool movement time, and pressure peak time are extracted from the time series. Subsequently, the Moving Average (MA) method is employed to mitigate the degradation trends in multiple physical performance indicators. Finally, a polynomial fitting model is utilized to fit the degradation trends in the experimental data of a certain solenoid valve, resulting in the performance degradation curve.

The distance measurement method based on deep learning for monocular vision transmission line inspection unmanned aerial vehicle(UAV) has the advantages of low cost, high efficiency, and intelligence. To guide the dense encoding information at different scales more effectively and address the issue of image clarity reduction due to the loss of high-frequency information during upsampling, this article proposes a Laplace U-Net (LUNet) network architecture. The architecture enhances the global information decoding capability with the proposed LU module, introduces a Laplacian pyramid residual structure in each reconstruction module, and utilizes the encoding information from different depths of the encoder to guide the reconstruction of high-quality depth images. The experimental results show that the relative error of the new method is less than 6%, which provides technical support for the inspection UAV to carry out the distance detection technology of the transmission line and the tree conveniently through the monocular camera.

This paper focuses on nickel metal hydride batteries that are monomers. By increasing the adaptive rate of BP neural networks, the batteries can increase discharge, with temperature, current, and end voltage acting as the network's input characteristics. Nickel metal hydride batteries are the primary backup battery used in mining power supplies. Their high stability and energy density make them ideal for improving battery SOC (charge) estimation accuracy and prolonging service life at the same time. In accordance with the difference between the training sample's actual and expected output, adjust the momentum term; in accordance with the backpropagation error's direction change, raise the adaptive learning rate. In contrast to the traditional BP neural network approach, there was a 78% increase in convergence speed and a steady estimation error within 10%.

With escalating environmental concerns, notably climate change and air pollution, electric vehicle (EV) power systems have gained paramount importance and research significance. Fulfilling the electricity demand of EVs necessitates efficient energy management strategies for high-voltage power supply systems. Traditional strategies, reliant on logical thresholds, exhibit limitations, thereby urging a shift towards investigating energy management strategies incorporating fuzzy logic. This article presents a simulation model for the high-voltage power supply system of EVs, evaluating two energy management strategies through simulation. Comparative analysis of experimental results reveals that the strategy based on fuzzy logic outperforms the conventional approach. Findings indicate that the proposed energy management strategy effectively distributes energy output across various components of the high-voltage power supply system, optimizing performance during starting, accelerating, cruising, and braking conditions, thereby enhancing overall system efficiency.

With the continuous advancement of high-density space launch missions, the vibration phenomenon generated during launch has gradually attracted attention. Research has found that most of these vibrations can have varying degrees of impact on towers and satellite rocket bodies. This paper proposes a vibration measurement method based on domestic Microelectromechanical System (MEMS) acceleration sensors for vibration characteristics related to structural mechanics such as large vibration impact, wide spectrum range, and low amplitude and frequency of vibration signals in the active section of the rocket launch during the actual launch process of a certain satellite launch site. This method is used for real time acceleration detection of X-axis, Y-axis, and Z-axis vibrations, uses MEMS sensors to collect vibration signals, completes the display of acceleration amplitude through a series of processes such as low-pass filtering, analog-to-digital A/D conversion, data acquisition and analysis. And through software simulation and hardware circuit testing, the feasibility of the scheme design and the reliability of the vibration measurement system were verified, providing solutions and technical support for the subsequent development and transformation of the launch site.

In order to improve the photovoltaic power generation efficiency of the Kunlun Station in Antarctica, it is necessary to simulate and model it in order to control it. However, current photovoltaic modeling is mostly based on numerical models, which is not conducive to subsequent system updates. On the basis of traditional mechanism models and combined with the actual environment in Antarctica, this article analyzes the impact of low temperature conditions on the system. By introducing environmental correction parameters, the photovoltaic cell model is optimized. Based on the Wiener process, a degradation model of battery capacity is proposed. The experiment shows that the optimized model can effectively simulate the Antarctic environment.

There is a situation where a single suspension point and single string insulator are used for the suspension of conductors at some crossing points of existing overhead transmission lines, which is prone to wire detachment and poses great safety hazards. This article proposes an improved YOLOv5 based method for identifying single suspension point and single string insulators at cross crossing points of overhead transmission lines. We devised a Lightweight Attention Mechanism Module (LAMCM) to bolster the network's feature extraction prowess and appended an additional prediction head atop the original YOLOv5's trio of prediction heads, aimed at enhancing the detection of targets across varying scales. Upon validating the augmented YOLOv5 algorithm on both the COCO dataset and a proprietary transmission line dataset, the outcomes revealed that our proposed algorithm surpasses numerous prevalent object detection methods. Specifically, on the COCO dataset, our algorithm achieved an impressive 79 FPS and 49.1% AP on a Tesla V100 platform. Meanwhile, on the self-constructed transmission line dataset, our algorithm registered a recognition recall rate of 96.8% and an accuracy rate of 94.3% for single suspension points and single string insulators, exhibiting enhancements of 1.7% and 1.8%, respectively, over the vanilla YOLOv5.

Neural radiance fields (NeRF) is an important work in the fields of novel views synthesis, representing scenes as 5D functions that model the color emitted by objects directly based on their positions and observed directions. The rendering approach of NeRF faces challenges in accurately capturing the rapid and complex variations in specular reflections on glossy surfaces. In the computer graphics, physically based rendering (PBR) facilitates the real-time rendering of realistic lighting effects. Drawing inspiration from this, we deeply analyze the reflectance render equation and propose PGlightNeRF to enhance the lighting representation of NeRF. Our study indicates that our approach can achieve training and rendering in a short timeframe, synthesizing high-quality specular reflections views.

In a large-scale interconnected power grid, when a fault occurs in the grid, the system current will be redistributed, and if left untreated, the overloaded line or transformer will be removed, which may cause a series of chain reactions or even system collapse. The grid transmission line fault analysis based on graph database proposed in this paper mainly relies on the powerful relational query and storage capability of graph database. Firstly, by constructing the graph of transmission lines, it is able to integrate the line equipment and their relationship information with each other, and store various information of transmission lines, such as the geographic information of the lines, the equipment models, and the operation status, etc., in the form of a graph database. Secondly, the graph database is utilized in combination with graph convolutional network to construct a graph convolutional network based on the graph spectrum to quickly locate fault points. Finally, by analyzing the operating status, maintenance history and other information of the line equipment, the possible fault time and fault type of the equipment can be predicted, so that corresponding maintenance measures can be taken in advance to avoid the occurrence of faults. After testing, the speed of recognizing faults is significantly improved.

A segmented temperature compensation bandgap reference with low output noise was designed based on the MGN 0.18um process in this paper. The circuit is powered by a 3.3V power supply, with a reference output voltage of 1.177V. In the temperature range of -45~125 °C and under the TT process angle, simulation results show that the temperature drift coefficient of the traditional structure can only reach 9.8 ppm/°C, and it is 1.08 ppm/°C for the bandgap reference using a new segmented temperature compensation. Compared with the traditional structure, the temperature coefficient is improved by nearly 10 times for the proposed reference generator. And also the amplifier used in this bandgap has chopping modulation, effectively reducing the impact of input offset voltage of the amplifier and 1/f noise of the bandgap. Simulation shows that at 100mHz, the bandgap equivalent output noise without the use of a chopping amplifier is 377uV/Hz, while the bandgap equivalent output noise with the use of a chopping amplifier is 29uV/Hz, with a noise reduction rate of up to 90%.

Based on the MGN18nmCMOS process, an off-chip capacitor-free LDO circuit is proposed and designed. The difference between the off-chip capacitor-free LDO and traditional LDO lies in the size of the output load capacitance. Because the on-chip load capacitance of LDO is small, it can be integrated into the chip, which greatly saves the area. But it has two problems: it is difficult to get the loop stability and the output voltage variation at load transients is large. In this design, active feed-forward frequency compensation is used to realize LDO loop stability. The output voltage detection circuit is used to provide the charging and discharging path for the power MOSFET and to get a fast regulation of its on resistance, resulting in fast regulation to output voltage without large variation at load transient. The simulation results show that the output voltage overshoot voltage and undershoot voltage are 538mV and 453mV respectively and can be stabilized in 30ns when the load current varies between 0 and 50mA in 10ns. Therefore, stability and response speed of the proposed capacitor-free LDO are greatly improved.

Based on the analysis of the operating principle of the stepper motor, an indexer structure combining the counter and digital-analog signal converter (DAC) is proposed and designed in this paper. Through the digital-analog signal converter, the digital signal output of the counter is converted into analog signal to accurately control the current flowing through the stepper motor. By the control of the indexer, six types of asynchronous long stepper motor working mode including the 1/32 subdivision, 1/16 subdivision, 1/8 subdivision, 1/4 subdivision, 1/2 subdivision and full step can be realized. Only a few number of control signals are needed to realize the switching of the working mode and the motor steering. The function and feasibility of the indexer are verified by simulation.

An overcurrent detection circuit for motor drive application is proposed and designed in this paper. The circuit utilizes the negative feedback clamp characteristics of operational amplifiers and off-chip high-resistance resistors to obtain an reference current with zero temperature coefficient current. This current flows through a small-size detection power mos transistor to generate a reference voltage used for motor driving over-current protection. The conduction voltage drop V_ds of the power transistor is transferred to a voltage on a sense resistor. The two voltages are compared, and the comparator flip point is the overcurrent point. The circuit is designed and simulated based on the 0.18μm MGN process. Simulation verification results show that under different PVT conditions, the overcurrent detection circuit has an accurate overcurrent point with error from -2.8% to +1.9%.

The natural conditions such as the solar irradiance intensity, temperature, cloudiness, etc, have brought about periodicity, randomness and uncertainty in photovoltaic (PV) performance characteristic. However, up to now, few studies focus on the integration of multiple factors. In addition, in the face of large-scale distributed photovoltaic array simulation, the high-efficiency computation n of scaled photovoltaics has not been deeply demonstrated. Therefore, this research optimizes the parameters of PV module and studies the relationship between irradiance intensity, temperature and the generation of photovoltaic. What’s more, on account of the PV array P-U characteristic curve appears repeatedly to change single-peak into multi-peaks under partial shading conditions which may cause the traditional maximum power point tracking (MPPT) algorithm fails to track the peak. An equivalent model for large-scale power generation with improved particle swarm optimization (IPSO) algorithm is proposed. Compared with traditional PSO algorithm , the IPSO method prevents the MPPT trapping in local maximum power point and finding the global maximum power point. Also IPSO accelerates the convergence speed and improves the algorithm stability by using the adaptive acceleration coefficients and inertia weight. The feasibility of this algorithm is verified based on Matlab simulation.

Partial Discharge (PD) is one of the manifestations of insulation deterioration in electrical equipment. In order to study the variation law of PD caused by air gap in cross-linked polyethylene (XLPE) cables, a PD simulation circuit was built in Simulink. Under the influence of different positions, sizes and power supply voltages of air gap, The variation characteristics of PD current are studied. The results show that the discharge current is related to the position, size and supply voltage of the air gap. When the air gap is closer to the conductor, the amplitude of discharge current is larger. The larger the air gap size, the larger the discharge current and discharge time. The greater the supply voltage, the greater the scope of PD. The results provide data reference for the follow-up research on PD localization.

Point clouds have played an important role in power transmission inspection. However, the huge data volume hinders the popularization and application of the technology. In this paper, the segmentation and visualization of point cloud data associated with power transmission lines have been addressed through an innovative approach, characterized by the integration of octree-based spatial partitioning and dynamic resampling. Initiated with the construction of an octree hierarchy, the point cloud data undergo segmentation, while a top-down strategy is concurrently enacted for the purpose of dynamic resampling, optimizing the efficiency of octree index construction. This is followed by the deployment of the octree index to selectively load point cloud data specific to individual transmission towers, culminating in a lightweight representation of the transmission line point clouds. The results derived from the research underscore the capability of the proposed methodology to substantially expedite the loading speed of transmission line point clouds, thereby offering a robust framework for the optimization of flight route planning during autonomous inspections of transmission lines.

To facilitate the automatic generation of power distribution line ledgers, a technique harnessing both Radio Frequency Identification (RFID) and deep learning is introduced. Drones equipped with RFID tools are utilised to capture information from power distribution pylons, culminating in the creation of a geographical map of the power distribution line, thereby providing essential data support for automatic ledger generation. Subsequent to this, various models are employed to detect images of power distribution line equipment, with the Yolov5-s model being specifically employed for the identification of equipment on distribution line poles. Detection outcomes are subsequently displayed on the geographical map of the distribution line, culminating in the automatic generation of the power distribution line ledger. Results from this study indicate that the introduced method can effectively and swiftly generate power distribution line ledgers.

This paper discusses OCR extraction techniques based on power metering inspection reports. With the rapid development of the electric power industry, the number of power metering inspection reports continues to grow, and the traditional manual processing method faces the challenges of inefficiency and error-prone. OCR technology has become an ideal choice for solving this problem, but the complex electric power symbols and units of measurement in the power metering inspection reports have brought new challenges to OCR technology. This paper details the application and principles of Convolutional Neural Network (CNN), YOLO algorithm in OCR extraction techniques for power metering inspection reports. CNN is used for feature extraction to integrate feature extraction, sequence modeling and transcription into a unified framework, while YOLO algorithm redefines object detection as a regression problem. These algorithms provide strong support for automated and intelligent processing of power metering inspection reports, which helps to improve data processing efficiency and accuracy.

In response to the issue of electrical interference in unmanned surface vessel compartments caused by poor electromagnetic compatibility of the antennas, a method of filling the compartment orifice with conductive material is proposed in combination with the structure of the unmanned surface vessel, to achieve the effect of shielding antenna electromagnetic interference. Firstly, by controlling the interfering variables, the interference source is located; then, based on the identification of the antenna as the interference source, the electric field strength of the antenna radiation at the equipment location is simulated and calculated. When rubber is used to fill the compartment orifice, the highest field strength inside the compartment reaches 50V/m. However, with the use of conductive materials for filling, the field strength at the equipment location is close to 0, achieving the effect of shielding antenna radiation interference and restoring normal operation of the compartment's electrical equipment. Finally, the effectiveness of this method is verified through experiments, and it is recommended to plan the distance between the unmanned surface vessel's antennas and equipment reasonably during the early stages of design.

This article aims to design an SRAM circuit with in memory operation function, and proposes a multiplication operation structure, including an in memory operation storage unit and an output pre charging circuit. This structure needs to be coordinated with other modules of the circuit for in memory calculation. The memory operation unit includes a transmission unit and a 6T-SRAM cell. The 6T-SRAM cell is used for writing weights, and the transmission unit is used for performing multiplication operations between weights and input signals, and transmitting the calculation results to the output pre charging circuit. This scheme can reduce the number of transistors required for in memory computing, thereby achieving the goal of reducing circuit area.

On the basis of classical theoretical analysis of transmission line parameters, a simulation model of dual circuit four split 500kV transmission line is constructed. When the parallel length ratio of the double circuit line varies from 10% to 100%, and the non-measurement line is grounded at both ends, Obtain the variation pattern of the zero-sequence impedance of the measured line, and obtained the general range of magnetic effects under the conditions of parallel operation of two single circuit lines. In addition, taking actual transmission lines as examples, the applicability of the model to conventional transmission line parameters is demonstrated, which has a certain guiding role for practical engineering.

An improved YOLOv5 for PCB defect detection named as YOLO-PDD is proposed in this paper. Based on this, a circuit board defect detection system which can achieve high accuracy and fast detection of small target defects under complex background noise interference is designed. The industrial cameras and LED light sources are used in the system to capture images of circuit board defects, forming a self-made original dataset. At the same time, stitching technology is adopted to enhance the original dataset. YOLO-PDD has made the following improvements to the original YOLOv5 algorithm: Firstly, a double-layer routing attention module BRA is introduced in the feature extraction network. Next, the backbone of YOLOv5 algorithm is replaced by DenseNet. Finally, a feature pyramid network which is bidirectional and weighted in nature called Bifpn is added to the YOLOv5 feature fusion network. The circuit board defect detection system software is built based on QT and can obtain real-time detection results and data statistics online. Experimental results show that the system can accurately detect defect targets in circuit board images at a speed of 47ms per image, achieving 92.28% mAP on our self-made dataset, which is superior to the average detection accuracy of current state-of-art original detection algorithms and has good practicality and effectiveness.

In order to effectively improve the accuracy of transient fault detection in smart meter circuits and minimize the outage time of the power grid, a transient fault detection method for smart meter circuits based on continuous wavelet transform is proposed. Analyze the basic circuit structure of the smart meter power module, clarify the power information fluctuation of the input side and output side when the circuit transient fault occurs; based on the equivalent electric field environment, collect the original signal of the smart meter circuit fault when the transmission direction of the smart meter circuit signal wave deviates; use the Butterworth filter to filter the original signal, and obtain the transient fault signal; based on the continuous wavelet transform, extract, Based on the continuous wavelet transform, extract and decompose the feature components in the signal, and output the spatial mutual information containing the continuous wavelet components (transient fault feature signal) to realize the detection of transient faults of smart meter circuits. The experimental results show that the proposed method can effectively obtain the power output information of shallow faults of the circuit, and has better transient fault detection effect. With the increasing number of iterations, the average value of monitoring accuracy of the proposed method is 0.924, and the highest value of monitoring accuracy is 0.979, which shows that the proposed method has high accuracy in detecting transient faults of smart meter circuits, and it can effectively recognize the characteristics of transient faults of circuits.

There is a diversity of grounding faults in distribution network lines, and different types of faults may lead to different states and signal changes. Therefore, a fault state grounding diagnosis method for distribution lines based on waveform similarity and random matrix is proposed. The principle of waveform similarity is used to determine whether the single-phase fault current is abnormal. By using the random feature matrix, the time series of the original fault data set is calculated, and the location of the fault point and the specific category of transmission fault is determined by identifying the fault spectrum of the data current and voltage signal. The single-terminal method and double-terminal method are used to diagnose the single-phase grounding fault of the distribution network accurately. The experimental results show that there is no confusion among the various fault features extracted by this method, and the application effect is better.

A single-phase ground fault on the valve side in modular multilevel converter high-voltage direct-current (MMC-HVDC) may cause commutation faults that can lead to commutation inrush current. The traditional second harmonic braking differential protection method will be invalid at this time. In order to solve the above problems, this paper proposes a new criterion for recognizing the excitation inrush current and fault current of MMC-HVDC converter based on the Hausdorff distance of continuous window, first convert the delta side current to the star side in which and then eliminate the zerosequence current from the original star side current, and then use it as the elements of sets A and B for Hausdorff distance calculation. When the Hausdorff distance is below a certain threshold for 1.5 milliseconds, the excitation inrush current due to out-of-area faults is happened. Otherwise, it is judged as the fault current due to faults in the area. Finally, the validity of the proposed criterion is verified by simulation based on PSCAD/EMTDC.

The distribution network in China usually operates with ungrounded neutral points. Installing suitable arc suppression coils on distribution lines can effectively control the range of influence of single-phase grounding faults. To configure arc suppression coils with reasonable compensation capacity, accurate measurement of the distribution line is necessary. However, current capacitance current measurement methods often do not consider the impact of zero sequence voltage on the measurement results. This article proposes a new method for measuring capacitive current. Optimized the wiring method of traditional signal injection method. A line to ground capacitance analysis model considering the zero sequence voltage of the power grid was constructed. By injecting multiple sets of abnormal frequency signals into the open triangular end of the voltage transformer, the capacitance of the line to ground can be effectively measured and analyzed. The method in this article does not require measurement of the system phase angle, and the required instruments and equipment are simple and convenient for on-site operation. This new type of abnormal frequency signal injection method has guiding significance for the prevention and control of single-phase grounding faults in distribution lines.

Due to its remarkable theoretical energy density and capacity density, the Al-air battery is regarded as the most competitive potential candidate for next-generation energy storage systems. However, attributed to its enormous liquid electrolyte storage and transportation infrastructure, the conventional Al-air battery is not appropriate for miniature applications. To get around the issue of pumps providing a continuous electrolyte flow for AABs, we created a microfluidic Al-air battery. Without pumps, the special capillary action in the paper tubes could carry electrolytes constantly and automatically. This battery produced a remarkable output of 16.74mW/cm² and an open circuit potential (OCP) of more than 1.6V using an alkaline electrolyte. The energy density and capacity density of this battery were 350.96mWh/g at 20mA/cm² and 963.31mAh/g, respectively. As a result, we suggested a commendable plan for compact and basic Al-air cells as reliable power sources.

Traditional transformer on-load tap-changers produce an arc when switching the winding tap action, and the arcing exotherm can affect the insulation performance of the contacts and accelerate the deterioration of the transformer oil, which ultimately leads to a significant reduction in the transformer’s voltage regulation performance and affects the safe operation of the transformer. This paper analyzes the thyristor gate characteristic curve in detail to find out the region of reliable driving thyristor; on this basis, combined with the design of ideal driving signal index, it puts forward the method of calculating and selecting the parameters of thyristor driving circuit based on the gate characteristic curve; finally, it constructs the thyristor-type electronic diverter switch to carry out the experimental validation, and it provides a reference scheme for designing and selecting the key components of the thyristor driving circuit in the power electronic on-load tap-changer.

With the rapid development of China’s economy, the demand for electricity is continuously increasing, leading to increased complexity in the power system. To process the exponentially growing data in the distribution network in real-time and accurately assist the power industry in making prompt decisions, this paper proposes a distribution network auxiliary decision-making system. It utilizes distributed computing to construct a streaming computing framework based on Storm, implementing the Very Fast Decision Tree (VFDT) algorithm, and a batch processing computing framework based on Spark, implementing the Random Forest algorithm. These frameworks enhance the scalability, real-time capability, and accuracy of big data analysis in the power grid. Through simulation experiments, we validate the efficient and accurate classification of various types of large-scale data using the VFDT algorithm based on Storm. Additionally, we verify the effectiveness and good scalability of the batch processing computing framework based on Spark for handling massive data processing.

With the rapid development of home energy management systems (HEMS), forecasting electricity loads is increasingly important for the operations of energy storage and usage in residential households. However, household electricity loads can be extremely uncertain due to the complex consumption behaviors of residents, posing challenges for existing hours-ahead forecasting methods. Moreover, there is a scarcity of research considering loads and weather conditions at different time scales for probabilistic load forecasting (PLF) methods, thus forecasts are dependent on trends of a limited spectrum. In this study, a decomposition-based method is proposed for the short-term probabilistic load forecasting (STPLF) of household electricity loads. Specifically, historical data are decomposed into trend components of different time scales using a novel structure. Based on these components, quantile forecasts are produced respectively and synthesized into total probabilistic forecasts for loads hours ahead. The effectiveness and superiority of the proposed method are demonstrated by the simulation results on household electricity loads.

The frequent occurrence of urban waterlogging seriously affects the safe operation of distribution network. The lack of differentiated planning for the waterlogging prevention of distribution facilities in the existing codes of distribution network infrastructure causes a large number of distribution facilities to be inundated by rain in extreme storm weather. Taking Guangde City of Anhui province as the research area, taking the ring grid cabinet of 10 kV distribution network as the research object, based on the local geographic information data, the distribution network location information data and the historical extreme rainfall disaster data, a two-dimensional unsteady FloodArea model was used to analyze the risk of flooding caused by short-duration rainfall in historical extreme weather, draw the 10kv distribution terminal inundation risk distribution map in Guangde City, and evaluate the equipment inundation risk. The results show that the FloodArea model can accurately simulate the extreme heavy rainfall process, and the simulation results are important for reducing the submergence risk level of existing distribution facilities and distribution network planning.

With the continuous emergence of new technologies in wind farms, in order to improve the performance of wind turbines, it is necessary to carry out technological transformation of wind turbines that have been put into operation and are relatively backward in technology. This paper proposes a technical and economic evaluation method of wind farm renovation based on the combination assignment method. Firstly, the evaluation index system of wind power project is designed from the two dimensions of wind turbine operation efficiency and economy, secondly, the combination of hierarchical analysis method and entropy value method is used to determine the weight of each index, and then the renovation program is evaluated based on the comprehensive evaluation model, and finally, the feasibility and validity of the proposed evaluation method is shown through empirical analysis.

In order to comprehensively assess the efficiency of wind farms and meet the needs of the development of China's wind power industry, this paper proposes a technical assessment model based on the comprehensive efficiency of the units on the basis of recognizing the current status of research on the energy efficiency assessment of wind farms, and introduces the assessment index system and the method of determining the weights. The effectiveness and practicability of the method are verified by analyzing the arithmetic examples of two wind farms.

At present, the constraint mechanism of multi-source fusion of distribution network operation data containing new energy is generally set as a single form, and the data fusion efficiency is low, resulting in an increase in the average absolute percentage error. Therefore, the design and analysis of multi-source fusion method of distribution network operation data containing new energy based on joint Kalman filter is proposed. According to the current measurement, first collect the characteristics of distribution network operation data, set up multi-source data fusion levels, adopt a multi-level approach, improve the overall efficiency of data fusion, and build a multi-level constraint mechanism. On this basis, a multi-source fusion model of distribution network operation data with new energy based on joint Kalman filter is constructed, and multisource data fusion is realized by mapping auxiliary processing. The final test results show that the average absolute percentage errors of the four randomly selected test areas within the coverage of H distribution network are well controlled below 1.2 after three cycles of measurement, which indicates that the designed multi-source fusion method of new energy distribution network operation data is more reliable and efficient with the help of the joint Kalman filter technology, The degree of data fusion is high, targeted, and has practical application value.

As a new mode of aggregating demand-side resources such as electric vehicles and air-conditioning loads, the virtual power plant can fully utilize the advantages of fast response speed, large frequency regulation potential and high regulation flexibility of demand-side resources. In this paper, we take the smart building-type virtual power plant aggregated with electric vehicles and air-conditioning loads as the research object, quantitatively analyze the users' willingness to frequency regulation using TSK fuzzy model, and establish a power allocation model considering the users' willingness to realize the intra-day frequency regulation control with the participation of demand-side resources in the frequency regulation auxiliary service. Finally, the effectiveness of the proposed strategy is verified through an arithmetic example.

The rapid development of renewable energy and the widespread adoption of electric vehicles (EVs) have brought new opportunities and challenges to the power system. This paper proposes a multi-timescale optimization method for virtual power plants involving various distributed resources such as electric vehicles, renewable energy, and energy storage systems. The method is based on the Second-order Cone Relaxation Alternating Current Optimal Power Flow Model (SOCR-ACOPF). By considering the coordination and cooperation among various components on both daily and intra-day timescales, energy optimization and management for virtual power plants are achieved. The application of second-order cone relaxation techniques to solve mixed-integer nonlinear programming problems effectively enhances the efficiency of optimization solutions. Finally, the proposed model's effectiveness and practicality are demonstrated through simulation experiments conducted on an IEEE 33-node distribution system.

A new non-invasive power monitoring method based on two-layer clustering and multi-task learning is proposed to solve the problems of power monitoring at microgrid level. Firstly, the historical operating data of microgrid users are pre-clustered based on density clustering algorithm, and then the membership matrix is obtained by fuzzy clustering. Then, a power identification model is constructed based on clustering results, and the shared layer of the model is used to extract the influence of external factors such as climate on the microgrid. Finally, the non-invasive power decomposition and monitoring of the microgrid are completed based on identification model and independent component analysis. The feasibility of the proposed method is verified by a case study of an actual integrated energy microgrid.

With the proposal of the "dual carbon" strategy, the development of China's new energy vehicle industry has ushered in new opportunities. This article focuses on the charging business scenarios in microgrid parks, relying on digital technologies such as artificial intelligence and big data, and deeply explores the value of data. By constructing a two-stage model of YOLOv5+LPRNnet, electric vehicle recognition is achieved, improving recognition accuracy while ensuring recognition speed, and achieving automated and standardized management of vehicle entry and exit charging stations, Effectively solving the problem of occupying space for gasoline vehicles, empowering business development with data, and significantly improving operational service efficiency.

Numerous studies have demonstrated the effectiveness of integrating fog computing with smart grids to reduce data latency, leveraging the low latency characteristics of fog computing. Within this resource-rich and emerging field, privacy-preserving data aggregation emerges as a critical service. However, fine-grained recording and transmission of electricity consumption data and related user data are still inevitable, including user privacy security, which is the biggest challenge faced by smart grids. Therefore, user privacy protection and communication security during the above user data aggregation stage have become crucial. This study introduces a novel and efficient method for secure data aggregation that preserves user privacy. By leveraging the BLS short signature technique, we ensure the authentication of individual user contributions. Concurrently, we harness the Paillier cryptosystem's additive homomorphism feature to encrypt the data, thus maintaining the confidentiality of users' sensitive information throughout the aggregation process. After conducting a thorough examination of both security and performance, our approach safeguards user privacy and the accuracy of data throughout the consolidation phase, all without significantly increasing computational load.

With the widespread integration of distributed power sources, analyzing the carrying capacity of distribution networks holds significant importance. This paper firstly approaches from a topological perspective, establishing a carrying capacity adjacency matrix based on improved nodal importance and maximum flow theory. Secondly, it defines the topological structure sensitivity according to the variation degree of the adjacency matrix, thereby analyzing the impact of distribution network topology on carrying capacity. Finally, by simulating the IEEE 33-node test system, it validates that the proposed method not only analyzes the carrying capacity of distribution networks but also ensures the safety of network topology, meeting the current development requirements of distribution networks.

In addressing the task allocation problem for an anti-UAV swarm protection system for power transmission lines, a solution based on swarm intelligence optimization algorithms is proposed. While ensuring the accomplishment of anti-drone tasks, an optimization objective is introduced, incorporating energy conservation and loss reduction. Simultaneously considering factors such as strike effectiveness and time constraints, a mathematical model for task allocation optimization is formulated. Subsequently, an improved optimization algorithm, namely the iPSO-GA algorithm, is introduced. By enhancing the Particle Swarm Optimization (PSO) algorithm and integrating it with the Genetic Algorithm (GA), this approach ensures a balance between convergence and diversity in the evolutionary process of the particle swarm. Compared to existing PSO algorithms, it exhibits superior global search capabilities. Simulation experiments demonstrate that this algorithm effectively enhances the task allocation capability of the anti-drone protection system, ensuring the safety of power transmission lines.

After the large-scale integration of wind farms into regional power grids, traditional high-frequency cutting generator strategies no longer fit the new way of grid operation. This paper conducts a specialized study on isolated power grids containing both hydro and wind power that may experience high-frequency issues. It fully considers the impact of wind farms and the problem of over-shedding in large-capacity single-unit generators. Combining the characteristics of the isolated grid and the properties of different units, an optimized method for high-frequency load shedding strategies is proposed. Simulation analysis is carried out, achieving good results.

With the deepening of the new round of power system reform, market players are more independent in their choice of suppliers, pricing and trading methods, and the characteristics of systematic, diversified and competitive electricity retail market are prominent. Electricity retailers need to be more scientific in their pricing to ensure they are competitive in terms of market share and profits. Therefore, a probabilistic model considering customer demand response and a two-stage robust model considering power demand uncertainty are proposed in this paper. The upper model sets the TOU price with the goal of maximizing the benefit of electricity retailers, while the lower model considers the uncertainty of electric vehicles. The simulation analysis of IEEE33-node power system shows that the two-stage robust model can guide the user's power consumption decision by adjusting the TOU price, balance the uncertainty of electric vehicles, and improve the benefit of electricity sales.

Efficient and accurate short-term load forecasting is an important guarantee for the safety, stability, and economic operation of the power system. Aiming at the problem of large prediction errors in peak and valley loads, a prediction model based on feature selection and neural networks is proposed through in-depth analysis of the working principles of feature selection and neural networks. Considering seasonal factors, the load was divided into four seasons: spring, summer, autumn, and winter. Pearson correlation analysis and Spearman correlation analysis were used to comprehensively analyze the climate and electricity price characteristics of the four seasons. A hybrid neural network combining convolutional neural network CNN and long short-term memory network LSTM was used. Using CNN to mine features between data to form new feature vectors, and then inputting this feature vector into LSTM for prediction. Simulate and verify the load forecasting model based on feature selection of CNN-LSTM using a power load dataset, and compare it with the full feature LSTM and feature selection LSTM models. The results show that there is a significant improvement in the accuracy of load forecasting after feature selection and using CNN models.

In order to implement the "dual carbon" goal and sustainable development strategy, this paper proposes a method for interval prediction of wind power through Bayesian-optimized bidirectional long short-term memory network. By combining the existing deep learning algorithms, the CEEMDAN-BO-QRBiLSTM model was constructed. The empirical analysis results show that the newly established model is feasible and has higher accuracy and better effect than other existing models.

To cope with the increasingly complex port power system scheduling problems, accurate port load forecasting is required. In this paper, a new port load forecasting method is proposed on the basis of the previous work by adopting the limit learning machine method based on the optimisation of simulated annealing algorithm. Compared with the ordinary limit learning machine method and time series prediction method, this method is faster and more accurate, and more suitable for engineering applications.

With the transformation of energy structure and large-scale access to renewable energy sources, the operation of the power system is facing more and more challenges. In the context of achieving the goals of carbon peaking and carbon neutrality, this paper uses linear regression method to forecast the power demand and new energy generation, uses linear programming method to establish the model of power generation system considering low carbon, and at the same time constructs the combination method of transmission channels of the whole network, optimizes the grid scheduling, and builds the multivariate energy storage system, to study the method of coordinated operation of the power system with source, network, and load/storage in the power system considering the carbon constraints. Based on the research method in this paper, the overall operation scheme of power system source-grid-load-storage coordination in compliance with carbon constraints can be realized in the target provinces and target power grids.

Due to the introduction of a large number of distributed resources and other new elements, the distribution network presents new characteristics that are different from traditional power grids, thus posing new requirements for the planning work of the distribution network. The future distribution network planning needs to meet the requirements of primary and secondary collaborative coordination planning. Collaborative coordination is reflected in the formulation of planning schemes, that is, while carrying out the primary network planning, it is necessary to simultaneously consider the configuration of relay protection, safety automatic devices, as well as distribution automation and communication configuration, so as to coordinate the primary and secondary schemes. This article mainly introduces the main steps of modern smart distribution network planning technology, and provides an example application of collaborative planning for distributed power generation access. The adoption of primary and secondary coordinated planning helps to achieve the development goals of modern distribution networks that are safe, reliable, cost-effective, clean, and low-carbon.

In the context of smart grid and new power system, distributed generations, energy storage systems, electric vehicles and other new power equipment are connected to the distribution network on a large scale, which makes the topology of the distribution network change more frequently. The traditional topology identification methods of the distribution network are limited by measuring equipment, and difficult to apply. Aiming at the difficulty of obtaining real-time topology of distribution network, this paper proposes a topology identification based on feature selection and Convolutional Neural Network. Firstly, XGBoost is used to calculate the feature importance of node voltage amplitude, and then modified according to the correlation analysis results to select topology identification features. This step can be used to reduce the time and space complexity of the model. Secondly, a distribution network topology identification model based on Convolutional Neural Network is constructed. The correlation between the feature data and the topology is learned by using the operation characteristics of convolutional and pooling and the characteristics of the model structure. Then the mapping rules are established. Finally, IEEE33-node distribution network is used to verify the effectiveness of the proposed method in feature selection and topology identification.

In order to maximize the revenue of electric selling companies, this paper establishes an optimal scheduling model of electric vehicles which takes into account the charging cost and scheduling deviation of electric vehicles. In the constructed model, electric selling companies guide electric vehicle owners to participate in real-time charging and discharging through real-time electricity price, and use evolutionary game to build an interactive income model between electric selling companies and electric vehicle users, so as to obtain the charging and discharging prices under the interaction and cooperation between electric selling companies and electric vehicle owners, providing reference for electric selling companies to implement optimal scheduling strategies motivated by real-time electricity price.

This paper presents a composite inverter circuit for Wireless Power Transfer (WPT) systems, addressing the limitations of reconfigurability in traditional inverters, and catering to the evolving demands of multi-port power converters. The proposed circuit consists of parallel-connected Buck converters, demonstrates versatility by seamlessly functioning as a DC source, or an inverter power source at the primary side of WPT systems. Simulation results conducted using PLECS software validate the proposed inverter's suitability for WPT systems. The circuit demonstrates comparable output performance to a multi-level bridge inverter capable of both voltage reduction and boost, also allowing for DC-DC module reuse in multi-port power electronic converters.

Severe convective weather is often accompanied by extreme weather conditions such as storms, resulting in significant instability of wind speed and temperature parameters. This instability can lead to large changes in airflow patterns in the wind field, making it more complex to predict fan response and power output. Therefore, accurate power prediction of wind farm under severe convection weather is of great significance to economic dispatching and safe and stable operation of power system. In order to make full use of the effective information from multiple data sources to further improve the prediction accuracy of wind farm generation power under severe convection weather, a generation power prediction method based on TabNet model is proposed. The influencing factors of wind power under severe convection weather are analyzed, and the TabNet model is used to achieve feature selection, and the TabNet model is used to predict wind power under severe convection weather. The findings indicate that the suggested approach can successfully utilize multi-variable time series data during intense convective weather conditions to forecast wind farm power generation, demonstrating a high level of prediction accuracy.

Predicting the oscillation evolution of hydraulic turbines is crucial to ensuring their safe and stable operation. However, accurately and effectively predicting the oscillation evolution of hydraulic turbines can be challenging due to their non-stationary and nonlinear characteristics. Therefore, this article proposes an oscillation evolution prediction model for hydroelectric units based on Optimal variational modal decomposition-Bilateral gated recurrent neural networks (OVMD-BIGRU). Firstly, the decomposition parameters of Optimal variational modal decomposition (OVMD) are determined using the central frequency observation method and the residual index minimization criterion, avoiding subjective factors. Then perform Variational modal decomposition (VMD), normalize each modal component, and establish a BIGRU model for prediction. Finally, reverse normalize and stack the modal results to obtain the final predicted vibration trend of the unit. This article designs a comparative experiment based on the unit data of a domestic hydraulic power station, and the findings indicate that the proposed model has good performance and high prediction accuracy.

This paper applies different machine learning-based methods for equipment renewal scale prediction in power systems, and compares the predictability of different methods by measuring efficiency and accuracy. Focused on equipment renewal of power enterprises, this paper uses historical data to model training and dynamic prediction of the equipment renewal scale in the future, and obtains an optimal prediction model. The main contributions and innovations of this paper are as follows: (1) A machine learning-based method for predicting the equipment renewal scale of power enterprises is proposed, which can dynamically predict the future equipment renewal demand according to the operation data and equipment data of power systems, thus helping the equipment management and investment decision of power enterprises. (2) By comparing prediction performances of the machine-learning based model and the alternative methods, this paper provides a benchmark for prediction model selection in the field of power system. The comparison is conducted among ordinary least squares, vector autoregressive model, tree model and multilayer perceptron neural network model, and finds out that the neural network model has obvious advantages in predicting the accuracy of the equipment renewal scale of power enterprises. Therefore, the neural network model is more suitable for subsequent prediction.

Accurate predictions of electricity load are crucial for the safe and stable operation, as well as the economic performance, of power systems. To overcome the inadequacy of feature selection and the difficulty in tuning model hyperparameters in traditional load forecasting models — issues that may lead the predictive model into local optima — this paper proposes a short-term electricity load forecasting method based on XGBoost-IARO-LSTM.Initially, the eXtreme Gradient Boosting (XGBoost) tree model is employed for feature selection. Subsequently, the Improved Artificial Rabbits Optimization (IARO) algorithm is enhanced using tent chaotic mapping, elite reverse learning strategy, and hybrid reverse learning strategy. The improved IARO algorithm is then utilized to tune the hyperparameters of Long Short-Term Memory (LSTM).Finally, the XGBoost-IARO-LSTM combined forecasting model is established to achieve accurate short-term load predictions. The proposed method is validated using real electricity load data from a specific region in northern China. The results indicate that the proposed approach exhibits faster parameter optimization speed and superior prediction performance.

When providing services for distribution network users, the throughput of expanded data of user industry is low, which leads to long data extraction time. Therefore, an incremental data extraction model for expanded data of distribution network users based on improved IAA algorithm is proposed. Based on the sparsity analysis results of the data of distribution network users' business expansion and installation, the improved IAA algorithm is used to incrementally learn the original data, import it into the data warehouse, read multi-source data, build an incremental data extraction model, compare the data reading speed and update speed of the distribution network users' business expansion and installation system, determine the data extraction time window, and extract incremental data. The experimental results show that the data throughput of this model and the time-consuming part of data extraction are improved, which provides a new idea for incremental data extraction.

Based on the theoretical framework of Pearl's causal model, this study conducted research in three main areas to address the lack of work on causal model construction in the urban fire domain. First, at the association level, the PC (Peter-Clark) algorithm was used to construct an initial causal network model in the urban fire domain to establish causal relationships between relevant factors. However, due to potential missing directed relationships between casual network nodes, the network structure was supplemented based on knowledge provided by domain experts. Second, at the intervention level, the study applied the backdoor criterion and the front-door criterion to estimate the causal effects between fire variables in the above causal network, and determined the variable nodes with the greatest causal effect through three main variables and the final direction of the network. Finally, at the counterfactual level, the study tested and verified the constructed causal network structure using methods such as adding random confounding factors, placebo interventions, and data subset validation. The experimental results show that the proposed urban fire causal network obtained by the framework of causal analysis has a certain robustness and interpretability, demonstrating the effectiveness and reliability of the methods used.

Power system faults bring serious threatens and challenges to the safe and stable operation of the system. The traditional methods of generating predictive fault sets mainly rely on manual experience and the generated fault sets are often mismatched with the system safety characteristics. Motivated by this issue, this paper proposes an automatic fault set generation method based on Attention-Bi-LSTM and XGBoost. Firstly, a system operation risk model is established to calculate power equipment fault severity and fault risk. Then utilize the Attention mechanism to enhance Bi-directional Long Short-Term Memory (Bi-LSTM) model and combining it with eXtreme Gradient Boosting (XGBoost) model to construct a probabilistic combined prediction model for the risk of power dispatching faults with weights determined by the error reciprocal method. Finally high-risk fault sets are then automatically generated based on the obtained predictions in terms of risk probability and severity magnitude. The proposed method is verified in a practical power system in China and the results show that the fault sets generated can better reflect the actual system operating state, improve the fault prediction efficiency and provide support for the safe and stable operation of the power system.

To solve the problem of distribution network line-broken fault, a judgment method for distribution network line-broken fault based on node balance information entropy is proposed. Firstly, a feeder node relationship information matrix model was established to describe the topological connection relationships and node types of each node in the distribution network feeder. Then, the node balance information entropy was established to describe the voltage imbalance situation of each node. Secondly, the feeder node balance information entropy relationship matrix is generated through the feeder node relationship information matrix and node balance information entropy. Finally, the line-broken fault area was located through the feeder node balance information entropy judgment matrix. Through example analysis, it is shown that this method can locate line-broken faults in the distribution network based on existing terminal collection information. The method is simple and reliable, can meet the needs of on-site applications, and has good promotion value.

The lawsuiting requirement of galvanic carriage users is affected by a variety of factors, including working days, seasons, weather conditions, user journeys, etc. These factors make the lawsuiting mode and time of galvanic carriages diverse and uncertain, making load prognosis difficult. Therefore, this article proposes an EV lawsuiting load forecast measure based on requirement comportment analysis and random forest algorithm. K-mean clustering is applied to analyze the comportment impress of lawsuiting load requirement. Because the driving of galvanic carriages will be affected by weather, road conditions and other factors, it is difficult to predict. Thus, it is imperative to recover the mixed impress in the traffic sequence. Based on this, the random forest blueprint is constructed to classify the lawsuiting load and complete the prognosis of the lawsuiting load. The experimental results show that the load forecast results of galvanic carriages have a lofty extent of fitting with the tangible value. The time is shorter and the application is better.

The conventional simulation analysis method for the lightning distribution characteristics of wind farms mainly uses the moment method to calculate the lightning distribution parameters, which is easily affected by the dynamic changes of a lightning waveform, resulting in a significant deviation between the simulated current amplitude and the actual current amplitude. Therefore, it is necessary to design a new method for simulating and analyzing the lightning distribution characteristics of wind farms based on spatiotemporal interpolation and Bayesian networks. Based on the spatiotemporal interpolation method, a lightning transient feature model for wind farm sites was constructed, and the simulation parameters of lightning distribution characteristics at wind farm sites were calculated using Bayesian networks. The example analysis results show that the deviation between the current amplitude and the actual current amplitude is small, which proves that the designed simulation analysis method has good analysis results, reliability, and application value, making a certain contribution to improving the operational safety of wind farms.

Power load forecasting is a key problem in power system operation and planning. As an effective method, grey correlation analysis has the advantages of overcoming data fuzziness and establishing incomplete models and can help determine the main factors affecting power load and improve the accuracy and reliability of prediction. Therefore, it has important background and significance in electric load forecasting research. This study uses the grey correlation analysis method to predict and determine the main factors affecting the power load by collecting power load data and possible influencing factors data. It is found that weather conditions, seasonal factors, holidays, and policy factors are the most influential factors on power load, and they have a high correlation with power load. In particular, seasonal factors have the most significant impact on power load. The results of this study provide important references for power system operation and planning and have guiding significance for the rational arrangement of power supply and demand and formulation of dispatching strategy.

During the charging process of electric vehicles, additional loads will be imposed on the power system, the stability and reliability of power supply cannot be guaranteed. In order to optimize the location of the charging station, an electric vehicle charging station selection algorithm based on the power demand side response model is proposed. Calculate the demand side demand for electricity, and based on the demand side response model, construct a mathematical model for selecting electric vehicle charging station stations and solve the model. The feasibility and effectiveness of this algorithm have been verified through experimental results analysis.

Aiming at the problem of low prediction certainty of existing load forecasting methods, this paper proposes a short-term load forecasting system that is based on the LSTM neural network. The original dataset is first processed by the median elimination method, Markov-Monte Carlo algorithm and improved hierarchical clustering algorithm. Then, the data are input into the LSTM prediction model optimized by the Adam method and SGD method for prediction. The example analysis proves that the method proposed in this paper has better prediction accuracy, which is helpful for the planning and movement of the power system.

The integration of distributed generators (DGs) poses great challenges to the topology, operation planning, control methods, and protection configuration of traditional distribution networks (DN). To solve the optimization control problem of large-scale multi-type DGs, energy storage systems (ESSs), and multi-type loads in DNs, this paper proposes a distributed active and reactive power optimization (ARPO) control method for active distribution networks (ADN). Firstly, the ADN is modeled, taking into account the reactive power (RP) support effects of DGs and ESSs inverters and the suppression of active power (AP) fluctuations in the ESSs, in addition to various grid operation constraints. Afterward, based on the established model, the improved Harris Eagle optimization algorithm (IHHOA) is used to optimize the AP and RP of the ADN to minimize economic efficiency (including operating costs and network loss costs) and voltage fluctuations. The proposed model and method not only ensure the economic operation of the ADN but also improve the voltage quality of the ADN and the self-consumption level of DGs. Finally, simulation verifies the effectiveness and feasibility of the proposed control method.

With the rapid growth of renewable energy resources, the structure of microgrids is becoming more and more complex, rendering the traditional centralized method more computationally intensive and poorly scalable. In order to realize the economic optimization operation of microgrids in a distributed way, this paper proposes a two-level decomposable penalty method based economic dispatch algorithm (DPED). The first level transforms the complex constraints into penalty terms of the objective function and simplifies the optimization problem with simple constraints. The second level adopts the forward-backward splitting method to decompose the complex objective function into multiple simple subproblems, which is convenient for the computer to solve in parallel. Finally, simulation results show the effectiveness of the proposed algorithm.

The source network load and storage data of Ultra High Voltage (UHV) power grid is very large, including massive real-time monitoring data and equipment status data. The backup of these data needs to occupy a lot of storage space, and the transmission and backup process may put forward high requirements on the network bandwidth and transmission speed. Therefore, an active de-duplication backup algorithm based on improved swarm algorithm is proposed for UHV power grid source data storage. The bee colony algorithm is improved and optimized as artificial bee colony algorithm. Simhash algorithm is introduced to remove load and storage data of UHV power grid. Based on this, the balanced binary tree is used to realize its backup. The experimental results show that: Compared with the traditional method, when using the research method to backup data, the storage space is minimum, the data deduplication rate is higher than 95%, and the time is always less than 100ms.

A proposal is put forward for an adaptive control method for composite energy storage in smart distribution networks, which utilizes a convolutional neural network to achieve accurate control over the bidirectional charging and discharging of energy. Utilizing convolutional neural networks to train composite energy storage parameters, introducing softmax classifiers to identify the discharge state of composite energy storage, simulating energy storage capacity, light intensity, and temperature as inputs to the convolutional neural network, and using genetic algorithms to solve the output value of composite energy storage control, achieving adaptive adjustment of composite energy storage in distribution networks. Based on the experimental findings, it has been demonstrated that the suggested approach can adaptively modify the temporal arrangement of power energy considering fluctuations in seasons and temperature. This enhancement significantly enhances the operational flexibility and dependability of composite energy storage in distribution networks.

With the digitalization of power grids and the development of new power systems, more and more IoT terminals are used in various scenarios of power grid business transmission and distribution, so as to realize full perception of system status and full service penetration in all links of electric energy production, transmission and consumption, and ensure the high-quality construction of power grids. The explosive growth of the number of IoT terminals (such as sensor devices, inspection robots, etc.) has transformed the system IT infrastructure from "boundary" to "borderless", and the network security boundary is gradually disintegrating, so the continuous anomaly detection security of power IoT terminals connected to the power grid system is becoming more and more important. To solve this problem, this paper proposes an anomaly detection method for power IoT terminals in a zero-trust environment. In this method, a BSREM credibility evaluation model based on behavioral sequence is designed to complete the anomaly detection task of IoT terminals, improve the information learning ability of the model by maintaining the information matrix, and expand the adaptation scope of the method. Finally, this paper uses simulation comparison to verify the BSREM model and LightLog model to evaluate the credibility of the device's traffic behavior and activity log. The results show that the BSREM model can complete the credibility evaluation task more efficiently.

Power quality in renewable energy generation-based power systems faces challenges like complicated generation mechanisms and evolution law, which turns power quality comprehensive evaluation into significant work. This paper proposes a power quality comprehensive evaluation method based on fuzzy clustering and the analytic hierarchy process (AHP), which takes opinions from the load into account. Firstly, a comprehensive power quality evaluation index system is established based on the power quality standard and load opinion evaluation index. In addition, the evaluation index levels are classified. Secondly, the fuzzy clustering analysis method is used to evaluate each power quality index and generate dynamic data sources clustering to determine the best threshold of a single index. Lastly, the power quality comprehensive evaluation model is developed considering power quality opinions from the load side and their evaluation index weights. The engineering data calculation example verifies the accuracy and feasibility of the proposed power quality comprehensive evaluation model.

This paper designs an active identification platform for critical risks in power systems based on multi-objective decision models to improve the safety and stability of the power system. Based on the design of a multi-objective decision-making model, a framework for the active identification of critical risks in the power system was designed. This framework consists of a data source acquisition unit, a data conversion processing unit, and a data storage unit. After designing different units, an active identification algorithm for critical risks in the power system was designed to effectively identify potential risks. The performance and effectiveness of the platform were verified through experiments. The experimental results show that the platform can accurately identify key risks in the power system and provide corresponding warning and decision support, effectively improving the safety and stability of the power system.

Aiming at the problems of solar curtailment and wind curtailment in the integrated energy system, an optimal control method considering the optimal consumption of clean energy is proposed. Regional Integrated Energy System (RIES) usually adopts multiple consumption technologies to solve problems such as curtailment of light and wind. On this basis, this paper establishes a RIES optimization model including Air Source Heat Pump (ASHP), photovoltaic units, wind turbines and energy storage systems. The source side adopts a robust optimization model including ASHP, which can not only remove the issue with wind energy output instability, but also expand grid for wind energy connection; the storage side uses the thermal energy time shifting properties of the heat storage device to establish a storage-discharge model to improve RIES. flexibility; finally, construct an optimal regulation model considering the optimal consumption of clean energy. The fake outcomes verify the potential of the suggested model. Compared with the traditional model, an optimal control model considering the optimal consumption of clean energy has more considerable renewable energy consumption capacity and economic benefits.

Fast-switching and fast-throwing device is installed at different open loop points of power supply in the power grid, which can provide backup power source for neighbouring supply zones and has important practical value of the reliability improvment of power syetem. The amplitudes and phase angles of voltages measured at two adjacent busbars are different because both sides of the loop closing positon power system are powered by different power sources. The most obvious impact of loop closing operation is the system disturbance and impact current. Therefore, the constraint condition of loop closing is analyzed, and the technical scheme for the fast-switching device is determined. The practicality of the fast-switching and fast-throwing technology applied at the loop closing position between neighbouring supply zones is verified based on the transient simulation data from regional power system of Baotou, China.

Given that the molten salt pipelines and equipment of Concentrating Solar Power (CSP) plants are equipped with Electric Heat Tracing (EHT) system in addition to the thermal insulation layer, a joint optimization design method for the insulation layer and EHT system is proposed. The design point that minimizes the total investment in insulation and the EHT system is defined as optimum design point A, while the point minimizing the sum of equipment investment and whole life cycle Operation and Maintenance (O and M) cost of the EHT system is defined as optimum design point B. Taking the typical low and high-temperature molten salt pipelines with large/medium/small diameters as examples, the distribution characteristics of optimal design points are analyzed and discussed. Results indicate that the insulation thickness, heat loss, corresponding EHT heating capacity, and amounts of heat tracing cables are very sensitive to the required insulation outer surface temperature. While the optimum design point A and point B are not the same under normal conditions, the selection of the optimum design point should be made based on comprehensive consideration including the specific design requirements and the space constraints for insulation installation. For low-temperature pipelines, the optimum design point A is selected with an insulation outer surface temperature of 8~9°C, and for high-temperature pipelines, the optimum design point A is still recommended with the insulation outer surface temperature of 16~18°C. The corresponding insulation thickness and EHT heating capacity are the optimal design values.

The aim of the fruit and vegetable object detection task is to enable the picking robots to identify objects efficiently and accurately. Although there are extensive studies in improving the accuracy and speed of fruit and vegetable object detection task, the methods still lack of responsiveness to different requirements of picking robots and algorithms for picking different objects. In this paper, based on YOLOv8, Lightweighting Sequential Polarized Self-Attention YOLOv8s (LSPA_YOLOv8s) was proposed to improve the accuracy and speed in grape object detection task. The new approach has three critical designs: (1) Incorporating the ResNet and multiple attention mechanisms to improve the PSA, the DSPSA module is proposed to improve feature extraction; (2) Replace the convolution block in C2f with PConv to get the optimized C2f_PConv module, take advantage of the nature of partial convolution to cut the number of model parameters and memory accesses; (3) Replacement of some Conv blocks in YOLOv8s with Involution to further increase network speed. Experiments show that the method performs well on the grape object detection task.

To enhance the spectral efficiency of conventional frequency hopping systems, Message-Driven Frequency Hopping (MDFH) technology has been proposed in recent years. This technique employs information sequences instead of pseudo-random sequences to select hopping frequencies, significantly improving modulation efficiency. However, this approach results in the receiver being unable to obtain the hopping pattern information, necessitating blind detection, which makes the system more susceptible to interference. This article aims to enhance the anti-interference capability of message-driven frequency hopping systems, integrating Cognitive Radio (CR) technology to propose a Cognitive-based Message-Driven Frequency Hopping Communication System (CR-MDFH). Initially, the system’s transmitter and receiver were improved by implementing a multihop mapping scheme based on frequency band division and an interference detection method based on deducting frequency-hopping bands, building on the foundation of traditional energy detection sensing technology. This enhancement has elevated the system’s frequency hopping pattern performance and reliability. Finally, through simulation, it is demonstrated that the proposed system exhibits excellent interference detection capabilities and superior anti-interference performance in scenarios with partial-band interference.

Differential privacy, as a provable privacy protection framework, has been widely applied in recommendation systems. However, the integration of existing collaborative filtering algorithms with differential privacy typically encounters two issues. Firstly, algorithms often use the same privacy budget for data from different time periods, introducing unnecessary noise and reducing performance. Secondly, when using gradient descent to solve collaborative filtering models, the allocation of privacy budgets is relatively simple, usually evenly distributed across each iteration. This allocation requires a predefined number of iterations. If the number is too small, the algorithm may terminate before reaching the optimal solution; if too large, each budget per iteration may be too small, leading to excessive noise. To address these issues, an adaptive privacy budget recommendation algorithm is proposed. Firstly, a dynamic time window-based privacy budget calculation algorithm is utilized to provide varying levels of privacy protection for rating data from different time periods, thus avoiding unnecessary noise. Then, an adaptive privacy budget gradient descent strategy is employed to solve the collaborative filtering model, which dynamically allocates privacy budgets based on the noise gradient of each round. This algorithm is compared with existing differential privacy-based recommendation algorithms on the Movielens-1M dataset, demonstrating superior predictive accuracy and lower sensitivity to privacy protection.

With the increasing adoption of LIDAR technology in recent years, the beam scanning system, an integral component of LIDAR, has garnered significant attention. The beam scanning system represents a laser scanning technology capable of achieving rapid, highly precise, and stable performance. However, conventional mechanical beam scanning systems, which are widely used, suffer from large size and susceptibility to wear and tear. Similarly, solid-state beam scanning systems often encounter issues such as limited scanning angles and slow scanning speeds. To address the shortcomings of existing solutions, we propose a novel solid-state beam scanning system based on Metasurface-integrated VCSEL as the laser light source in this paper. By redirecting the outgoing laser light from the Metasurface-integrated VCSELs, we generate a largearea spatial point cloud in space, thereby facilitating beam scanning tasks. Additionally, we validate the efficacy of our proposed system through experimentation using an FPGA demo board to simulate the VCSEL driver chip. This experimental verification successfully demonstrates the real-time dynamic beam scanning capabilities of our proposed scheme.

Water rescue operations often face challenges related to timeliness and safety using conventional methods. In this paper, we present a solution to address these challenges by introducing an omnidirectional mobile boat that incorporates advanced technologies. The boat utilizes the YOLO algorithm for detecting individuals in distress within the water and the MPU6050 hardware Digital Motion Processor for accurate posture calculation. Additionally, a path planning algorithm has been implemented to enhance rescue efficiency. Compared to traditional rescue methods, our solution significantly mitigates the risk of personnel injuries during rescue operations while enhancing overall rescue efficiency. Furthermore, our solution effectively resolves the challenge of omnidirectional boat movement through the implementation of a modular three-motor layout scheme. This layout enables prompt rescue of individuals in distress from any location. Moreover, the modular design simplifies the operation and maintenance of the omnidirectional boat, enhancing its usability. Additionally, our solution offers the advantage of low manufacturing costs, making it easily scalable and deployable.

Brushless DC motor is widely used in industrial control because of its good mechanical characteristics, large starting torque and excellent control performance. In this paper, the powerful adaptability of fuzzy PI to nonlinear system is proposed for the speed regulation design of double-closed-loop brushless DC motor. Through the fuzzy system online PI value, the output of the speed regulation system is greatly improved compared with the PI system. After the simulation test, the system has good dynamic performance when using the fuzzy PI control algorithm, which is more superior and effective than the traditional PI control.

As quantum computing continues to advance, the complexity of control systems for superconducting qubits also increases. This paper presents a genetic algorithm (GA) approach for optimizing the control systems in quantum computing. Focusing on minimizing costs, maximizing resource utilization, and reducing device count, the study demonstrates the algorithm's effectiveness in configuring control systems efficiently. Experimental results from five quantum chips, ranging from 64 to 1024 bits, demonstrate significant improvements in cost efficiency and device management after 500 generations of optimization. The findings highlight the algorithm's adaptability to different quantum system scales and its potential in enhancing the scalability and efficiency of quantum computing architectures.

Embedded controller chips play an important role in IoT, vehicle-to-everything, industrial internet, and other fields. However, with the proliferation of the Internet of Things, embedded terminal devices accessing the network face a great risk of vulnerability attacks. Due to the limited resources and high real-time requirement of embedded devices, traditional desktop operating system security measures are difficult to transplant and deploy. This paper proposes a mitigation strategy to maximize vulnerability of embedded systems based on ARM architecture. Based on the theory of moving target defense and the idea of heterogeneous redundant actuator, an equivalent variant is constructed to mitigate the attack of buffer overflow vulnerability. The hardware resources include IROM, IRAM, and the main controller STM32F103VET6 of the Cortex-M3 kernel, thus building heterogeneous redundant actuator and reducing the probability of embedded system intrusion. After implementing the heterogeneous redundant actuator of the embedded system, theoretical verification is also carried out in combination with the ARM-Shellcode experiment. The experimental results show that the proposed method can alleviate the attack using the buffer overflow vulnerability type.

Software testing is a crucial method for ensuring the quality of Android applications. Typically, the failure of test case execution suggests the presence of an error within the application. However, within Android multi-threaded environment, factors such as changes in the execution order of threads can produce inconsistent outcomes for specific operations across multiple test executions, thereby leading to inconsistent test results. In such instances, developers are unable to determine whether errors exist in the program based solely on test results, significantly impacting the reliability of software testing. This phenomenon is referred to as "flaky test." During the development of Android applications, the absence of an appropriate synchronization mechanism to regulate the order of variable access among threads can easily trigger Read- Write races, contributing to flaky tests. In response, this paper introduces a method for detecting flaky tests in concurrency-related Android applications. This method involves capturing the event execution trajectory during test execution and identifying Read-Write races among events along the trajectory. It introduces two key metrics, Event Significance and Inter-Event Influence, to construct an Event Conflict Network. Utilizing this network, the method calculates the priority of the Event Conflict Network and employs priority-based Event Scheduling to identify flaky tests by inducing Read-Write races within as few test executions as possible. The effectiveness of this methodology was assessed using 22 test cases from Android applications. Experimental results demonstrate that our method successfully detected all 22 flaky tests in the dataset, achieving a 33.7% reduction in average detection time compared to existing state-of-the-art methods.

This paper describes the importance of transmission towers in the power system, their key role in long-distance power transmission, and the potential risks posed by the natural environments and severe weather they face. Traditional tower monitoring methods such as GPS, BeiDou and static levels have limitations such as high price and inconvenient installation. To solve this problem, a we3D-based panoramic informatization platform construction scheme is proposed. Firstly, the transmission tower model is analyzed, then the BIM model is built and imported, and the geographic model is created at the same time. The platform interface design is simple, and the data server configuration is well docked. The platform can monitor the tower status in real time, providing strong support for the prevention of tower tilting and wire collapse, ensuring the safe operation of transmission lines, which is of great significance for the "West-East Power Transmission" strategy. This program is affordable, easy to install, and reduces the risk of external damage. Through real-time monitoring, it can reduce the occurrence of accidents and avoid large-scale power outages and social losses. The construction of the panoramic informatization platform provides a strong guarantee for the stable operation of transmission lines.

In response to the problems of pure theory, lack of body feeling, and abstract ambiguity in traditional electric power fire safety, this study combines virtual reality technology to design a new type of electric power fire safety system. The fire scene is modeled and system functions are developed using 3ds Max software and Unity 3D development engine. The process and methods of scene modeling for conference rooms and fire extinguishers are described, Simultaneously, the Gaussian distribution principle is used to determine the smoke concentration at any location during a fire, providing support for highly reproducing real-life fire occurrences. Experiments have shown that when facing a sudden power fire with an area of 100mm2, the disposal results of the research system can reduce property losses by 1.3 million yuan.

Satellite networks are on the brink of becoming pivotal components in future communication systems, thanks to their extensive coverage and multicast capabilities, especially in the context of the emergence of software-defined satellite (SDS) and software-defined satellite networks (SDSN), which enable the provisioning of services within the orbit. However, the changes in satellite network topology inevitably result in a degradation of service quality and significantly improve the expenditure for ensuring the quality of service for users. It's noteworthy that existing research predominantly focuses on deploying service function chains (SFCs) within satellite networks, with relatively less attention devoted to the intricate challenges associated with SFC readjustments caused by the movements of satellites. To address this pressing issue, we introduce the innovative definition of "satellite node stability." This definition effectively utilizes information on the adjacent time-slice network topology to evaluate the importance of nodes within the SFC orchestration process in satellite networks. Furthermore, we propose a methodology for assessing the quality of SFC orchestration schemes within satellite networks. Building upon this approach, we present a tailored SFC orchestration method designed to suit the unique characteristics of satellite networks. Simulation results indicate that this approach can enhance the overall revenue for service providers by approximately 9% without compromising the service deployment success rate. Even more noteworthy is its ability to reduce the total expenditure on service function chain readjustments resulting from satellite movements by around 20%

The present paper primarily presents the development and application of an intelligent maintenance system for winter construction of transmission tower foundations. Leveraging Internet of Things technology and data analysis techniques, this system enables real-time monitoring and intelligent maintenance during the winter construction process, thereby enhancing the quality and safety of transmission tower foundations. Research findings demonstrate that this system exhibits high practicality and reliability, offering a novel solution for winter construction of transmission tower foundations.

This paper explores the enhancement of active suspension systems in vehicles, focusing on improving stability and passenger comfort under varying road conditions. We introduce the Soft Actor-Critic (SAC) algorithm, a sophisticated reinforcement learning method, to advance the efficiency and responsiveness of automobile active suspension systems. SAC is adeptly tailored to navigate the complex dynamics of vehicle suspension, offering an intelligent alternative to traditional control strategies. Our experimental results demonstrate the SAC algorithm's effectiveness, with significant improvements in ride quality and vehicle stability over passive suspension systems. The SAC-controlled active suspension achieves smoother vehicle body displacement and reduced oscillations, underscoring its superior performance. This research marks a notable contribution by implementing SAC in active suspension control, signifying a major step forward in intelligent vehicle control. It sets the foundation for future explorations that could extend SAC's application beyond automotive to other adaptive control systems. Potential future directions include integrating SAC with various vehicle control systems, leveraging advanced neural networks, and real-world testing, further establishing reinforcement learning as a key driver in automotive technology advancement.

With the increasing number of projects undertaken by electronic equipment development enterprises and the increasing complexity of products, the proportion of project material support management in the entire product development cycle will further increase, and information technology is needed to achieve cost reduction and efficiency improvement. The project material supporting management system designed in this article meets the functional requirements of material procurement, material transfer outbound, and material transfer inbound related to supporting work. Through the deep integration of product material list management, material return management, procurement management, and obsolete inventory monitoring and analysis with the project material reservation mechanism, it realizes the standardization of material supporting process, improves work efficiency, and reduces the backlog of inventory. The application results indicate that the system has improved the refinement level of enterprise project material supporting management.

With the development of new power systems, the safe and stable operation of the power grid is of great significance to social and economic development, which puts crucial requirements on the operational skills and business level of power grid employees. Traditional power training cannot meet the increasingly iterative and updated scale of power grid equipment. Targeting to design a more practical and smart training method, this article constructs a cloud training system suitable for power training, virtual reality (VR) technology is utilized to build the 3D models of power equipment. Based on the architecture model of an online cloud classroom system, a cloud training system is constructed, forming an integrated evaluation system that integrates education, learning, and examination. This digital training system could provide desirable guidance for constructing the digital training evaluation system for power systems in the future.

To widely apply deep learning technology, it is necessary to deploy neural network models on different hardware platforms. Multi-core Digital Signal Processors (DSPs), as one of these platforms, exhibit the potential to accelerate deep learning in specific applications and scenarios due to their energy efficiency advantages. CINN, as an AI compiler, is been proposed for deploying DNN models on diverse hardware devices. However, due to differences in hardware architectures, there is a risk of poor utilization of memory and computing resources, leading to increased execution time. In this work, we propose several extensions to achieve optimization specific to DSPs, and allow CINN to automatically generate high-performance tensor programs for our DSPs through a search approach. Experimental result shows that there is a significant improvement in execution performance for the automatically generated tensor operators containing reduction operation compared to manual schedule.

Partial discharge (PD) is one of the common forms of faults in high-voltage equipment, and it is also one of the most important forms of fault manifestation, with different discharge modes for different types of defects. Its identification and localization are of great significance for equipment condition monitoring and fault prevention. This article proposes a local discharge pattern recognition method based on deep learning algorithms. The method uses two deep learning algorithms, Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM), to preprocess the sensor signals, extract effective feature information, and perform pattern recognition through the combination of the two algorithms, achieving good recognition results. This article describes in detail the principles and implementation process of the algorithm, and verifies the effectiveness of the algorithm in simulated experiments.

To ensure the safe and smooth operation of heavy-haul trains and to prevent safety issues such as wheel slip/slide or derailment, a creep ratio controller with soft constraints is designed. The adhesive constraint controller adopts an obstacle Lyapunov function-based approach, considering the dynamics model of heavy-haul trains and taking into account the adhesion mechanism between the wheel and rail. It constructs creep ratio constraint conditions to enhance safety. However, the BLF (Barrier Lyapunov Function) controller suffers from issues such as reduced convergence time with increasing controller gains and increased overshoot. To address these issues, the Finite Time Asymmetric Barrier Lyapunov Function (FT-ABLF) controller is designed using finite time control theory, improving the controller's static accuracy and transient response performance. Simulations are conducted using the Simpack and MATLAB/Simulink platforms, and the simulation results validate the effectiveness of the proposed design approach.

In the use of barcodes for commodity identification, the barcode recognition of non-planar commodities has the problem of a high error rate. Hence, designing commodity intelligent systems based on image correction was carried out. The proposed fisheye-image aberration correction technique was used to eliminate the aberrations in the image and improve the barcode recognition accuracy. At the same time, it was combined with the hardware system to recognize commodities rapidly. Through experiments, it is verified that the system has the function of commodity recognition and can improve the accuracy of non-planar commodity barcode recognition.

Sequencing rules that determine the order of jobs have been well studied and adopted in real-life manufacturing due to their efficiency. However, the rules for assigning jobs among distributed factories are scarcely studied. For dynamic distributed job-shop scheduling problem with new job arrivals, two assigning rules named MOPT and MOW are proposed to minimize the total tardiness. Their performance is evaluated in twenty-four production settings. The combinations of MOPT and MOW with different ratios are also evaluated to find better rules, result shows that the best ratio is interrelated with their independent performance. The research on the performance of proposed rules cooperating with different sequencing rules suggests that assigning decisions have a greater influence than sequencing decisions.

The international Chinese textbook evaluation system is designed based on deep learning algorithm. The evaluation system includes 5 units: user management, online evaluation, data management, evaluation result query, teaching material quality analysis. Based on the evaluation index system of teaching materials, convolutional neural network(CNN) learning expert textbook evaluation samples, the CNN textbook evaluation model is constructed. The system can count the proportion of language content in the syllabus, and the degree of intelligence is higher.