A prototype wide-field-of-view (WFOV) star tracker camera has been fabricated and tested for use in spacecraft navigation. The most unique feature of this device is its 28 degree(s) X 44 degree(s) FOV, which views a large enough sector of the sky to ensure the existence of at least 5 stars of mv equals 4.5 or brighter in all viewing directions. The WFOV requirement and the need to maximize both collection aperture (F/1.28) and spectral input band (0.4 to 1.1 micrometers ) to meet the light gathering needs for the dimmest star have dictated the use of a novel concentric optical design, which employs a fiber optic faceplate field flattener. The main advantage of the WFOV configuration is the smaller star map required for position processing, which results in less processing power and faster matching. Additionally, a size and mass benefit is seen with a large FOV/smaller effective focal length (efl) sensor. Prototype hardware versions have included both image intensified and un-intensified CCD cameras. Integration times of

An instrument has been designed to demonstrate 5 arc second autonomous, all-stellar attitude determination on a NASA Spartan spacecraft. The instrument includes a CCD star camera that provides centroid measurements of up to 5 simultaneous star images with 10 frames of data per s. The performance of the camera and processing techniques have been studied in an extensive test program including star measurements taken under realistic conditions at the Jet Propulsion Laboratory Table Mountain Observatory. Results of these tests are presented and compared to laboratory and simulation data. At slue rates less than 0.1 deg/s the camera accuracy was found to vary from 5 to 15 arc seconds per image, depending on star magnitude. Q

This paper presents a general expression for the transmission function of the discrete frequency-versus-radius reticle and compares such a reticle to the more common continuous reticle. A discrete form of the frequency-versus-radius reticle has an integer number of chopping cycles on a single radius. The discrete form limits the resolution of the reticle in the radial direction, but this limit is not severe for small target images. However, since no phase reversal occurs, electronic processing is simplified. Also, a tracker is presented that utilizes two discrete frequency-versus-radius reticles that can track large objects and eliminates target cancellation.

Conventional thermal images which use single-element detectors usually incorporate mechanical scanning systems which move the detector's instantaneous field of view in a regular pattern to cover the imager field of view. Alternative techniques are possible which utilize discrete reticle masks to multiplex several resolution elements onto the same detector in order to increase the total measurement time per resolution element. Benefits and limitations of these techniques are discussed.

An appealing approach to implementing spinning reticle trackers in the pupil plane is presented. The advent of research in spatial light modulators and addressable mirror arrays may allow achievement of pupil plane processing in the next few years. Reticles were generated with the general FM reticle equation, and FFTs of the reticles were obtained. Magnitude and phase plots of these reticles are shown. The magnitude plots are easily understood; however, the phase plots will take more study before implementation is tractable.

Frequency modulated (FM) reticles can be used to create imaging systems that operate simultaneously from the ultraviolet to the far-infrared regions of the optical spectrum. The reticle modulates different spatial locations at different temporal frequencies, and these locations are optically condensed onto a single detector. Modulation takes place prior to splitting into spectral bands, resulting in excellent pixel-to-pixel registration of the different spectral images. Sensitivity is better than the spinning filter wheel technique sometimes used to achieve good registration. Modulation also allows multiplexing many pixels onto fewer detectors by utilizing the bandwidth of the detectors more effectively. Thus, a single detector may perform as a linear array and a linear array may perform as a staring array.

The problem of time-optical guidance design for the low-altitude, subsonic, vertical plane approach of a sensing air vehicle to a fixed final point is studied using a nonlinear, constrained, optimal control problem formulation. Multiple, competing optimization criteria are included as state equality constraints for design trade-off analysis. A sensor line-of-sight (LOS) limit is adjoined to the system Hamiltonian as a control/state variable inequality constraint to provide a sensor performance influence on the guidance design. Numerical results are provided that illustrate the optimal guidance, the sensor LOS environment along the optimal maneuver, and the influence of the sensor limit on the guidance law design.

This paper describes the overall configuration and performance of a comet approach and landing system (CALS), a space-borne sensor package for navigation toward a low-gravity celestial body in an interplanetary environment. The sensor system is aimed at satisfying the requirements of the Rosetta/CNSR (comet nucleus sample return) mission, whose objective is to land on a cometary surface and to retrieve samples that will be analyzed on the ground after Earth re-entry. Several constraints at the mission and spacecraft level make the configuration of a suitable sensor package a quite complex problem. The baseline system includes the following sensors, all mounted on a high-precision gimballed platform: (1) high-resolution camera, for detection of the comet at large distance and mapping at medium/short distance for ground-assisted landing site selection; (2) wide-angle camera with data processing equipment (star and target tracker), able to track simultaneously the irregular comet image and the surrounding stars for autonomous navigation; (3) laser topographic mapper for autonomous topography-assisted navigation in the final descent phase; (4) multitask radar altimeter for the on-board measurement of range, attitude, 3-axis velocity and surface roughness, with a microwave sounder section for the determination of subsurface structure and composition.

The Indian Remote Sensing Satellite IRS-1A, launched in March 1988, is a three-axis stabilized, polar sun synchronous satellite orbiting at an altitude of 904 km. Two types of earth sensors are used for pointing and control of the satellite. One is a pair of conical scanning sensors using a rotating germanium wedge prism. The other is a static horizon sensor operating on the principle of radiation balancing. The latter sensor used novel normalization technique for removing the effects due to radiation gradients, which is one of the main problems of this type of sensor. The in-flight performance of the sensor is quite satisfactory with very low noise behavior. However, there were certain problems noticed in the acquisition mode of operation of the sensor which were traced to the heating of the IR filter due to direct sun viewing by the sensor near the poles. Based upon this experience, the configuration for IRS-1B to be launched in 1991 was modified. This paper briefly describes the configuration, the flight performance, and the modification carried out in the future models of the sensor.

Indian remote sensing satellites (IRS) carry a set of star sensors using CCDs for precise determination of satellite attitude. The star sensor data is used for precision annotation of imagery received from the satellite. These satellites are three-axis stabilized in polar sun- synchronized orbits. Two types of star sensors have been developed. The first type is based on a linear CCD and the second one is based on an area CCD. They use thermo-electric coolers to cool the detectors. A comparison of both configurations is given.

The second-generation Indian communication satellite INSAT-II is scheduled for launch in 1991-92. The primary attitude sensor in this satellite is a two-axis scanning infrared earth sensor. The satellite carries three earth sensors, one being used exclusively in transfer orbit. The earth sensor contains an oscillatory scan mechanism which has been qualified for 10 years life in orbit. A mirror is made to oscillate on a taut band at low frequency by a sensor and torquer system. Two sensors nominally scan in the east-west direction while the third sensor scans in the north-south direction. The sensor has features to work in presence of sun/moon interference. The paper describes the elaborate qualification tests and characterization carried out in thermo-vacuum conditions. The sensor has a thermal design which was verified by a thermo-vacuum solar simulation test.

This paper describes a novel method of hierarchical asynchronous distributed filtering called the Net Information Approach (NIA). The NIA is a Kalman-filter-based estimation scheme for spatially distributed sensors which must retain their local optimality yet require a nearly optimal global estimate. The key idea of the NIA is that each local sensor-dedicated filter tells the global filter 'what I've learned since the last local-to-global transmission,' whereas in other estimation architectures the local-to-global transmission consists of 'what I think now.' An algorithm based on this idea has been demonstrated on a small-scale target-tracking problem with many encouraging results. Feasibility of this approach was demonstrated by comparing NIA performance to an optimal centralized Kalman filter (lower bound) via Monte Carlo simulations.

An improved ring laser gyro-based inertial navigation unit (RLG INU) has recently been delivered. The new RLG INU is an all-attitude strapdown inertial navigator. Designed to provide superior attitude performance, this new RLG INU also provides navigation performance well within the Air Force Standard Navigator requirements for a strapdown system. The new design expands upon F-15 and Air Force Standard Navigator INU technology. A number of techniques were employed to achieve superior performance. The sensor assembly was redesigned to increase stiffness and reduce coning characteristics. Filtering algorithms were redesigned to reduce noise under 2.5 arc seconds. Coning algorithms were redesigned to reduce compensation errors. Attitude computation rates were increased to 300 Hz to reduce data delays. A faster microprocessor was incorporated in order to implement the improved coning compensation, filtering techniques, and higher attitude computation rates. New calibration techniques were developed because of the navigation and attitude precision of this navigator. Extensive ground testing has shown the navigator's performance to be well within requirements. Performance does not appear to degrade at either temperature extreme, nor in a rigorous vibration environment. Limited flight testing has demonstrated the RLG INU's navigation performance. More extensive flight test results should be available later this year. This paper provides the design rationale for this unit, its ground/flight performance, and plans for further testing.

An interferometric fiber optic gyro (IFOG) measures rotation by comparing the propagation time of optical signals transmitted in opposite directions around a coil of optical fiber. An important constraint on accuracy is the thermally induced nonreciprocity which can occur when there is a localized time-dependent thermal expansion in a section of the fiber. In order to prevent measurement error due to thermally induced nonreciprocity, the commonly used IFOG coil geometry, called quadrupole winding, features layers after the innermost wound in pairs, each layer pair beginning with a fiber element on the opposite side of the innermost layer from that of the preceding layer pair. This alternating layer pair arrangement is needed so that any thermal gradient present in the coil structure will be symmetrical about the center of the fiber loop, with the important desirable result that time of flight variations for the two opposite transmission directions due to time varying thermal expansion cancel one another. Although the principle is sound, such windings are difficult to realize in practice because the alternating layer pair geometry tends to allow winding flaws which degrade precision and make reliable production of the winding difficult. OPTELECOM has devised methods for making flawless level quadrupole windings between flanges which facilitate maintaining the precision required for IFOG coils. Under DARPA/U.S. Army Contract DAAH01-90-C-0520, the authors investigated application of these methods to low manufacturing cost IFOG coil winding implementation. A summary of IFOG coil winding constraints, discussion of winding techniques which lead to low-cost winding while satisfying constraints, and results of initial cost reduction work are presented.

ASTRO-type systems for autonomous attitude determination in space are introduced. They recognize stars which are registered in the field of view (FOV) of a star sensor by using an on- board star catalogue for subsequent attitude determination in the 2nd equatorial coordinate system. The star sensor utilizes a Peltier-cooled CCD array in the focal plane of a 1.4/100 mm lens which realized a FOV of 5.3 degree(s) X 8 degree(s). A first microprocessor system reduces the CCD readout data to the images of the registered stars and subsequently computes the exact star positions. A second microprocessor system realizes the attitude determination and is able to handle up to three star sensors simultaneously. Utilizing more than one star sensor in a nearly orthogonal configuration allows a 3-axes attitude determination with high accuracy. The first ASTRO type system--the ASTRO 1--was designed and manufactured from 1984 to 1987 and launched in November 1989 on Module D to the Soviet MIR space station. Test results demonstrate that the in-orbit performance is better than specified. The star sensor registered stars up to 8.0 mv and the measurement accuracy for attitude determination is characterized by 1 - 2 arcsec. Currently, the advanced ASTRO 1M system is under development. It serves as a subsystem of the attitude measurement and control system (AMCS) of the Soviet SPEKTR-RG satellite, where it provides both initial attitude information and an absolute attitude reference to update the gyros.

An experimental study is conducted to investigate temperature effects on the polarization characteristics of three polarization-maintaining (PM) fibers designed to operate in the 1.3 micrometers wavelength region. A conventional single-mode (SM) fiber is used for a comparison. The polarization stability of the test fibers is evaluated as a function of temperature (between -60 degree(s)C and 100 degree(s)C). Significant changes are observed in the extinction ratio at -60 degree(s)C for one of the PM fibers containing a UV-cured buffer coat. It is believed that lateral compressive forces due to the buffer coat shrinking onto a nonuniform glass fiber at low temperature cause microbending and mode coupling. While the test fibers containing the UV-cured buffer coats are in general more sensitive to temperature changes, some fiber designs seem to perform better according to the experimental results.

A feasibility study of an active 3-D terrain mapping system for helicopter landings is presented. The system acquires the 3-D shape of the terrain beneath the helicopter by scanning the terrain with a laser beam and imaging the illuminated spot by two position-sensing devices. The measurement accuracy of the system is assessed, and a Kalman filter for accuracy improvement is devised. The acquired terrain shape is displayed continuously to the pilot on a perspective display, showing both the terrain shape and helicopter spatial position.

Electrooptical sensors provide a covert way of computing range during helicopter flight. The optical flow-based computation of range provides range information only in certain distinguishable parts of the image. The regions where range information is available can be increased by performing texture analysis and object segmentation in the image. This paper reviews some of the literature on texture segmentation methods with a view towards applying them to images containing both man-made and natural objects at varying ranges. Two algorithmic approaches are given and their application to a real image is demonstrated. Results indicate that it will be necessary to combine several different texture measures and methods in a hierarchical way in order to achieve an object segmentation which is useful in enhancing range information.

Image-based ranging has emerged as a critical issue in the low altitude operation of flight vehicles such as rotorcraft and planetary landers. These flight regimes require ranging systems for recovering the geometry of the terrain and obstacles for use with guidance algorithms. The development of a ranging equation combining image irradiance together with various order spatial partial derivatives and the vehicle motion parameters is discussed. The ranging equation is in the form of a polynomial in scene depth. Two-dimensional linear filters are then used to compute the coefficients of this polynomial to result in a fast image-based ranging algorithm. Performance of the algorithm is demonstrated using laboratory images.

This paper addresses the real-time estimation of rigid-body angular motion in 3-D space. The method obtains covariances of the 'true' 3-D angular velocity, the angular velocity measurement and the measurement noise, from the time averages of the outputs of an array of nine linear accelerometers and the outputs of three orthogonal gyroscopes. These statistics are used by the estimator to calculate the 3-D angular velocity of the rigid-body. The multisensor technique performance is evaluated through a computer simulation. The results indicate the new method leads to more accurate angular velocity values than are obtained conventionally.

The coherent launch-site atmospheric wind sounder (CLAWS) is a lidar atmospheric wind sensor designed to measure the winds aloft at space launch facilities to an altitude of 20 km. Candidate lidar systems analyzed for use in CLAWS include Nd:YAG, Ho:YAG, and CO2. Detailed simulations were carried out by Coherent Technologies, Inc. The results of our development studies include: (1) definition of lidar sensor requirements, (2) definition of a system to meet those requirements, and (3) a concept evaluation with recommendations for the most feasible and cost-effective lidar system for use as an input to a guidance and control system for a missile or spacecraft launch. A field test program will begin in August 1991, in which the ability of CLAWS to meet NASA goals for increased safety and launch/mission flexibility at Kennedy Space Center (KSC) will be evaluated with regard to maximum detection range, refractive turbulence, and aerosol backscattering efficiency at the three lidar wavelengths. It is found that the shorter wavelength solid-state lasers will afford better performance (longer detection range), are more energy efficient, and are more compact for operation in the humid, postvolcanic aerosol environment found at KSC. Finally, the Ho:YAG (2.1 micrometers ) lidar gives the best performance at an eyesafe wavelength and would be applicable for detecting winds aloft during descent as well as during ascent.

A four-center NASA team has undertaken to develop and demonstrate mature technologies applicable to autonomous guidance, navigation, and control (GNC) systems for application to the National Space Transportation System in full cognizance of its operational, safety, and performance requirements, as well as its cost constraints. Attention is to be given to GNC launch/landing weather assessment, ascent guidance, ascent load relief, and system failure during ascent. Preliminary results indicate that a ground-computed atmospheric steering profile can achieve near-optimum performance as well as high cost effectiveness.

The association of optronic and radar sensors significantly improves comt'at aircraft potential. After having studied the fusion and estimation algorithms required by this association,Thomson-CSF has set-up a ground experiment in order to evaluate the performance of a combined radar-optronic system in short and medium range aerial combat.A system of this type can be used in association with fire-control systems (gun or missile). The experiment combines a RDM radar, an ATLIS TV tracking pod and will be completed later with an ASPIC IRST. The potential interest of the combined use of radar and optronic is presented with the experimental conditions and the main results obtained