The characteristic absorption lines of carbon monoxide gas cross with that of other gases in near-infrared and midinfrared bands, the detection results are easily disturbed. Especially in some environments such as coal mines and petrochemicals, it is more important to avoid cross interference, in order to achieve accurate measurement of carbon monoxide. In this paper, a DFB laser beam at 2330nm was selected and scanned for absorption line of carbon monoxide. Although interference of other gas components was avoided, the overlap of carbon monoxide and methane spectra still existed. Firstly, the absorption lines of methane and carbon monoxide were studied. According to the theory of molecular spectroscopy, the FWHM of methane and carbon monoxide absorption lines are different. Once, there is methane gas in the background, the harmonic signals will be different due to different line shape. The second harmonic signals with different modulation coefficients are simulated. So the troughs width is selected as the evaluation function of the harmonic signal characteristics for component judgement. Secondly, after component determination, when mixed gas is present, adjacent methane absorption peaks are also scanned for deduction. The influence of temperature is also calculated and the evaluation function is modified again. Finally, the free calibration measurement of carbon monoxide concentration in the mixture is realized. In summary, for the working conditions of complex background gas such as coal mine and petrochemical. This paper presents a calibration-free method for reliable carbon monoxide concentration detection based on a TDLAS technique.
An algorithm of Rayleigh noise compensation in dual-end configured distributed temperature sensor (DE-DTS) is proposed in this paper. A 2 km long multi-mode fiber is used to calculate the attenuation of the light within Anti-Stokes and Stokes bandwidth, and figure out the isolation of the wavelength division multiplex through both Anti-Stokes channel and Stokes channel. Experiments are taken out to validate the proposed Rayleigh compensation algorithm. As a result, the Rayleigh noise in both Anti-Stokes component and Stokes component is compensated, and the temperature error of the dual-end configured distributed temperature sensor is revised. With the help of the proposed algorithm, a dual-end DTS can reach to the absolute accuracy of 1.09°C (RMSE) between 180°C to 300°C, which significantly monishes the temperature error compared to the sensor without Rayleigh noise compensation.