Research On Image Reconstruction Algorithm Of Gas Concentration Field Detection Technology Based On TDLAS Technology

It is a key measure to realize the optimal combustion control,improve combustion efficiency and reduce pollutants to effectively detect the combustion state of the boiler chamber in power plant on line and make combustion diagnosis.Tunable laser absorption spectroscopy(TDLAS)is a non-contact gas detection technology that utilizes spectral absorption characteristics.Its advantages include high sensitivity to small changes,strong real-time response and effective anti-interference.It has good compatibility to measure gas parameters in real time under bad operating conditions of combustion space.Between internal combustion chamber of a stove or furnace space is huge,the combustion state complex and combustion product concentration change strong sexual characteristics,to achieve low error of two-dimensional gas concentration field reconstruction results is a necessary condition for enough projection Angle in order to obtain adequate projection data,gas concentration detecting system existing in many application way of laser and laser receiver position,the incomplete projection data leads to lower precision of image reconstruction,bottlenecks are widely used in industrial field.In this paper,TDLAS technology is mainly used to measure the CO2 concentration of combustion products in the boiler chamber and conduct research and experiment on the image reconstruction algorithm of two-dimensional gas concentration field.(1)The basic principle of tunable laser absorption spectroscopy and the important parameters for detection are described,and the characteristics of various measurement techniques are summarized.(2)This paper studies the basic principle of image reconstruction method,compares the advantages and disadvantages of various image reconstruction methods that can be used for gas concentration,and comprehensively analyzes several factors that influence the reconstruction results of gas concentration distribution.The performance difference of several typical iterative algorithms in image reconstruction precision is studied by combining error analysis with simulation of field model.(3)Taking carbon dioxide gas as the main detection target,the laser signal with the central wavelength of 2004.5484nm was selected as the laser light source and the experimental requirements were combined to build a large-scale experimental platform for simulating power plant boilers with the same proportion reduction.In addition,mechanical transmission device is adopted to automatically scan and detect laser signals,greatly increasing the number of projected optical paths.To a certain extent,the problem of limited optical paths caused by fixed layout for laser detection is solved,which effectively saves detection cost and provides data support for system image reconstruction to improve reconstruction accuracy.(4)An optimized image reconstruction algorithm combining iterative algorithm and gaussian process regression(GPR)is proposed,which can effectively reduce errors and improve reconstruction accuracy.The optimization algorithm constructs the objective function based on robust estimation,and the algebraic iterative algorithm is developed to solve the objective function effectively.GPR method was used to predict the gas concentration distribution of the fine mesh,and more accurate information of the gas concentration distribution could be obtained.The simulation results of the model and the experimental results of the measured data verify the effectiveness of the optimization algorithm in improving the reconstruction accuracy.