Optimization Of Gas Multi-parameter Reconstruction Algorithm Based On Tunable Dioxide Laser Absorption Spectroscopy

As a common energy conversion method in industrial production,combustion has always been a hot research direction of scholars.In the process of combustion,a variety of mechanisms will be coupled,so that a variety of physical parameter fields with different distribution of pressure,temperature,concentration will be produced in the combustion process.Therefore,if the combustion situation can be accurately monitored online,it is of great significance for scholars to study the combustion related fields.Among various combustion measurement methods,tunable absorption spectroscopy has great potential in engineering application because of its advantages of high selectivity,high real-time performance and high economy.Tunable absorption spectrum technology can be divided into direct absorption technology and wavelength modulation technology.Compared with direct absorption technology with very low dynamic response,wavelength modulation technology itself has a higher signal-to-noise ratio.However,the complex calculation process of harmonic demodulation seriously affects the development of its engineering application,which makes wavelength modulation technology have a higher demand for hardware resources.And the time response is longer.To solve this problem,this paper intends to carry out the research on gas multi-parameter reconstruction algorithm based on tunable absorption spectrum technology,aiming to optimize the current calculation theory of wavelength modulation technology,improve its accuracy and speed,and reduce the computational resources and time response required for wavelength modulation technology to achieve the same accuracy.(1)Based on the gas absorption theory of Bill Lambeau’s Law,this paper analyzes in detail the reconstruction and calculation characteristics of gas parameters such as concentration,temperature and pressure in the process of absorption spectrum measurement.At the same time,the harmonic demodulation process of wavelength modulation technology is deduced in the form of Fourier series,and the disadvantages of this method are analyzed,which makes a paving for the study of the optimization process of wavelength modulation technology.(2)The practical expressions of each harmonic in the modulation process are derived by means of Taylor series.In the process of data post-processing,a special digital filter is designed according to the simulation results,and the numerical simulation results when the filter is applied to the wavelength modulation technology are given.Compared with the traditional digital filter,this filter technology has been greatly improved in anti-noise ability and calculation speed.(3)A new multi-parameter optimization algorithm of harmonic combined iterative method is proposed,and genetic algorithm is used to complete the multi-parameter optimal reconstruction process of the signal.Compared with the slow speed of this kind of intelligent optimization algorithm,a kind of gradient descent optimization algorithm is designed.Under the premise of ensuring accuracy,this method can greatly increase the speed of iterative convergence,and reduce the computing resource requirement of wavelength modulation technology in real machine measurement application.(4)The corresponding experimental verification process is designed.The first half of the experiment mainly includes the construction of optical path and the selection of laser related devices.The second half of the experiment mainly describes some improvement processes when the data post-processing algorithm is combined with FPGA hardware.The process of compiling the core filter algorithm into the hardware algorithm by verilog hardware description language is studied,and the structure of the hardware algorithm is given.Finally,the accuracy improvement of wavelength modulation method compared with direct absorption method is shown in the analysis of experimental results.