Font Size: a A A

Design And Research Of TDLAS-based Process Visualization For Methane Sensing Applications

Posted on:2023-02-23Degree:MasterType:Thesis
Country:ChinaCandidate:X F WangFull Text:PDF
GTID:2531307022998079Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Optical methane gas sensor based on tunable diode laser absorption spectroscopy(TDLAS)has attracted increasing attention because of low cross interference,high sensitivity and rapid response.Based on Beer-Lambert law,a whole-process visual TDLAS gas sensing system has been constructed,and system optimization,real-time calibration and data analysis have been investigated.The main research is conducted as follows:(1)The concept and development of optical gas sensing technology are briefly introduced,and the basic principles of TDLAS and wavelength modulation spectroscopy(WMS)are introduced.The Beer-Lambert law are analyzed,and the relationship between harmonics and gas concentration is discussed for subsequent simulations and experiments.(2)Simulink software is employed to establish and simulate the complete TDLAS process.The gas absorption line pattern and line width are both added into the simulation model to analyze their effect on harmonic waveform and laser source selection.The gas temperature and pressure influence on the absorption line and harmonics are investigated,to optimize the laser modulation in WMS and enhance subsequent experiments.(3)A TDLAS gas sensing system was experimentally built up,and the visualization interface was established by Lab VIEW.Analog circuits are employed to drive the laser source and stabilize the laser operating temperature within ±0.02 °C.The dual-channel digital phase-locked algorithm is designed to lock the phase difference between the reference signal and the measured signal.The interface is established to monitor the working status of each essential module simultaneously and improve data analysis and subsequent experiments.(4)The VMD-SG filtering algorithm is introduced into data analysis to improve the SNR ratio,and the system stability is improved by 27% and 12.5% respectively,compared to the original signal and wavelet domain denoising algorithm.Furthermore,the RBF neural network is introduced and leads to improved fitting accuracy between gas concentration and harmonics,compared to linear fitting and least squares fitting,with absolute error of 50 ppm.
Keywords/Search Tags:Tunable laser absorption spectroscopy, Wavelength modulation spectroscopy, Visualization, VMD-SG algorithm, RBF fitting
PDF Full Text Request
Related items