Multi-component Gas Analysis Based On Photoacoustic Spectroscopy

Gas detection is receiving more and more attention in industrial and agricultural production and in people’s daily lives,and the requirements for gas detection technology are also increasing.In today’s increasingly demanding world,most detection technologies are no longer able to meet the requirements.Photoacoustic spectroscopy gas detection technology detects gas concentration by detecting the amplitude of acoustic waves generated by photoacoustic effect,which has the characteristics of no gas consumption,good stability,high sensitivity and excellent selectivity compared with other gas detection methods.However,in different fields,the detection of single component gases is no longer suitable for most of the detection scenarios,and more often multi-component gas detection is required.However,when multi-component gas detection is performed,cross-absorption interference between different types of gases can lead to inaccurate results due to complex environmental factors and low concentrations of the gases to be measured.In order to reduce the cross-interference phenomenon between different gases and improve the accuracy of multi-component gas detection,different methods are used for modeling correction to improve the final detection results.Firstly,this paper studies the detection principle of photoacoustic spectroscopy technique based on gas absorption spectroscopy theory and photoacoustic effect,and lays the foundation for the realization of photoacoustic spectroscopy gas detection.Second,the main influencing factors in the process of photoacoustic spectroscopy gas detection are analyzed.The effects of temperature and pressure on the parameters and performance of the detection system are studied,and the main mathematical relationships corresponding to the temperature and pressure of the detection gas are established to realize the corrections under different environmental conditions.The noise that interferes with the appearance of photoacoustic signals in the detection system is analyzed and measures are taken to correct it.Finally,in the use of photoacoustic spectroscopy technology for multi-component gas detection,for the multi-component gas exists cross-absorption eventually affect the detection results of the phenomenon,respectively,using least squares,linear regression analysis,support vector machine method for analysis,and for different methods exist in the shortcomings of the optimization,respectively,after test verification,can improve the accuracy of the detection results to a certain extent.In order to better solve the problem of cross-talk in multi-component gas detection,the non-dominated sorting genetic algorithm with elite strategy(NSGA2)and discrete binary particle swarm algorithm(BPSO)in neural network algorithm are also selected to build models and perform multi-component gas concentration calculation respectively,compare their relative errors and verify the feasibility.On this basis,NSGA2 and BPSO are combined and further optimized in depth to address the defects in the neural network,and finally a new NSGA2-BPSO-RBF algorithm is proposed,which optimizes the model weights and thresholds and achieves the improvement of model performance and the relative error is reduced from 6% to 1%.It is verified through experiments that the NSGA2-BPSO-RBF algorithm can effectively eliminate the cross-influence between gases and improve the accuracy of detection results.