Multicomponent Trace Gases Detection Based On Photoacoustic Spectroscopy

There are a variety of poisonous and harmful trace gases in the atmosphere, which can cause serious damage to human body. Photoacoustic spectroscopy (PA) technology can detect the concentration of gases by detecting the acoustic amplitude produced by photoacoustic effect. Compared with other gas detection technology, photacoustic spectroscopy has many advantages, such as no gas-consuming、great stability、high sensitivity、good selectivity and so on. Due to the complexity in the environment of the gas detection field, the low concentration of the measured gas and the cross absorption between multicomponent gases, how to improve the stability and accuracy of PA apparatus has been the focus of the searchers.To begin with the molecular Energy Transition Theory, this paper analyzes the heating and forming process of the photoacoustic effect. The distribution of sound pressure in cell is provided and the relationship between PA signal and gas concentration is acquired. On the basis of theory the modules of PA detection system are introduced. The noises in system are divided into two classes and analyzed according to the coherence.The cell is the key part of the whole PA system, and this paper mainly discusses the cell design. PA signal transmission model is established to simulate the generation of PA signal, and the simulation results make a guidance to the design of the geometric structure of cell. Aimed at various system noises, corresponding denoising methods are summarized.The absorption spectra of NO, NO2, CO and H2O are calculated by referring to the HITRAN database and the calculation formula of gas absorption coefficient. Based on the gas absorption lines of characteristic spectrum, the curve that the gas absorption coefficient changes along with pressure and temperature is obtained by simulation.Gas PA system is designed in this paper and the system sensitivity is calculted through the theoretical analysis. Subsequently, RBF model is established to detect the concentration of NO in connection with the cross absorption between multi-component gas. Compared to the true value, the calculation results error is less than2%, the experimental results verify the feasibility of this model.