Study on data processing method of Raman scattering in complex combustion field

Spontaneous Raman Scattering (SRS) is an advanced technique for flow field diagnosis. It is an irreplaceable method in the measurement of combustion flow field because it can obtain the concentration information of various components quantitatively and single time without the influence of laser energy. Since the scattering cross-section of raman molecule is very small, the signal is weak and easy to be disturbed by noise. When the diagnosis object is a complex turbulent combustion field in an aero-kerosene-fueled engine, the interference becomes very serious, resulting in a serious decrease in the signal-to-noise ratio (SNR) of the raman scattering signal, and even the weak raman scattering signal is completely submerged by all kinds of strong interference, which makes it difficult to apply the raman scattering technique to the diagnosis of such complex combustion field.To suppress the noise interference and extract the signal accurately is a difficult problem to be solved in the application of Raman scattering technology. The commonly used filtering method has a good effect on the general combustion field diagnosis, but it is difficult to solve the complex combustion field of the engine. In this paper, multiframe average method and Savitaky-Golay filtering method are used to suppress noise, W-test is used to solve the optimal parameters in the weighted average algorithm (SG) of moving window, and the mixed distribution model of Newton descent method is used to extract the optimal parameters in the probability density. Raman scattering signals with high intensity and high signal-to-noise ratio are obtained in the complex combustion field of the engine. The high precision parameters such as the concentration of the main components in the complex combustion field of the engine are obtained by raman scattering signal inversion, which provides technical and data support for the design of the combustion chamber of the engine and the optimization of the combustion experimental parameters.