Extraction of forward small angle particle scattering light pulse signal based on variable forgetting factor RLS algorithm

Abstract

Small angle light scattering measurement is more relevant for determining the size of solid aerosols, but small scattering angle measurement results will be interfered by larger stray light. The key technology is to suppress the background noise caused by the Fraunhofer diffraction of the laser light source and the Rayleigh scattering of atmospheric molecules, so as to improve the resolution of weak scattered light signal of strong light absorption small particle aerosol. An adaptive filtering method of forward small angle aerosol scattering signal is proposed based on recursive least-square (RLS) algorithm. By analyzing the characteristics of small-angle aerosol detection signals, the forgetting factor in traditional RLS is optimized, so that it can not only distinguish aerosol scattered light signals from stray light signals, but also dynamically adjust according to the amplitude change under different particle size and absorbance. In order to verify the filtering effect, small angle scattering light pulse extraction experiments of aerosols with different absorbance and different particle sizes were conducted in a simulated smoke box. Experiments show that the proposed variable forgetting factor RLS algorithm can effectively suppress stray light signals caused by laser light sources and atmospheric molecules. When the aerosol detection signal appears, the algorithm has fast convergence speed and tracking speed, which highlights the aerosol pulse signal well. Compared with the traditional method, the resolution of the processed aerosol scattering pulse signal increases dramatically and has a great advantage in the extraction of weak scattering pulse signal.