Mathematical Modeling and Parameter Optimization of Fine Particle Sensors Based on Laser Light Scattering

Fine particles have promoted air quality monitoring need of humankind after their harmful effects on human health are pathologically verified. To monitor fine particle concentrations at a low cost, laser light scattering technique is favored by many researchers. While several cheap and interesting sensors are developed, little attention is paid to the measurement model, which in turn bound the sensor precision. In order to alleviate this situation, a simplified model is proposed, which can describe the nonlinear relationship between particle volume and single particle angular light flux by considering the effect of Rayleigh scattering and Fraunhofer diffraction simultaneously. The model can be employed in the signal processing of commercial low-cost particle sensors based on laser light scattering. A data-driven approach is adopted to find the optimum parameters of this model by considering different kinds of fine particles. Based on Trust-Region and Levenberg-Marquardt methods combining with robust methods, the optimum parameters of this model are obtained. To verify the effectiveness of the proposed model, comparative experiments are conducted on: a particle sensor with and without parameter optimization, a DustTrak II 8530 monitor and a reference TEOM PM2.5 monitor. The results indicated that the optimal model improves the measurement accuracy of the fine particle mass concentration by 10%.