Multi-wavelength method based on global optimization for particle size distribution

The particle size distribution (PSD) of particle medium plays an important role in the field of particle science, so the inversion of PSD is of great significance. To study spherical particle PSD, a multi-wavelength detection model based on a global optimization algorithm called OptQuest nonlinear programming (OQNLP) is established in this paper and an experiment to verify the reliability of the system is designed. The numerical results show that the selection of detection wavelength has great influence on the results of PSD inversion. The relative error of PSD parameters is minimized by choosing the wavelength at the peak of extinction coefficient curve of appropriate particle size. Both simulation and experimental results indicate that the five-wavelength method has the highest testing accuracy. When high accuracy is not required, choosing the four-wavelength method is the most suitable testing method. Furthermore, the universality of the model is also confirmed for the Rosin-Rammer (R-R) function, normal (N-N) function, and lognormal (L-N) function.

Automated summary

This paper establishes a multi-wavelength detection model based on the OptQuest nonlinear programming algorithm to study the particle size distribution (PSD) of spherical particles. An experiment is designed to verify the reliability of the system. The selection of detection wavelength greatly affects the results of PSD inversion, and choosing the wavelength at the peak of the extinction coefficient curve of appropriate particle size minimizes the relative error of PSD parameters. Both simulation and experimental results indicate that the five-wavelength method has the highest testing accuracy.