Application of Stochastic Particle Swarm Optimization Algorithm to Determine the Graded Refractive Index Distribution in Participating Media

Lin-Yang Wei,Hong Qi,Ya-Tao Ren,Li-Ming Ruan
DOI: https://doi.org/10.1016/j.infrared.2016.07.024
IF: 2.997
2016-01-01
Infrared Physics & Technology
Abstract:Inverse estimation of the refractive index distribution in one-dimensional participating media with graded refractive index (GRI) is investigated. The forward radiative transfer problem is solved by the Chebyshev collocation spectral method. The stochastic particle swarm optimization (SPSO) algorithm is employed to retrieve three kinds of GRI distribution, i.e. the linear, sinusoidal and quadratic GRI distribution. The retrieval accuracy of GRI distribution with different wall emissivity, optical thickness, absorption coefficients and scattering coefficients are discussed thoroughly. To improve the retrieval accuracy of quadratic GRI distribution, a double-layer model is proposed to supply more measurement information. The influence of measurement errors upon the precision of estimated results is also investigated. Considering the GRI distribution is unknown beforehand in practice, a quadratic function is employed to retrieve the linear GRI by SPSO algorithm. All the results show that the SPSO algorithm is applicable to retrieve different GRI distributions in participating media accurately even with noisy data. (C) 2016 Elsevier B.V. All rights reserved.
What problem does this paper attempt to address?