Application of the biogeography-based optimization algorithms to inverse transient radiation problem

Three biogeography-based optimization (BBO) algorithms, i.e., the basic BBO, the improved biogeography-based optimization algorithm based on local search strategy (ILSBBO), and the hybrid ILSBBO combined with the Simplex Method based on the K-means clustering method (KSM-ILSBBO), were developed to solve the inverse transient radiation problem in nonhomogeneous multilayer participating media. The transmittance and reflectance on the boundaries simulated by the finite volume method were served as input for the inverse analysis. The absorption coefficients, scattering coefficients, and geometric position of the middle layer in a three-layer nonhomogeneous slab were estimated by these three BBO-based algorithms, respectively. The influence of measurement errors on the retrieval results was also investigated. The results demonstrate that the BBO, ILSBBO, and KSM-ILSBBO algorithms all could obtain reasonable results when one, two, or three unknown parameters are retrieved. The KSM-ILSBBO could get more accurate results even when four parameters are estimated simultaneously, and the retrieval accuracy of KSM-ILSBBO is higher than that of BBO and ILSBBO algorithms. The KSM-ILSBBO is proved to be suitable for solving inverse radiation problems.