Particle Swarm Optimization Algorithm For Radiative Inverse Problem

The radiative heat transfer problems are composed of the direct and the inverse problems. This paper calculates the inverse radiation problems by the Multi-Phase Particle Swarm Optimization (MPPSO) algorithm.The MPPSO algorithm is one of the intelligent optimization methods. After introducing the theory and formula of the basic Particle Swarm Optimization (PSO), the PSO with inertia weight and the MPPSO, three famous benchmark optimization functions Sphere, Rastrigin, Griewank are used to test the convergence speed and the performance of these PSO algorithms. The results reveal MPPSO is better than others.Meanwhile, the present study uses MPPSO to solve the inverse problem of the particles size distribution, which could be usually described by one given mathematic function. An inverse calculated model is built based on Mie light scattering theory, which is used for the R-R, Normal and Lognormal distributions. The MPPSO is proved to be available to retrieve the reasonable particles size distribution, even with the errors.The theoretical infrared models of the artificial satellite have also been set up, and the infrared radiative intensity that the detector catches in a day has been calculated and discussed. The influence of inner heat source and the angle between the solar battery sailboard and sunlight has been illuminated too. Based on the direct problem, inverse problems have been solved to retrieve the inner-heat and the angle between the solar battery sailboard and sunlight by using the MPPSO method. The results show without measurement errors, the present MPPSO is robust and could obtain the satisfied estimation.