Applying student psychology-based optimization algorithm to optimize the performance of a thermoelectric generator

Abstract

ABSTRACT In this study, two algorithms, the student psychology-based optimization (SPBO) and the mutation/particle swarm optimization (MPSO/PSO) were invoked to optimize the output power of a TEG module. The results indicate that compared to the MPSO/PSO, the SPBO has a better global searching ability under a low population size, while involving fewer adjustable parameters. The original SPBO algorithm was modified by introducing a mutation (M-SPBO). Then, the global searching ability, convergency, and repeatability of the M-SPBO were evaluated. The mutation subprogram improved the global searching ability of the original SPBO. With a high population size, the repeatability was also improved.