Multi-objective optimization of thermoelectric conversion systems by RSM and NSGA-II

A thermoelectric module directly converts heat into electricity. However, optimizing the thermoelectric conversion system (TECS) to maximize the power output is necessary. In addition to improvements in the thermoelectric materials and modules, improvements in the spatial layout of the TECS are equally important to achieve its high performance. This study is intended to investigate the effect of different spatial parameters in TECS. A computer model was designed to simulate the performance of the TECS. This paper investigates the impact of three key factors on the performance of a thermoelectric system: the thermocouple duty cycle, the spacing of thermoelectric modules, and the number of modules. These factors are combined with a Response Surface Model (RSM) and a Non-dominated Sorting Genetic Algorithm II (NSGA-II) to develop a fast method for optimizing the system's performance. This method saves significant time for simulation calculations without compromising computational accuracy.