Bio-inspired methods for fast and robust arrangement of thermoelectric modulus

Abstract

This paper aims to evaluate the ability of some well-known bio-inspired metaheuristics for optimal arrangement of thermoelectric cells mounted in a thermal component. In real life applications, proper arrangement of thermoelectric modules plays a pivotal role by maximising the generated electricity. However, some defects such as the increase in total maintenance cost is often associated with the use of thermoelectric cells. Hence, it is mandatory to contrive a policy which guarantees the maximum electricity generation while keeps the maintenance cost in lowest level. Here, authors use both adaptive neuro-fuzzy inference system (ANFIS) and experimental data to model the power generation and maintenance cost of thermoelectric cells. At the next step, they engage some famous bio-inspired metaheuristic algorithms, i.e., bee algorithm (BA), particle swarm optimisation (PSO) and the great salmon run (TGSR) to arrange the thermoelectric cells in a cost effective manner. The gained results indicate that the proposed algorithms are highly capable to find an efficient arrangement for thermoelectric cells within a rational duration. Besides, through independent runs, it is observed that metaheuristics show acceptable robustness for the current case study.