An improved metaheuristic-based MPPT for centralized thermoelectric generation systems under dynamic temperature conditions

This paper proposes a multi-peak maximum power point tracking (MPPT) method based on the Global Flying Squirrel Search-Particle Swarm Optimization (GFSS-PSO) for centralized thermoelectric generator (TEG) systems operating under uneven temperature distribution conditions. Conventionally, metaheuristic-based MPPT methods mainly focused on indicators such as tracking speed, oscillation amplitude, and system efficiency. However, the real-time global search ability of conventional metaheuristic-based MPPT methods designed for photovoltaic systems may not be suitable for the gradual temperature change in the thermoelectric scene. A strong global search capability also can add to the computational burden and increase the power loss in the search process. To solve these problems, the GFSS-PSO algorithm introduces improved position updating method and multi-threshold restart mechanisms to reduce energy loss and improve the dynamic performance under temperature change. The proposed method has been compared with the perturb and observe method and several state-of-the-art metaheuristic-based MPPT algorithms. Simulation results confirm that GFSS-PSO demonstrates exceptional performance and generates higher energy levels compared to perturb and observe, grey wolf optimizer, and flying squirrel search optimization methods during the search phase under dynamic temperature conditions. The improvements achieved by GFSS-PSO are remarkable, with energy levels increasing by 118.3%, 105%, and 102.2% respectively. Finally, experiments are conducted to verify the effectiveness of the proposed algorithm in a real-time digital system.