Arithmetic optimization algorithm based MPPT technique for centralized TEG systems under different temperature gradients

Abstract

Since centralized thermoelectric power generation (TEG) system presents multiple local maximum power points (LMPPs) at different temperature gradients (DTG), thus its optimal power harvesting is difficult to realize via conventional approaches. Therefore, an efficient arithmetic optimization algorithm (AOA) is applied to realize maximum power point tracking (MPPT) of centralized thermoelectric power generation system at different temperature gradients to improve the energy exploitation and utilization. AOA is utilized to efficiently and reliably identify unique global maximum power point (GMPP) in multiple LMPPs, which replicates the distribution mechanism of main arithmetic operators during mathematic calculation to find the best solution from a set of randomly generated candidate solutions. Compared with other well-known algorithms, AOA owns the distinctive superiorities of simple implementation structure, as well as few control parameters need to be tuning. Meanwhile, its own random and adaptive parameters selection principle greatly boost the AOA convergence performance. In addition, AOA shows strong ability to avoid falling into local optima thanks to its proper balancing between global exploration (diversification) and local exploitation (intensification). For further validation of the practicability of AOA based MPPT, two case studies are conducted, e.g., start-up test and step change of temperature. Experimental results show that AOA can achieve the optimal energy harvesting with minimal power fluctuations in comparison with the other three optimization algorithms, e.g., the energy produced by AOA is 34.44%, 6.03%, and 8.17% higher than that of perturb and observe (P&O), particle swarm optimization (PSO) and grey wolf optimization (GWO) respectively under the step change of temperature.