An efficient artificial gorilla troops optimizer-based tracker for harvesting maximum power from thermoelectric generation system

The thermoelectric generator (TEG) device opens a new lighted avenue for saving waste heat by transferring it into electrical energy. Due to the unique features of the TEG systems, they have materialized as promising alternatives for green power production. Nonetheless, the TEG generated power is mainly influenced by the distribution of temperature levels added to the value of the load. To enhance the utilization efficiency of the TEG, proposing a reliable maximum power point tracker (MPPT) approach to detect the maximum global power of it is an essential solution. Two remarkable drawbacks are detected in the conventional MPPT, low dynamic response and high oscillations around MPP. Also, operating the TEG system at non-uniform heat distribution generates multi-local peaks in the TEG system characteristic which is difficult to be monitored by traditional trackers. Therefore, this paper proposes an efficient artificial gorilla troops optimizer (GTO) to simulate MPPT with TEG array operated at various temperature distribution conditions. The proposed approach is examined on 9 x 9 and 15 x 15 TEG arrays operated at seven cases including normal, non-uniform row, non-uniform column, long wide, diagonal, internal, and random temperature distributions. For investigating the proposed approach performance versus the state-of-the-art techniques, set of MPPT approaches including comparison to incremental resistance (INR), particle swarm optimizer (PSO), seagull optimization algorithm (SOA), and cuckoo search (CS) have been implemented under the same case studies. The comparisons reveal the superiority of the proposed GTO-MPPT in monitoring the peak global harvested power in time less than 1 s. The SOA is located at the second rank after GTO in the fetched power and the time consumed for achieving the steady state operation. Meanwhile, the INR, PSO, and CS fall in local maximum power points. Moreover, for 9 x 9 array, the minimum error between the simulink model and the proposed GTO-MPPT is 0.01% at internal heat distribution pattern while it is 0.09% for 15 x 15 array at normal operation. The results proved the efficiency of employing the GTO-MPPT compared to the other considered methods.