Optimal design of high efficiency wireless charging system for electric vehicle: AHO–GBDT technique

Abstract

This paper proposed a hybrid technique that optimizes the design of wireless charging systems (WCSs) for electric vehicles (EVs). The proposed technique combines the Archerfish hunting optimizer (AHO) and the gradient boosting decision tree (GBDT) algorithms, hence named the AHO–GBDT technique. The proposed strategy's goal is to enhance the efficiency, minimize cost, and increase coupling coefficient. The AHO algorithm optimizes the design parameters of WCS for EVs. The GBDT dynamically adjusts and optimizes the hyper-parameters, fine-tuning the algorithm for optimal performance. The outcome displays that the proposed strategy can greatly enhances the wireless charging system's efficiency. The proposed strategy is done in MATLAB software and assessed for performance with different existing strategies. The AHO–GBDT algorithm outperforms particle swarm optimizer (PSO), color harmony algorithm (CHA), and owl search algorithm (OSA) in terms of efficiency, which achieves 98.5%, while compared to 97.2%, 96.8%, and 96.4%, respectively. The proposed method shows the highest efficiency, power, and coupling coefficient compared to other existing the PSO, CHA, and OSA methods.