Optimizing power quality and placement of EV charging stations in a DC grid with PV-BESS using hybrid DOA-CHGNN approach

Abstract

Electric vehicle battery chargers have power electronic transformers, which causes significant distortions in electrical energy obtained from distribution system and numerous issues with power quality. This paper presents a hybrid method for optimizing energy quality and placement of Electric VehicleCharging Stations (EVCS) with Photovoltaic with Battery Energy Storage System (PV-BESS) in DC grids. The proposed method combines Dollmaker Optimization Algorithm (DOA) and Contrastive Hyper graph Neural network (CHGNN), referred as DOA-CHGNN technique. The primary goal of proposed strategy is to reduce voltage drop, Total Harmonic Distortion (THD) and increase system's efficiency. The DOA method is used to enhance assignment of EVCS in delivery system. The CHGNN method is utilized to predict the EV load. The MATLAB environment is used to assess and compare the proposed method with other existing techniques. The proposed approach determines betterfindings compared to existing methods like Jellyfish Search Algorithm (JSA), Hybridized Whale Particle Swarm Optimization (HWPSO) and Deep Neural Network (DNN). The proposed methods achieves a THD of 0.9 %, Total cost of 5,520,000$, the execution time of 0.41 s and an efficiency of 98 %.The proposed DOA-CHGNN method outperforms existing techniques, achieving improved THD, higher efficiency, and lower costs in optimizing EVCS placement with PV-BESS in DC grids.