Optimized Placement of EV Charging Stations With Integrated Photovoltaic and Battery Storage Systems Using Spatial Bayesian Neural Network

Abstract

Electric vehicles (EVs) are gaining popularity worldwide as a means to reduce carbon emissions (CE) and dependence on fossil fuels. However, the limited availability of public charging stations remains a major barrier to adoption. This study proposes a hybrid strategy for solar distributed generation (DG) and EV charging stations (EVCS) installation in distribution systems (DSs). The novelty of this paper lies in the innovation of the spatial Bayesian neural network (SBNN) and greater cane rat algorithm (GCRA). It is therefore called GCRA–SBNN. The main goals of this proposed approach are to lower power costs and losses while simultaneously enhancing the voltage profile of the system. The proposed GCRA technique is employed to optimize the rate of EVCS units in the DS, and the SBNN method is used to predict optimal locations for EVCS. Then, the proposed strategy is implemented in the MATLAB working platform, and the existing procedure is used to calculate the results. The proposed strategy outperforms all current approaches, including the gray wolf optimizer (GWO), genetic algorithm (GA), and particle swarm optimization (PSO). The existing method shows power losses of 69.428, 71.3, and 72.406 kW, while the proposed method results in a reduced power loss of 68.4272 kW. For the voltage profile, the present technique generates values of 0.932, 0.928, and 0.918 p.u, whereas the proposed strategy achieves a higher voltage profile of 0.9378 p.u. This improvement indicates that the proposed technique not only minimizes power loss but also improves voltage stability. The proposed technique provides better performance than the current method. Overall, these advances show that the proposed strategy provides a more efficient and stable solution for power DSs.