Season Aware Solar Photovoltaic-Based Efficient EV Charging System Using A Novel JFL-BiLSTM And RF-SHO

Abstract The surge in Electric Vehicle (EV) adoption has intensified energy demand, prompting the integration of Solar Photovoltaic (S-PV) systems for optimization. To address the challenge of fluctuating load demand with seasonal changes, a novel approach is proposed. This system combines Laguerre Polynomial-based Ramp Rate Method (LP-RRM) to smooth DC power from S-PV, Generalized Space Vector Modulation-based Switching Regulators (GSVM-SR) to tackle grid interruptions, and Rosenbrock Function-based Sea-Horse Optimization (RF-SHO) for optimal Charging System (CS) selection. Pre-processing the dataset identifies seasons and extracts features, which are fed into Joint Fusion Layer – Bidirectional Long Short Term Memory (JFL-BiLSTM) for load demand forecasting. The chosen CS is determined from a constructed graph based on user requests. The proposed technique effectively mitigates blackout risks and facilitates efficient EV charging operations, yielding promising results according to simulation outcomes.