Efficient Management of Electric Vehicle Parking Lot Allocation in Day-ahead and Reserve Markets Through the Application of Predefined Time Adaptive Neural Network

The rapid growth of electric vehicles (EVs) is reshaping energy demand and presenting significant challenges in managing charging operations efficiently and cost-effectively. Unpredictable EV arrival and departure times often result in grid imbalances and increased operational costs. To address these challenges, this research proposes a hybrid method, MBGO-EPTANN, which combines the multiplayer battle game-inspired optimizer (MBGO) for dynamic scheduling and the efficient predefined time adaptive neural network (EPTANN) for accurate electricity demand prediction within EV parking lots. The EPTANN component enables precise planning of charging and discharging schedules, while MBGO optimizes the scheduling to minimize grid costs and maximize overall profit. The method is implemented and evaluated against existing techniques on the MATLAB platform. In terms of total cost, the proposed MBGO-EPTANN approach records €27,567, substantially lower than Copula generative adversarial network (CGAN) (€47,577), particle swarm optimization (PSO) (€49,567) and remora optimization algorithm (ROA) (€55,351). This demonstrates the proposed approach’s potential to enhance EV charging management by reducing costs, improving grid efficiency and supporting better integration of EVs into the power system.