Optimization of Electric Vehicle Charge Scheduling in Multiple Parking Lots: A Method Based on Metaheuristics and High Performance Computing

This paper presents a centralized model based on metaheuristics to solve the problem of optimal Electric Vehicle (EV) charge scheduling in multiple parking lots. A centralized optimization model using two-level particle swarm optimization was developed, to find the optimal allocation of power between parking lots by a central aggregator, and to find the optimal EV charging schedule for each parking lot within the system, in order to minimize the cost to charge all EVs. High Performance Computing (HPC) techniques are used to combat the high resource requirements and long simulation time associated with the problem size, to increase the scalability of the solution, and to complete the optimization in an acceptable real-time interval. The parallelized centralized optimization model was validated against its sequential version and a decentralized model on an HPC cluster, where it provided more optimal, lower cost solutions than the decentralized model, and provided average computation speedups of up to 139 times faster than the sequential model. The model was tested with scenarios from 3 to 27 parking lots, on the MATPOWER 18, 33, 69, and 141-bus distribution systems.