Discrete Stochastic Control for Energy Management With Photovoltaic Electric Vehicle Charging Station

This paper develops an intelligent energy management system for optimal operation of grid connected solar powered electric vehicle (EV) charging station at workplace. The optimal operation is achieved by controlling the power flow between the photovoltaic (PV) system, energy storage unit, EV charging station (EVCS) and the grid. The proposed controller is developed considering the PV availability, grid loading and the EV charging load data. This information is modelled using Markov decision process (MDP) to develop a control strategy that eliminates the conventional problem of immediate recharging of energy storage unit after each EV charging by setting a target state of charge (SOC) level. This maximizes the use of PV power for EV charging and minimizes the impact on the grid. To test the operation of the proposed controller, a charging station powered by a 5 kW PV system with 35 kW energy storage unit connected to grid is developed through numerical simulations and experiment. The experiments were carried out for three different conditions under varying irradiance profile and load profile for multiple days. The results estimated the EV load and PV power and optimized the energy storage unit SOC between 0.3–1. Further, the energy management strategy minimized the impact of energy exchange between the grid and charging station by a factor of 2.