Stochastic Smart Charging of Electric Vehicles for Residential Homes with PV Integration

Using electric vehicles (EVs) in residential homes integrated with solar photovoltaic (PV) array can address many problems associated with environmental issues and energy demand. EVs are useful as long as they utilized with the smart energy system (SES) to optimally and smartly charge the EV batteries. In order to improve the SES performance for optimally and effectively charging the vehicle, the impact of uncertainties and random parameters have to be considered in the EV models. Furthermore, online control methods like model predictive control (MPC) should be used to eliminate the effect of the error uncertainties and random parameters model on the system performance during the actual operation. In this paper, the EV departure time and required energy consumption during driving are modeled by Markov chain and conditional probability. Also, the PV performance and home load demand are modeled by PVWatt model and adaptive neuro-fuzzy inference system (ANFIS) respectively. Afterward, an MPC is designed to optimally charge the EV, while maintaining the desired battery energy level. The simulation is performed and the results demonstrate the effectiveness and enhancement of the proposed method.