Two-stage stochastic home energy management strategy considering electric vehicle and battery energy storage system: An ANN-based scenario generation methodology

Abstract

This study implements two-stage stochastic programming in a smart home application to reduce the electricity procurement cost of an ordinary household. In this concern, vehicle to home (V2H) capability of the available electric vehicle (EV) is used in coordination with battery energy storage system (BESS) under control of a home energy management system. The stochastic decision variables are the charge-discharge power of these components. The uncertainties derived from the power production of the roof-mounted solar photovoltaic panels, household’s load demand, real-time electricity price are assimilated into the problem. Besides, to create the stochastic process, an artificial neural network (ANN) is trained using historical time series. Furthermore, as one of the main contributions, a proper analytical battery degradation cost model is integrated into the problem. Hence, different schemes such as with and without degradation cost, with and without BESS and uncoordinated charging are investigated under various charging rates. Also, the sensitivity of the problem for different charging rates of the EV and BESS is analyzed. Furthermore, the influence of probable future battery storage cost reductions on the home energy management system is investigated. Eventually, the efficiency of the stochastic programming method is analyzed by the value of stochastic solution (VSS) metric.