Optimal EV charge-discharge schedule in smart residential buildings

Abstract

The work in this paper provides an approach to generate the optimal charge/discharge schedule for an EV plugged into a smart building. The building is assumed to be equipped with local solar generation and to be subjected to time-varying electricity pricing schemes such as time-of-use pricing (TOUP) or real-time pricing (RTP). Artificial neural networks (ANNs) are used to forecast the building's hourly energy consumption and solar generation from the point in time when the EV gets plugged-in until it is expected to depart. The optimization is defined as a linear programming problem that is solved using MATLAB's optimization toolbox. Degradation costs are accounted for to ensure that the additional cycling of the EV battery, which accelerates battery degradation, does not offset the benefits of short-term optimization. The optimal schedule is constrained to ensure that battery charge/discharge operational limits are respected, and that the EV is fully charged by the time the user is ready to leave the residence. Simulation models developed in MATLAB coupled with real weather, load, and pricing data are used to evaluate the adequacy of the proposed algorithm.