A dynamic programming framework for optimal home scheduling

Abstract

Adoption of smart devices, distributed generation and home energy management systems is expected to increase among proactive energy users due to their ability to reduce electricity bills and minimize energy wastage. Critical component to successful deployment of these technologies is a tried-and- tested solution to the optimal scheduling problem for an end-user. This paper proposes a dynamic programming (DP) framework for the optimal home scheduling problem involving rooftop solar generation, controllable loads, and battery storage, while considering uncertainties arising from solar generation and load forecasts. A linear programming solution is devised to solve the stochastic DP equations in a computationally tractable manner. The proposed approach is validated against a typical mixed integer linear program for home scheduling. Results presented emphasize the effectiveness of the proposed scheme in handling actual scheduling problems.