Home energy management with generic thermal dynamics and user temperature preference

In this paper, we develop a novel home energy management solution that aims to minimize the electricity cost and guarantee user comfort in terms of preferred home temperature. Specifically, we consider a typical household system with a Heat Ventilation and Air Conditioning (HVAC) system and various types of loads. We formulate the home energy scheduling problem considering generic thermal dynamics represented by a look-up table, thermal comfort constraints, and specific characteristics of different electric loads. The assumed generic thermal dynamics overcome limitations of other approximate equation-based thermal dynamics typically employed in the literature. However, the empirical thermal dynamics makes the energy scheduling problem a complicated non-linear optimization problem, which is difficult to tackle. Therefore, we develop a decomposed solution approach where the scheduling of HVAC system and other loads are optimized in two different steps. We show that the HVAC scheduling problem is a dynamic programming problem and develop an algorithm to find its optimal solution considering the user comfort constraint Given the optimal HVAC scheduling solution, the scheduling problem for remaining loads is transformed into a mixed integer program whose solution can be found by using an available optimization solver. We then present numerical results to demonstrate the effectiveness and correctness of our proposed solution and its relative performance compared with the conventional design.