Distributed optimization algorithm for heat pump scheduling

This paper introduces a method that deals with Demand Response using a distributed optimization algorithm in a single or multiple building and supplier setup. On the household level, responsive loads are intelligently controlled in order to adapt to renewable power supply. In this paper, heat pump (HP) heating systems, in general one kind of controllable load, are considered as shiftable loads. These systems comprise an electrical heat pump and a thermal storage that can temporarily decouple the electrical demand of the heat pump from the building´s thermal demand. Thus, the heat pump load can be shifted with higher flexibility. Wind and solar energy generation or a combination of both is considered as the renewable supplier. The proposed distributed algorithm optimizes the schedules of demand and supply by minimizing the supplier´s imbalance costs and the disutility experienced by the customer. Applying Dual Decomposition (DD) leads to the single objectives for supplier and costumer. A global optimum for the overall system is identified through an iterative process based upon the local optima. The algorithm´s performance is validated in three different test scenarios including single and multiple buildings as well as suppliers.