Two-level agent-based scheduling for a cluster of heating systems

The ongoing decentralization of the electrical power system encompasses the installation of residential electrical loads and generators, such as electro-thermal heating devices. Clusters of these devices in combination with thermal storage can play a significant role in future power system control as they introduce load and generation flexibility into the system. Hence, control concepts which are attractive for both power system level and end user side are required. This work proposes a novel decentralized scheduling approach for electro-thermal heating devices based on a multi-agent system architecture. As a major novelty, the approach addresses both of the aforementioned appeals on two levels without central control. First, local planning with individual objectives is performed. Second, the heating devices cooperate to find the combination of local plans serving best a given energy system objective. This approach is fully decentralized and has a strong end user involvement. The performance of the proposed approach related to end user and system level objectives is compared to that of centralized approaches. Simulation results show that the new method can successfully satisfy end user objectives and provide energy fluctuation balancing. Finally, drawbacks compared to centralized scheduling are analyzed.