Demand Control Management in Micro-grids: The Impact of Different Policies and Communication Network Topologies

This paper studies how the communication network affects the power utilization and fairness in a simplified power system model, composed by three coupled layers: physical, communication and regulatory. Using an agent-based approach, we build a scenario where individuals may cooperate (by removing a load) or not (by keeping their loads or adding one more). The agent decision reflects its desire of maximizing the delivered power based on its internal state, its global state perception, a randomized selfishness related to its willingness to follow demand-side control requests, and the state information received from its neighbors in the communication network. Our focus is to understand how the network topology and errors in the communication layer affect the agents’ behavior, reflected in the power utilization and fairness for different demand-side policies. Our results show that for optimal power utilization and fairness a global knowledge about the system is needed. We show that close to optimal results can be archived with either a demand control signal or a global pricing for energy.