On Distributed Optimization for Supply Demand Coordination in Cyber Physical Energy Systems

It is of great practical interest to coordinate the stochastic supply and uncertain demand on electricity in a cyber physical energy system (CPES) such as urban energy internet and micro grid of buildings, especially by scalable distributed optimization algorithms. We consider this important problem in this paper and make the following major contributions. First, we consider the coordination between the supply of wind power generation and the charging demand from a fleet of shared electric vehicles in urban cities. The problem is formulated as a Markov decision process with average cost over finite stages. Second, we propose index policy, which is suitable for distributed implementation. Third, we numerically compare the index policy with Q-learning algorithms on case studies. The results show that index policies are scalable and achieve good performance in general. We hope this work sheds insight on distributed optimization for supply demand coordination in CPES in general.