ADMM-based decentralized demand response method in electric vehicle virtual power plant

To fulfill demand response to a grid's control command, electric vehicles (EVs) are usually managed via a virtual power plant (VPP), and how to manage the EVs for optimal response is a key issue. Such a problem is defined as an optimal tracking problem (OTP) in the literature, and an improved OTP model considering network constraints is studied in this paper. In order to circumvent the computational difficulty of a centralized solution and to save extra cost of centralized infrastructures, an alternating direction method of multipliers (ADMM)-based decentralized optimization algorithm is proposed. With this algorithm, the network constraints that couple different EVs' charging power together are relaxed, and the OTP model is thus decomposed into parallel single-EV charging subproblems that can be solved in a distributed manner. Case studies on a demonstration system have verified the effect of the proposed algorithm in terms of optimality and convergency. Impact of a penalty on the algorithm's convergency is also investigated, and a varying penalty strategy is suggested for better convergency.