Convergence Speed Analysis and Optimization for Distributed Control of Virtual Power Plant

Distributed control can be applied to a virtual power plant (VPP) to allocate electrical power among distributed generators (DGs). By using consensus algorithm and selecting the incremental cost of each DG as the consensus variable, economic dispatch of the VPP can be realized in a distributed manner. Therefore, only local communication among DGs is required and some performance limitations caused by centralized control can be avoided as well. However, the slow convergence speed is of concern in the distributed control. In this paper, mathematical analysis is presented to reveal factors influencing the convergence speed of the consensus algorithm. Specially, a convergence speed optimization method is developed to reduce the iterations, including communication network (CN) topology designing, edge weights optimization, and leader selecting. The performance of the optimization method is demonstrated through several simulation cases.