A new decentralized architecture and a fully parallel distributed algorithm for charge scheduling of plug-in electric vehicles

Abstract

In this paper, a distributed charge scheduling problem for plug-in electric vehicles (PEVs) is considered. It is shown that the problem of charge scheduling of PEVs in the centralized framework can also be recast into a Resource Allocation problem. A differentiation from existing literature is that this paper presents a new configuration of network architecture connecting charging station to PEVs for which a discrete-time fully parallel distributed primal-dual algorithm with fixed step size is proposed. In the proposed configuration, the existence of aggregator is removed, and the interactions are allowed among individual PEVs as well as between PEVs and the smart charging station (SCS) while private information of each PEV and the SCS is preserved. The heterogeneity in PEV charge specifications is considered in the formulated problem. The distributed algorithm can converge to an optimal solution of the problem for any convex cost functions of PEVs/utility with arbitrary initialization over any non-switching connected undirected networks without solving a sub-problem at each time step. Performance of the algorithm does not depend on the structure of the graph for convergence. Finally, a numerical example on load regulation is provided to demonstrate the results.