Decentralized scheduling of PEV on-street parking and charging for smart grid reactive power compensation

Abstract

Power quality is one of the major concerns in modern power systems, especially within smart grids where the power distribution is more dynamic and vulnerable. With drastically more plug-in electric vehicles (PEV) penetrating into the existing power distribution system, Vehicle-to-grid (V2G) technologies have attracted increasing research attention. This paper explores the potential of managing the charging pattern of PEVs for smart grid reactive power compensation. With PEVs' bidirectional AC chargers viewed as mobile reactive power resources, the scheduling of PEVs for parking and charging at distributed on-street stations is formulated into a multi-objective resource allocation problem. One objective is that stations should be allocated with adequate and timely resources (PEVs parked with an appropriate charging pattern) to compensate the time-varying reactive power of the grid. The other objective is that PEV owners should be provided with satisfying parking services with as low monetary cost as possible. We solve this multiobjective optimization problem by using the Normalized Normal Constraint (NNC) method to obtain a set of well-distributed Pareto optimal solutions. A decentralized algorithm based on Lagrangian decomposition is then used to make the optimization scalable as the number of PEVs increases. Simulation results demonstrate the satisfying quality of the obtained Pareto optimal solutions, among which one will be selected by the optimization system according to the grid requirement on the power quality.