Research On Optimal Charging Scheduling For Electric Vehicles Based On Multi-agent Systems

Electric vehicles(EVs)can effectively alleviate environmental pollution and energy depletion.It is a general trend to develop EV vigorously.However,with large amounts of EVs plugged-in,there will be a negative impact on the power grid and traffic network.It is necessary to optimize the charging of EVs using coordinated charging strategies.Given that the centralized control mode has a large amount of computation and it is easy to cause "dimension disaster",a hierarchical management framework for EVs based on multi-agent system(MAS)is proposed in this paper.In the framework,there are three layers including distribution network,local operator and EV.Under the MAS,the optimization strategy based on improved greedy algorithm and the hierarchical optimal scheduling strategy based on distributed control are proposed,considering the node voltage security constraint.The former starts from the mathematical model of EV optimal scheduling problem and solves the problem by designing reasonable greedy strategy to avoid getting into local optimum and greatly reducing the calculating time.The latter starts from the control management framework and the mechanism and realizes the optimal scheduling of the EVs based on the distributed control method,which can solve the large-scale EVs optimal scheduling problem.Simulation results show that the two scheduling strategies based on multi-agent system not only have high computational efficiency but also take the interests of different agents into account,which are suitable for the optimization of electric vehicle scheduling.Uncoordinated charging of EVs will have great influence on traffic network.An intelligent charging and navigation system for EVs based on user's travel demand is proposed.Navigation model considering traffic impedance,charging cost and charging waiting time is established to coordinate and optimize the traffic network,distribution network and EVs.Research results show that the comprehensive optimal strategy is able to realize the coordinated management of transportation network,distribution network and EVs,such as balancing the traffic of various sections,improving voltage regulation and reducing travel distance,charging fees and waiting time of EVs.Finally,an initial charging simulation platform based on MAS is established by using Java on the Eclipse platform to validate the hierarchical architecture and coordinated scheduling strategy of EVs.