Abstract: Aiming at the serious impact of the uncoordinated charging of electric vehicles on the distribution network, this paper designs a multi-objective scheduling model for coordinated charging and discharging based on K-means clustering. Firstly, we take private cars as research objects for the uncertainty modeling of charging load. Secondly, according to the spatial distribution of the electric vehicle charging pile, the effective clustering is achieved, and the equivalent node and the corresponding agent are formed. The first stage model is constructed to minimize the deviation between the peak-valley difference and the scheduling of agents. At the same time, the second stage model takes the minimum user charging and discharging cost as objective and each electric vehicle charging power as decision content. Then, two objective functions achieve comprehensive optimal through simplified handling. Finally, we adopt particle swarm optimization algorithm on the MATLAB platform to solve the model. The example simulation results show that the proposed scheduling optimization model has remarkable effect in peak cutting and improving user economy.

Key words: coordinated charge, clustering, scheduling, particle swarm optimization