Analysis of prediction and clustering for uncertainty of EV charging station behavior on V2G platform

Abstract

The charging behavior of electric vehicles has uncertainty, which leads to uncertainty in the load of the V2G platform. In this paper, a clustering model of charging load at the charging station of the V2G platform containing uncertainty factors is put forward, and analyze the uncertain charging behavior and charging power characteristics. To ensure the reliability of the uncertainty quantification results, an efficient Latin Hypercube Sampling method is chosen to make the extracted samples comprehensive and homogeneous. The load prediction of electric vehicles charging behavior with uncertainty is carried out by Monte Carlo method. Finally, the typical load data of electric vehicles charging stations in Chongqing V2G platform is used as an example to verify the superiority of the uncertainty charging load clustering and prediction method proposed in this paper.

Keywords

• Electric vehicle station
• Uncertainty in charging behavior
• Load clustering algorithm
• V2G platform