Research on Data-driven Cluster Analysis of Electric Vehicle Charging Sessions

With the rapid increase in electric vehicles (EVs) in recent years due to their energy saving and emission reduction characteristics, the mismatch between EV charging facilities and EVs is becoming more and more evident. Mining EV massive behavioral data to explore user behavior characteristics is of great significance for the construction and renovation of charging facilities in the urban planning process. This paper proposes a clustering method for EV charging behavior based on random forest and Gaussian mixture model, aiming to mine typical charging behavior patterns from big data. Firstly, the historical EV charging data is preprocessed to extract valid information and improve data quality. Secondly, a feature subset is selected based on the random forest algorithm to obtain the input sample set for clustering. Then, the Gaussian mixture model is used to cluster the dataset output by the random forest. Finally, the proposed method is validated using actual charging data from a certain area in Jiangsu Province. The results show that the proposed method can effectively extract charging behavior clusters with different feature distributions from massive data, providing a feasible idea for mining vehicle networking big data.