Volume 65

Virtual Acquisition Method of Charging Station Data Considering Similarity Day and Similarity Station Wenhui Yang, Kang Xiong, Minglei Bao, Yi Ding, Junchao Cheng, Xuanqi Wang

Abstract

The rapid development of electric vehicle (EV) has led to a widespread distribution of charging stations (CS). The real-time and accurate acquisition of EV charging load plays an important role in charging load forecasting and power grid dispatching. However, failures of power acquisition devices or communication lines lead to missing collected data at some time points, making the virtual acquisition of charging load of great significance. Currently, most studies on virtual acquisition focus on distributed photovoltaics, while few focus on that for charging load. To this end, a virtual acquisition method of charging station data considering similarity day and similarity station is proposed. The gray correlation degree and cosine similarity are used to form a similarity composite index for selecting similar days. The Pearson’s correlation coefficient is used to select similar stations considering historical data from each CS. A training set is constructed using the characteristic data of similar stations on similar days. After that, the LightGBM model is trained and perform virtual acquisition of charging station data within the region. Finally, the model is validated using actual data from a certain province in China as a case study.

Keywords virtual acquisition, electric vehicle, similarity day, similarity station