Research on Power Consumption Prediction of Electric Vehicle Charging Station

With the explosive development of electric vehicles, the demand for charging facilities has increased significantly, which has become a key factor affecting the planning of power grid construction. In view of the charging station number is larger, the overall amount of data is bigger, and a single station data dimension is low, the data quantity is little, study is less, the traditional methods can’t be applied, this paper has carried out the charging power prediction model based on deep learning research, to provide technical support for the new charging station planning.

I. Introduction

The electricity consumption of charging stations is closely related to surrounding charging stations, regional conditions, road factors, weather factors, etc. As a result, many unknown situations are caused, which makes the subsequent prediction research difficult. At present, whether domestic or foreign research on this issue still has not made much progress. On the one hand, traditional physical model analysis methods are difficult to establish suitable, accurate and complex physical models, and difficult to accurately analyze the correlation between various factors on the other hand. Therefore, it’s necessary to combine the historical data of existing EV charging stations and all kinds of external data, comprehensively apply the feature extraction algorithm and deep learning network model, and study the electric power consumption prediction model of EV charging stations [1].