Short-term EV load forecasting method considering complex meteorological factors

With the intelligent and modern development of power systems, the types of power electric vehicle (EV) loadsare increasing. Moreover, the influence of meteorological factors on EV load is becoming more prominent. This paper presents an EV load forecasting method that considers the influence of various meteorological factors. Firstly, principal component analysis is employed to analyze the relationship between different meteorological factors and EV load, and the meteorological factors with a greater impact on EV load are selected as characteristic variables.Then, a short-term EV load forecasting model considering meteorological factors is established by constructing a double-layer long short-term memory (LSTM) neural network. The ant colony algorithm is used to optimize the parameters of the double-layer LSTM neural network model to determine the optimal short-term EV load forecasting model. Finally, by comparing the short-term EV load forecasting effect maps and errors of two regions with and without meteorological factors, it is concluded that the forecast algorithm proposed in this paper has higher prediction accuracy.