Electric Vehicle Charging Demand Prediction Based on Traffic Flow Volume and Fuzzy Reasoning

Abstract

Although our country's electric vehicle (EV) industry is booming, the coverage rate of EV charging facilities on highways is still low. Therefore, the key step in charging station site planning is to scientifically and rationally plan highway EV charging stations while determining the charging demand of EVs on highways. Focusing on highway EV charging demand, this article proposes a prediction method based on traffic volume and fuzzy reasoning. First, the influencing factors of highway charging demand are analyzed from five aspects: road network structure, EV parameters information, road congestion information, weather temperature, and user decision. The highway topology structure, traffic volume model, charging probability model, and power consumption model are established according to different influencing factors. Second, the Road Section Transmission Model (LTM) calculates the traffic volume of different road nodes, and the user's travel decision is employed to determine the shortest path for EV travel. Then, the Adaptive Neural Network Fuzzy Algorithm (ANFIS) is utilized to analyze the power consumption and charging probability. Finally, the Monte Carlo algorithm is employed to obtain the charging load demand of EVs. User travel characteristics and the highway network structure are employed to establish a more accurate and practical model. The charging probability and power consumption models obtained based on the ANFIS algorithm are more in line with the actual situation. Compared with the singular use of the Monte Carlo algorithm, the charging load prediction result based on traffic volume and fuzzy reasoning method is more accurate and closer to the actual value.

Keywords: expressway, Electric vehicle, charging demand prediction, LTM model, fuzzy neural network model