An XGBoost-based Electric Vehicle Battery Consumption Prediction Model
Tao Ma, Yusen Zhang, Xiangxin Nie, Xinchao Zhao, Yexing Li
Abstract
Electric vehicle battery management system is the current energy control manager of new energy vehicles, responsible for the task of real-time monitoring of various parameters of the entire vehicle battery. The State of Charge (SOC) is an important part of the battery management system for electric vehicles. The accurate estimation of SOC determines the charging and discharging control of electric vehicles and the optimization management strategy of vehicle driving to a certain extent, which directly affects the service life of the vehicle battery and thus the driving performance of the vehicle. Therefore, the study of SOC prediction for electric vehicles is of great significance. Due to the complex internal structure of the battery, this article mainly uses the principal component analysis method to select battery parameters for dimensionality reduction and specifically considers the battery terminal voltage, discharge current, and battery temperature. The impact of the XGBoost algorithm is then used to build a prediction model to establish a SOC prediction model for SOC estimation.