Risk Prediction of Power Battery Based on Logistic Regression Algorithm

Today, when we advocate the development of new energy vehicles, the improvement of battery management system is an important factor in the development of electric vehicles. Therefore, the estimation of the state of power batteries has become the key factor to ensure the safe driving of the vehicles. Traditional research on the low residual charge of power battery mostly uses the charge and discharge data of single battery under ideal experimental conditions, which may not be able to adapt to the real complex and changeable driving conditions and separate from the real driving environment. To solve this problem, this work relies on the national big data alliance of new energy vehicles and uses the data-driven method to build a time-varying, battery voltage, battery current and other angles to reflect the current working status of the remaining power (SOC) of the power battery. The type of power battery used in this paper is lithium iron phosphate battery, and the alarm group and prediction group are designed. The data of the alarm group are the collected power battery data detected as low residual power (SOC), and the data of the prediction group are the driving data of vehicles from May 9, 2019 to November 5, 2020. The logical regression algorithm is used to extract the battery characteristic information of low residual power, establish a model, and use the model to predict the data of the prediction group.