Power Battery Temperature Prediction Based on Charging Strategy Classification and Improved Adaptive GA-BP

During the charging process of electric vehicles (EVs), the temperature of the power battery plays a critical role in ensuring safety. Excessive heat can accelerate battery aging, leading to potential safety hazards. Therefore, accurate prediction of the temperature of the power battery is essential to effectively prevent overheating. However, due to the charging strategies of different types of vehicles, accurately modeling and simulating temperature prediction for automotive battery packs has been challenging, resulting in limited prediction accuracy. To address this issue, this article proposes a power battery temperature prediction method based on charging strategy classification and BP neural network by leveraging existing charging data from EVs. First, the k-nearest neighbor classification algorithm, utilizing a Gaussian kernel function, is employed to classify the charging strategies. Subsequently, the scaling factor, chaotic mechanism, adversarial learning strategy, and adaptive maximum mixed correntropy Kalman filter are combined with adaptive genetic algorithm (GA), is applied to optimize the BP neural network for temperature prediction. The proposed method is validated using actual charging data, the results demonstrate its ability to enhance the prediction accuracy of power battery temperature. This research provides a promising solution for accurately predicting battery temperature by considering charging strategies, contributing to the improved safety of EVs.