基于决策树的异常电池分类方法

提出了一种基于决策树算法的异常电池精准定位和分类方法，帮助人们在电池维修或更换时能够迅速确定故障单体的位置和类型，从而实施准确的维修更换方法，提高故障处理的效率。搭建电池仿真模型获取异常电池的充放电循环数据；以电压数据为基础，训练用于异常电池分类的决策树算法，使用试验数据和云端实车数据对构建的模型进行验证。验证结果表明，该方法能够准确判断异常电池单体在电池组中的位置和异常类型。在不同验证数据上，该方法的分类准确率高达98%以上，能够有效筛选出动力电池组中的异常电池。该结果说明了提出的决策树算法在动力电池异常分类中的有效性和准确性。

Classification of Defective Batteries Based on Decision Tree

This paper proposes a method based on decision tree to locate the defective battery cells and determine the defect type quickly. The proposed method is aimed at improving the efficiency and accuracy of battery repair. Initially a battery model was simulated to obtain the charging and discharging cycle data of damaged batteries. And then the decision tree algorithm for defective battery classification was trained by using voltage data. Finally the experimental data and the actual vehicle data in the cloud were used to verify the constructed model. The experimental results show that the classification accuracy of this method is more than 98% for different testing data, and the defective batteries in the power battery pack can be identified effectively. The results demonstrate the validity and accuracy of decision tree algorithm applied in power battery classification.