EV Lithium Ion Battery Fault Diagnosis Dataset

To train and test two ANN models, two separate datasets are developed based on battery modelling in MATLAB. State of Charge (SOC), Temperature and Voltage are the input features in both datasets. Where as against output label column in first dataset 0 represents healthy battery and 1 represents unhealthy battery. Similarly in second dataset 0 represents healthy battery and 1 represents battery has over discharge fault in battery and 2 indicates short circuit fault in the battery.

为训练并测试两个人工神经网络（Artificial Neural Network, ANN）模型，本研究基于MATLAB平台的电池建模流程构建了两套独立数据集。两套数据集均以荷电状态（State of Charge, SOC）、温度与电压作为输入特征。其中第一套数据集中，输出标签列的0代表健康电池，1代表非健康电池；第二套数据集中，0同样代表健康电池，1代表电池出现过放电故障，2代表电池存在短路故障。