摘要
为了保证电动汽车安全可靠,须对动力电池组的健康状态进行精确评估。研究以容量作为健康状态评价指标,以·150次循环试验中的表征参数数据作为数据样本,利用高斯过程回归方法,建立表征参数和容量之间的数学关系模型,得到动力电池组容量的演变规律,完成了基于数据驱动的动力电池组健康状态评估。模型仿真结果表明,模型最大相对误差为0.0137%,评估结果最大相对误差为0.3610%o残差分析结果表明,该高斯过程回归模型合理有效。
In order to ensure the safety and reliability of electric vehicles,the state of health of the power battery pack must be accurately evaluated.In the evaluation,the capacity is used as the evaluation index of the health state.The characterization parameter data in the 150-cycle-test is used as the data sample.By machine learning method,Gaussian process regression model between the characterization parameters and the capacity is built,in order to obtain the trend of the power battery pack capacity under the test conditions,and to realize data-driven power battery pack health assessment.The model simulation results show that the maximum relative error of the model is 0.013 7%,and the maximum relative error of the evaluation is 0.361 0%.The residual analysis results show that the Gaussian process regression model is reasonable.
出处
《机电一体化》
2019年第3期3-9,共7页
Mechatronics
关键词
动力电池组
数据驱动
健康状态评估
高斯过程回归
power battery pack
data-driven
evaluation of state of health
Gaussian process regression
