欢迎访问《太阳能学报》官方网站,今天是 分享到:
ISSN 0254-0096 CN 11-2082/K

太阳能学报 ›› 2022, Vol. 43 ›› Issue (5): 518-524.DOI: 10.19912/j.0254-0096.tynxb.2022-0550

• • 上一篇    下一篇

基于深度神经网络的梯次利用电池健康状态评估

魏梓轩, 韩晓娟, 李炫   

  1. 华北电力大学控制与计算机工程学院,北京 102206
  • 收稿日期:2022-04-21 出版日期:2022-05-28 发布日期:2022-11-28
  • 通讯作者: 韩晓娟(1970—),女,博士、教授,主要从事规模化储能方面的研究。wmhxj@163.com
  • 基金资助:
    国家自然科学基金(61973114)

STATE OF HEALTH ASSESSMENT FOR ECHELON UTILIZATION BATTERIES BASED ON DEEP NEURAL NETWORK

Wei Zixuan, Han Xiaojuan, Li Xuan   

  1. School of Control and Computer Engineering, North China Electric Power University, Beijing 102206, China
  • Received:2022-04-21 Online:2022-05-28 Published:2022-11-28

摘要: 随着大量退役电池梯次利用,对退役动力电池健康状态的准确估计是保障电池梯次利用安全高效运行的前提。针对上述问题,提出基于深度神经网络学习的梯次利用电池健康状态评估方法。根据不同循环次数下梯次利用电池充放电性能的差异性,从梯次利用电池物理特性角度挖掘影响梯次利用电池老化特征的主要参数,利用皮尔逊法计算电池老化特征与梯次利用电池健康状态的相关系数,选取较高相关度特征作为深度神经网络的输入,建立基于深度神经网络学习的梯次利用电池健康状态评估模型。通过美国国家航空航天局Ames卓越预测中心的锂离子电池测试数据仿真实例验证了该文方法的有效性。仿真结果表明,与传统神经网络相比,深度神经网络学习可明显提高梯次利用电池健康状态的预测精度,为退役动力电池健康状态评估提供理论依据。

关键词: 梯次利用电池, 神经网络模型, 电池建模, 健康状态评估, 相关性分析, 电池老化特征

Abstract: With a large number of retired batteries being used for secondary use, an accurate estimation of the health status of retired power batteries is a prerequisite for ensuring safe and efficient operation of battery secondary use. To address the above problem, a deep neural network learning-based method is proposed to evaluate the health status of the retired batteries. According to the difference of the charge and discharge performance of the battery under different cycle times, the main parameters affecting the aging characteristics of the battery are excavated from the perspective of the physical characteristics of the battery, and the correlation coefficient between the aging characteristics of the battery and the state of health of the battery is calculated by the Pearson method, select the high correlation feature as the input of the deep neural network, and establish a battery health state evaluation model based on the learning of the deep neural network. The validity of the method was verified by simulating the lithium-ion battery test data from the NASA Ames Prediction Center of Excellence. The simulation results show that compared with traditional neural networks, the deep neural network learning can significantly improve the prediction accuracy of the health status of the battery and provide a theoretical basis for the evaluation of the health status of retired power batteries.

Key words: secondary batteries, neural network models, battery modelling, health state assessment, correlation analysis, battery

中图分类号: