摘要
新能源汽车电池剩余使用寿命(Remaining Useful Life,RUL)预测不准确会导致电池失效,增加交通事故的风险。在开展电池RUL预测时,若不能及时对电池质量数据去噪,会使预测精度降低。为有效提升电池剩余寿命预测精度,提出粒子群算法下新能源汽车动力电池RUL预测方法。通过分析动力电池的工作原理,获取电池工作过程中能耗衰减给电池寿命带来的影响,确定电池寿命预测影响因素,并对其分析确定电池现阶段衰退点,完成电池质量数据采集。通过对采集数据去噪,获取电池质量完整数据,结合Bi-LSTM网络构建新能源汽车动力电池剩余寿命预测模型,并使用粒子群算法对模型优化,实现对新能源汽车动力电池剩余寿命的精准预测。实验结果表明,上述方法电池剩余寿命预测时的均方根误差低于0.025%,且预测效果好。
Inaccurate prediction for the RUL of batteries may lead to battery failure and increase the risk of traffic accidents.If the noise of battery quality data cannot be removed in time during the RUL prediction of the battery,the prediction accuracy will be reduced.In order to effectively improve the accuracy of prediction,a method for predicting the RUL of power batteries of new energy vehicles based on particle swarm optimization algorithm was proposed.After analyzing the working principle of power battery,we could get the impact of energy consumption degradation on battery life in the process of battery operation and then determine the influencing factors of prediction.Meanwhile,we ana-lyzed the current decay point of the batery,thus completing the data collection of battery quality.Moreover,we got complete battery quality data by denoising the collected data.After that,we built a prediction model of RUL by the Bi-LSTM network and used the particle swarm optimization algorithm to optimize the model,thus achieving accurate prediction for the RUL of new energy vehicle power battery.Experimental results show that the root mean square error of the method in predicting the remaining battery life is less than 0.025%,so the prediction effect is good.
作者
王丹
倪龙飞
王福忠
WANG Dan;NI Long-fei;WANG Fu-zhong(School of Intelligent Engineering,Huanghe Jiaotong University,Jiaozuo Henan 454950,China;School of Electrical Engineering and Automation,Henan Polytechnic University,Jiaozuo Henan 454950,China)
出处
《计算机仿真》
北大核心
2023年第11期77-81,共5页
Computer Simulation
基金
河南省科技攻关项目资助(232102241028)
河南省工程技术研究中心(266)。
关键词
粒子群算法
新能源汽车
动力电池
剩余使用寿命
预测方法
Particle swarm optimization algorithm
New energy vehicles
Power battery
Remaining service life
Prediction method
