摘要
锂离子动力电池是当前新能源汽车的主要储能方式,其容量预测非常重要。本文提出了一种利用差分整合移动平均自回归(ARIMA)模型和反向传播(BP)神经网络集成的方法来预测锂电池容量。首先,建立ARIMA模型,预测前期电池充放电容量数据;随后,建立相关的BP神经网络模型进行长期预测。用500组中的前420组数据预测后80组,预测结果实现R≈0.9,表示BP神经网络非常适合预测长期的锂电池容量数据。最后,利用安庆师范大学动力电池实验室提供的锂电池对提出的方法进行测试,实验误差在±5%以内,拟合值效果覆盖值也大于85%,预测效果良好。
Lithium-ion power battery is the main energy storage method for new energy vehicles at present,so the prediction of lithium battery capacity is very important.An autoregressive integrated moving average(ARIMA)model and a back propagation(BP)neural network integrated method were proposed in the paper to predict the capacity of lithium batteries.Firstly,the ARIMA model was established to predict the battery charge-discharge capacity data in the early stage.Then a longterm prediction model was set up by BP neural network.The BP neural network model was established to predict the last 80 groups based on the former 420 groups of data in the last 500 groups.The value of R in the prediction result was approximately 0.9,indicating that the BP neural network was very suitable for predicting the long-term capacity data of lithium batteries.The proposed method was validated by using the measurement data of lithium battery provided by the Power Battery Laboratory of Anqing Normal University.The experimental error is within±5%,and the coverage value of the fitting value is greater than 85%,and the prediction effect is good.
作者
张朝龙
卢阳
杨璇
胡靓靓
ZHANG Chaolong;LU Yang;YANG Xuan;HU Liangliang(School of Electronic Engineering and Intelligent Manufacturing,Anqing Normal University,Anqing 246133,China)
出处
《安庆师范大学学报（自然科学版）》
2022年第2期15-18,共4页
Journal of Anqing Normal University(Natural Science Edition)
基金
国家自然科学基金(51607004)
安徽省高等学校虚拟仿真实验教学项目(2019xfxm59)
安徽高校自然科学研究重点项目(KJ2020A0509)
国家级大学生创新创业训练计划项目(202010372049)。
