摘要
在电池管理系统中为了使荷电状态量SOC(state of charge)估计精确,提出以遗传算法优化最小二乘支持向量机(LS-SVM)的方法对电池的SOC进行预测的模型。在电池变流情况下对SOC进行研究,以标准工况下的实验数据作为样本,以电池的电流、电压及温度作为训练模型的输入,SOC作为输出建立模型,使之能很好地适用于混合动力汽车用电池在变电流状态下的实时SOC估计。研究结果表明:该预测模型预测精度高,其最大相对误差小于3%,平均相对误差小于2%,且与神经网络预测结果相比具有更强的实用性。
In order to make the SOC (state of charge) accurately estimate in the battery management system, the least square support vector machine (SVM) was used in the SOC estimation model. Under the varying current condition, the inputs of this model were current, voltage and temperature, and the output was the value of SOC, and the model was used in the hybrid electric vehicle. The results show that this model's forecasting precision is high, the max relative error is 3%, and the mean relative error is 2%. The model is more practical than the neural network.
出处
《中南大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2013年第1期135-139,共5页
Journal of Central South University:Science and Technology
基金
湖南省自然科学基金资助项目[11JJ3059]
关键词
混合动力
SOC预测
最小支持向量机
遗传算法
hybrid vehicle
SOC estimation
least square support vector machine (SVM)
genetic algorithm