基于MFO-ELM的荷电状态软测量模型

Soft Sensing Model of State-of-Charge Based on MFO-ELM

近年来,极限学习机(extreme learning machine,ELM)被广泛应用于解决非线性回归问题,针对SOC难以直接测量的难题,提出一种基于飞蛾火焰算法(moth-flame optimization algorithm,MFO)优化ELM的SOC预测模型。通过极小化预测误差来优化确定ELM参数的最优值并获得精确的SOC预测模型MFO-ELM,根据可测量电池电流、电压、温度和极化电阻参数来预测SOC值。随后,将MFO-ELM模型的性能与ELM模型进行了比较。结果表明:MFO-ELM预测精度高,误差不超过5%,是一种更好的SOC预测技术。

State-of-charge(SOC)is the equivalent of a fuel gauge for a battery pack in an electric vehicle.Determining the SOC becomes an important issue in all battery applications.However,the physical and chemical properties of the battery are complex;the SOC forecasting involves a rather complex nonlinear data pattern.In the recent years,the extreme learning machine(ELM)has been used widely to solve nonlinear regression.In order to solve the problem that SOC is difficult to measure directly,a SOC prediction model based on ELM optimized by moth-flame optimization algorithm(MFO)is proposed.MFO is adapted in this study to determine the optimal values of ELM parameters by minimizing prediction error.An accurate predictive model of MFO-ELM is obtained to forecast the SOC in the short term.The SOC of a battery is estimated from measurable battery parameters such as current,voltage,temperature and polarization resistance.Later,the performance of the MFO-ELM model is compared thoroughly with the ELM forecasting model.The results indicate that MFO-ELM is a better SOC prediction technology with high prediction accuracy and less than 5%error.