电池放电曲线的精准拟合探究

The Study of the Accurate Fitting of Battery Discharge Curve

为有效预测铅酸电池的剩余放电时间,探究了放电曲线的拟合问题。给出保障低电压段拟合质量的采样数据取舍方法及拟合函数选择原理,针对特定的精度评估指标MRE,设计完整的样本提取程序,运用微积分技术建立计算MRE的简化公式;全方位比较多项式拟合、样条拟合、线性回归、多元回归等各种拟合手段,指出其优缺点和适用条件,避免应用的盲目性;采用适当的计算方法,获得所涉3个问题的全部数值结果。

To predict the discharge time of lead-acid batteries, this paper explores the fitting of the dischargecurve of lead-acid batteries and puts forward the method of sampling data and the theory of choosing the fitting function. Based on MRE, the complete process of sampling is designed, and the simplified equation ofcalculating MRE is set up with calculus. The different fitting methods like linear regression, multiple regression, polynomial fitting and spline fitting are compared, and the advantages and the disadvantages and thecondition for application are pointed out to avoid blind application. With the proper algorithm, all the data involved in the three problems are achieved.