摘要
根据铅酸蓄电池在放电过程中内部电化学反应导致外部电特性变化的特点,提出一种基于支持向量机原理的电解液密度辨识模型。利用支持向量机理论非线性回归的特性,简化测量电解液密度的过程,在恶劣环境下检测动力电池的电解液密度更显其优越性。预测实验表明,采用改进的交叉验证预测模型具有泛化能力强、稳定性好的特点,并且在小样本的条件下能达到预期的辨识精度。
This paper presents an identification model of electrolyte density,which based on the Support Vector Machine Theory,according to the feature that the chemical reaction of interior electrics leading to the characteristic change of exterior electrics in the discharge process of the lead-acid battery.This model simplifies the process of measuring the electrolyte den-sity by using the nonlinear regression characteristics of the Support Vector Machine Theory,and it works better when meas-uring the electrolyte density of power battery in severe environment.Prediction experiment shows that the improved cross-validation pridiction model is featured by good generalization capability and stability,and can reach the expected identifying accuracy on small sample.
出处
《计算技术与自动化》
2014年第3期27-30,共4页
Computing Technology and Automation
关键词
电解液密度
支持向量机
交叉验证
参数辨识
electrolyte density
support vector machine
cross validation
parameter identification
