摘要
利用支持向量回归算法(ε-SVR)非线性逼近能力强、收敛速度快、具有全局最优解的特性,对电动汽车动力电池的荷电状态(Stateof Charge,简称SOC)估计方法进行了研究,确立了动力电池的电压、电流、输入输出功率与SOC之间的非线性关系。通过对比BP神经网络算法进行了仿真实验。结果表明,利用ε-SVR算法能更准确地逼近实际SOC值,获得更好的估计效果。
Support vector regression algorithm （ε-SVR）,which has good nonlinear approximation ability,quick convergence rate and global optimal solution, is proposed to estimate the battery state of charge （SOC）, and the nonlinear relationship between batteries＇voltage,current,input and output power and SOC is established.By comparing with the BP neural network algorithm, the simulation and experiment have been carried on.The results indicate that the SVR algorithm can more accurately approximate the actual SOC value and obtain better estimative performance.
出处
《电力电子技术》
CSCD
北大核心
2009年第5期78-80,共3页
Power Electronics
基金
国家自然科学基金项目(60874016)
高等学校博士学科点专项科研基金(200804220047)
山东省优秀中青年科学家奖励基金(2007BS01012)~~
关键词
电动汽车
电池/支持向量回归
荷电状态估计
electric vehicle
battery / support vector regression
state of charge estimation
