摘要
电动汽车驱动电动机的MAP图是传动系统构件选择与优化的重要依据,其直接影响整车动力性和经济性评价.在分析发动机统一模型(Willans模型)的基础上,通过变量映射将发动机Willans模型应用于电动机模型,建立了电动汽车驱动电动机的Willans模型.针对多重多项式拟合精度不高的问题,提出采用最小二乘支持向量机(LSSVM)方法拟合电动机的Willans模型,并利用粒子群算法优化LSSVM关键参数.计算结果表明:在较低可用平均有效压力区域,基于LSSVM的电动机Willans模型与多项式电动机Willans模型差别不大;在较高可用平均有效压力区域,基于LSSVM的电动机Willans模型明显优于多项式电动机Willans模型.
The efficient MAP of electric vehicle driving motor is very important for choosing andoptimizing transmission system components , which affects the dynamic performance and economy of whole vehicle. Base on the unified dimensionless engine model （ Willans line model） and variable mapping method, the Willans line model of electric vehicle driving motors was established. To solve the low accuracy problem of polynomial fitting , the least squares support vector machine （ LSused to establish Willans line model of electric motor. The results showthat the electric motor Willansmodels based on LSSYM and polynomial both have good accuracy in low available mean effective pressure region, whilethemodel baseon LSSYM has better accuracy in high available mean effectivepressureregion.
出处
《江苏大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2016年第4期381-385,共5页
Journal of Jiangsu University：Natural Science Edition
基金
中国博士后科学基金资助项目(2015M571680)
江苏省"六大人才高峰"项目(2014-JXQC-004)
江苏省普通高校研究生科研创新计划项目(CXLX13_677)
