• CN:11-2187/TH
  • ISSN:0577-6686

›› 2009, Vol. 45 ›› Issue (2): 36-40.

• 论文 • 上一篇    下一篇

基于多目标遗传算法的混合动力电动汽车控制策略优化

张昕;宋建峰;田毅;张欣   

  1. 北京交通大学机械与电子控制工程学院
  • 发布日期:2009-02-15

Multi-objective Optimization of Hybrid Electric Vehicle Control Strategy with Genetic Algorithm

ZHANG Xin;SONG Jianfeng;TIAN Yi;ZHANG Xin   

  1. School of Mechanical Electric and Control Engineering, Jiaotong University
  • Published:2009-02-15

摘要: 混合动力电动汽车是一个高度复杂的非线性系统,并且影响其控制策略的参数较多,要对这样的系统进行优化,常规的优化算法显得无能为力,模型的精确程度也直接影响了选取参数的可靠性。应用汽车动力性、排放性高级模拟分析软件AVL CRUISE,联合Matlab/Simulink软件,建立合动力电动城市客车整车动态性能仿真分析模型,以百公里油耗和排放指标为优化目标,运用多目标遗传优化算法,针对欧洲、日本及中国的城市公交循环工况对混合动力系统工作模式的选择和能量流的分配进行全局优化,减少了运算时间,获得一组可靠的可行解,精确地确定出控制逻辑参数。该解集在很大程度上同时提高了原车的燃料经济性和排放性能,并且为混合电动车的设计和控制提供了一个适宜的选择范围,设计者可以按不同的要求进行不同的方案选择。

关键词: 多目标遗传算法, 混合动力电动汽车, 控制策略优化

Abstract: Hybrid electric vehicle(HEV) is a very complicated non-linear system, whose performance is affected by lots of control parameters. To optimize this system, the routine optimization approach is inefficient, and the reliability of optima depends on the precision of model. An HEV bus dynamic simulation model is built for performance analysis by the interlinking of advanced software AVL CRUISE and MATLAB/Simulink. For achieving minimum fuel consumption and emissions, a multi-objective genetic algorithm(GA) optimization method is applied to getting the optima of work modes and energy distribution in different city bus cycles, which finds the compatible control logic parameters and saves much time. The feasible solution can improve the fuel economy and emissions simultaneously and provide wider choices for different requirements in HEV design.

Key words: Hybrid electric vehicle, Multi-objective genetic algorithm, Optimization of control strategy

中图分类号: