中国机械工程 ›› 2014, Vol. 25 ›› Issue (19): 2692-2697.

• 机械基础工程 • 上一篇    下一篇

基于马尔可夫链的混合动力汽车模型预测控制

孟凡博;黄开胜;曾祥瑞;张尧   

  1. 清华大学汽车安全与节能国家重点实验室,北京,100084
  • 出版日期:2014-10-10 发布日期:2014-10-14

Model Predictive Control for Hybrid Electric Vehicle Based on Markov Chain Approach

Meng Fanbo;Huang Kaisheng;Zeng Xiangrui;Zhang Yao   

  1. State Key Laboratory of Automotive Safety and Energy,Tsinghua University,Beijing,100084
  • Online:2014-10-10 Published:2014-10-14

摘要:

采用模型预测控制的思路,选取加速踏板位置不变、指数衰减、按马尔可夫链随机分布三种不同的车辆未来转矩需求预测模型,运用动态规划算法进行转矩分配的优化,并与基于规则的混合动力控制策略进行对比。仿真结果表明,马尔可夫链模型预测控制在混合动力车辆能量分配上具有优化潜力。最后提出了模型预测控制应用于混合动力汽车能量分配的进一步研究方向。

关键词: 混合动力汽车, 模型预测控制, 能量分配, 动态规划

Abstract:

This paper used model predictive control method and choosed three different torque prediction models such as the acceleration pedal invariability, exponential decay and Markov stochastic process. Based on different prediction models, the dynamic optimization method could be used to optimize the toque ratio with comparison of the hybrid control strategy based on rules. The simulation results indicate that model predictive control for hybrid electric vehicle energy distribution has potential to improve the fuel economy. At last, this paper gives out the advanced research direction for the application of model predictive control on hybrid electric vehicle energy distribution.

Key words: hybrid electric vehicle, model predictive control, energy distribution, dynamic programming

中图分类号: