汽车工程 ›› 2024, Vol. 46 ›› Issue (5): 784-794.doi: 10.19562/j.chinasae.qcgc.2024.05.005

• • 上一篇    下一篇

智能汽车轨迹跟踪多目标显式模型预测控制

赵树恩1(),王盛1,冷姚2   

  1. 1.重庆交通大学机电与车辆工程学院,重庆 400074
    2.武汉理工大学智能交通系统研究中心,武汉 430063
  • 收稿日期:2023-09-18 修回日期:2023-11-21 出版日期:2024-05-25 发布日期:2024-05-17
  • 通讯作者: 赵树恩 E-mail:zse0916@163. com
  • 基金资助:
    国家自然科学基金(52072054)

Multi-objective Explicit Model Predictive Control for Intelligent Vehicle Trajectory Tracking

Shuen Zhao1(),Sheng Wang1,Yao Leng2   

  1. 1.School of Mechatronics and Vehicle Engineering,Chongqing Jiaotong University,Chongqing 400074
    2.Intelligent Transportation Systems Research Center,Wuhan University of Technology,Wuhan 430063
  • Received:2023-09-18 Revised:2023-11-21 Online:2024-05-25 Published:2024-05-17
  • Contact: Shuen Zhao E-mail:zse0916@163. com

摘要:

针对现有智能汽车轨迹跟踪控制算法难以同时保证跟踪精确性、横向稳定性、舒适性以及控制实时性的问题,提出了一种基于多目标优化和显式模型预测控制理论的轨迹跟踪控制策略(MO-EMPC)。首先,建立考虑跟踪精确性、横向稳定性、舒适性的多目标函数及约束。然后,针对传统MPC控制实时性低的问题,设计基于EMPC的多目标优化轨迹跟踪控制器,通过引入多参数二次规划(MPQP)理论,将反复在线优化求解过程转化为等价的分段仿射系统(PPWA),离线计算得到最优显式控制律以供实时控制调用,减少在线运算时间。最后,基于CarSim/Simlink联合仿真方法,将所设计控制器的轨迹跟踪多目标优化效果与MPC控制效果进行对比验证。研究结果表明,所提出的轨迹跟踪策略在保证良好的跟踪精度前提下,横向稳定性、舒适性方面的表现更优于MPC控制器,且算法在线运行速度提高56.63%。

关键词: 汽车工程, 轨迹跟踪, 多目标优化, 智能车辆, 显式模型预测控制

Abstract:

For the problem that existing intelligent vehicle trajectory tracking control algorithms are difficult to simultaneously ensure tracking accuracy, lateral stability, comfort, and real-time control, a multi-objective optimized trajectory tracking control strategy (MO-EMPC) based on the theory of explicit model predictive control (EMPC) is proposed. Firstly, the multi-objective function and constraints considering tracking accuracy, lateral stability, and comfort are established. Then, for the problem of low real-time performance of traditional MPC control, a multi-objective optimized trajectory tracking controller based on EMPC is designed, which transforms the repeated online optimization and solution process into an equivalent segmented affine system (PPWA) by introducing in the multi-parameter quadratic programming (MPQP) theory, and calculates the optimal explicit control law offline for real-time control to be invoked, so as to reduce the online computation time. Finally, based on the CarSim/Simlink joint simulation method, the trajectory tracking multi-objective optimization effect of the designed controller is compared and verified with the MPC control effect. The results show that the proposed trajectory tracking strategy outperforms the MPC controller in terms of lateral stability and comfort under the premise of ensuring good tracking accuracy, with the online operation speed improved by 56.63%.

Key words: automotive engineering, trajectory tracking, multi-objective optimization, intelligent vehicle, explicit model predictive control