首页|考虑侧倾和输入饱和的自动驾驶汽车鲁棒多维切换路径跟踪控制

考虑侧倾和输入饱和的自动驾驶汽车鲁棒多维切换路径跟踪控制

Multi-dimensional switched path tracking control for autonomous vehicles considering roll and input saturation

扫码查看
原文链接
  • 维普
  • 万方数据
为解决自动驾驶路径跟踪控制中的侧倾稳定和执行器饱和等问题,提出了一种鲁棒多维切换H∞路径跟踪控制方法以平衡控制器设计难度与控制性能.首先,基于M3D切换系统框架和侧向载荷转移率,构建了工况随动的路径跟踪控制模型;其次,采用鲁棒不变集理论和辅助反馈矩阵方法处理输入饱和的非线性问题,设计了一种鲁棒M3D切换控制器;最后,在多项式二次Lyapunov函数和M3D切换系统理论框架下,给出了保证跟踪系统闭环渐近稳定的判据.在CarSim-Simulink联合仿真环境下,对所提方法的性能进行验证.结果表明,所提方法相比于现有的MPC控制方法,路径跟踪性能提升19.4％,侧倾稳定性能提升23.4％,能够保证车辆在高速大转向等极端工况下具备良好的路径跟踪性能和侧倾稳定.
To solve the problems of roll instability and actuator saturation in autonomous vehicle path tracking control,a robust multi-dimensional switched H ∞ path tracking control method was proposed to balance the complexity of controller design with control performance.Firstly,a condition-dependent path tracking control model was constructed based on the multi-mode multi-dimensional(M3D)switching system framework and the lateral load transfer rate.Secondly,the robust invariant set theory and the auxiliary feedback matrix meth-od were used to deal with the nonlinear problem of input saturation,and a robust M3D switching controller was designed.Finally,the criteria ensuring asymptotic stability of the tracking system were established within the poly-quadratic Lyapunov functions and the M3D switching system theory.The performance of the designed method was validated in a CarSim-Simulink co-simulation environment.The results demonstrate that the pro-posed method outperforms the existing model predictive control(MPC)control methods,with a 19.4％im-provement in path tracking performance and a 23.4％enhancement in roll stability performance.This strategy ensures that vehicles maintain excellent path tracking performance and roll stability under extreme conditions such as high-speed large turns.

autonomous vehiclepath tracking controlswitched systemroll stability

董锋威、殷国栋、徐利伟、采国顺、沈利霖、王磊、朱壮瑞

展开 >

东南大学机械工程学院,南京 211189

自动驾驶汽车 路径跟踪控制 切换系统 侧倾稳定

国家自然科学基金资助项目国家自然科学基金资助项目国家自然科学基金资助项目江苏省科技成果转化专项资金资助项目广西科技重大专项资助项目吉林大学汽车仿真与控制国家重点实验室开放基金资助项目清华大学汽车安全与节能国家重点实验室开放基金资助项目清华大学汽车安全与节能国家重点实验室开放基金资助项目

523942635202512152372379BA2021023AA2206800120210104KFZ2201KFY2215

2024

10.3969/j.issn.1001-0505.2024.05.024

东南大学学报(自然科学版)
东南大学

东南大学学报(自然科学版)

CSTPCD北大核心
影响因子：0.989
ISSN：1001-0505
年,卷(期)：2024.54(5)