arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.04882math.OC
收藏

协同李雅普诺夫函数中的混合极小值搜索:未知控制方向下的鲁棒全局镇定

Hybrid Minimum-Seeking in Synergistic Lyapunov Functions: Robust Global Stabilization under Unknown Control Directions

Mahmoud Abdelgalil, Jorge I. Poveda

AI总结:

针对仿射控制系统中控制方向动态不确定性与拓扑障碍导致无光滑全局控制李雅普诺夫函数的问题,基于混合动态包含李括号平均结果提出混合极小值搜索反馈律,实现鲁棒全局实用稳定性,应用于流形镇定与车辆避障并验证有效性。

摘要:

我们研究仿射控制系统中的鲁棒全局镇定问题,重点关注控制方向的动态不确定性以及存在阻止光滑全局控制李雅普诺夫函数存在的拓扑障碍。基于文献[abdelgalil2023lie]中提出的混合动态包含的李括号平均新结果,我们提出一类新型通用混合反馈律,通过识别一组适当选择的协同李雅普诺夫函数的极小点实现鲁棒全局实用稳定性。作为结果的具体应用,我们针对未知控制方向下各类流形上的鲁棒全局镇定问题综合不同的混合高频高幅反馈律,并为描述完整和非完整模型的运动学模型车辆设计避障控制器。通过利用混合系统的李括号平均,我们还展示了所提混合极小值搜索反馈律如何克服持续(有界)时间内的可控性缺失问题。数值仿真结果用于说明主要结果。

英文摘要:

We study the problem of robust global stabilization in control-affine systems, focusing on dynamic uncertainties in the control directions \emph{and} the presence of topological obstructions that prevent the existence of smooth global control Lyapunov functions. Building on a recently developed Lie-bracket averaging result for hybrid dynamic inclusions presented in \cite{abdelgalil2023lie}, we propose a novel class of universal hybrid feedback laws that achieve robust global practical stability by identifying the minimum point of a set of appropriately chosen synergistic Lyapunov functions. As concrete applications of our results, we synthesize different hybrid high-frequency high-amplitude feedback laws for the solution of robust global stabilization problems on various types of manifolds under unknown control directions, as well as controllers for obstacle avoidance problems in vehicles characterized by kinematic models describing both holonomic and non-holonomic models. By leveraging Lie-bracket averaging for hybrid systems, we also show how the proposed hybrid minimum-seeking feedback laws can overcome lack of controllability during persistent (bounded) periods of time. Numerical simulation results are presented to illustrate the main results.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑