状态不确定性下安全性的递归CBF框架
A Recursive CBF Framework for Safety under State Uncertainty
浏览论文内容
中文总结 AI 辅助
针对状态不确定性下CBF安全控制缺乏递归可行性的问题，提出强制执行安全子集前向不变性的递归CBF框架，确保系统始终存在鲁棒安全控制输入。
中文摘要 AI 辅助
控制屏障函数（CBFs）在安全关键控制中的实际应用常常受到状态知识不确定性的阻碍。现有的鲁棒CBF方法虽然解决了状态不确定性问题，但往往缺乏递归可行性保证，或在不确定性水平较高时失效，从而使系统进入不存在安全控制输入的区域。为解决这一问题，我们提出了一种强制执行递归CBF的新框架。该方法不仅仅是确保原始安全集的不变性，而是强制执行安全区域中保证存在鲁棒安全控制输入的子集的前向不变性。这一整体框架确保系统永远不会偏离到模糊区域，无论状态不确定性水平如何，都能提供持续的可行性和安全性保证。
英文摘要
The practical implementation of Control Barrier Functions (CBFs) for safety-critical control is often hindered by uncertainty in the knowledge of the state. While existing robust CBF methods address state uncertainty, they often lack recursive feasibility guarantees or fail when uncertainty levels are high, allowing the system to enter regions where no safe control input exists. To resolve this, we propose a novel framework of enforcing recursive CBFs. Rather than merely ensuring the invariance of the original safe set, this approach enforces the forward invariance of a subset of the safe region where a robustly safe control input is guaranteed to exist. This holistic framework ensures that the system never strays into ambiguous regions, providing continued feasibility and safety guarantees, regardless of the level of state uncertainty.
发表机构
- University of California at Los Angeles（加州大学洛杉矶分校）
- California Institute of Technology（加州理工学院）
机构由 AI 辅助整理，请以论文原文为准。
