Safe Control for Nonlinear Systems Under Faults and Attacks via Control Barrier Functions

• 成果类型: Article
• 署名作者: Zhang, Hongchao; Li, Zhouchi; Clark, Andrew
• 署名单位: Washington University (WUSTL)
• 刊物名称: IEEE TRANSACTIONS ON AUTOMATIC CONTROL
• ISSN/ISSBN: 0018-9286
• DOI: 10.1109/TAC.2025.3571338
• 发表日期: 2025
• 关键词: FAILURE COMPENSATION CONTROL; TOLERANT CONTROL-SYSTEMS; verification; flight; reachability; guarantees; STABILITY; tracking; DESIGN

摘要

Safety is one of the most important properties of control systems. Sensor faults and attacks and actuator failures may cause errors in the sensor measurements and system dynamics, which leads to erroneous control inputs and hence safety violations. In this article, we improve the robustness against sensor faults and actuator failures by proposing a class of fault-tolerant control barrier functions (FT-CBFs) for nonlinear systems. Our approach maintains a set of state estimators according to fault patterns and incorporates CBF-based constraints to ensure safety under sensor faults. We then propose a framework for joint safety and stability by integrating FT-CBFs with control Lyapunov functions. By utilizing redundancy, we proposed high order CBF-based approach to ensure safety when actuator failures occur. We propose a sum-of-squares-based approach to verify the feasibility of FT-CBFs for both sensor faults and actuator failures. We evaluate our approach via two case studies, namely, a wheeled mobile robot system in the presence of a sensor attack and a Boeing 747 lateral control system under actuator failures.