Output-Feedback Safe Navigation of High-Order Multiagent Systems Under Disturbances and Measurement Errors
研究了高阶移动多智能体系统在存在扰动和测量误差时，如何通过观测器估计未测量状态并设计控制器，实现安全导航控制目标。
This article studies the problem of output-feedback safe navigation for high-order mobile multiagent systems (MASs) in the presence of disturbances and measurement error. We provide a theoretical framework to extend the barrier Lyapunov function’s negative gradient method for safety-critical control to the case where accurate measurements of agent states are unavailable. Specifically, an observer helps estimate the unmeasured states, and these estimations are used in the control design process. We consider a worst case scenario together with a modified gradient-based virtual control to avoid sensitivity caused by estimation errors, and the system velocity is driven to track the virtual control. The safe navigation control objective is achieved even though the system states are not measured accurately, which is essential for realistic MASs with low-quality sensors. Moreover, this framework for general high-order systems is suitable for a wide range of agent dynamics. We conduct several numerical simulations to observe the behavior of the system under the output-feedback controller.