Abstract:This article addresses the problem of safety-critical control for multiagent systems with unknown dynamics in unknown environments. It has been shown that stabilizing affine control systems to desired (sets of) states while optimizing quadratic costs subject to state and control constraints can be reduced to a sequence of quadratic programs (QPs) by using control barrier functions (CBFs) and control Lyapunov functions (CLFs). One of the main challenges in this approach is obtaining accurate system dynamics of all components in the system, which is especially difficult when online model identification is required given limited computational resources and system data. We address this problem by proposing a robust framework (to unknown dynamics including uncertainties) through defining adaptive affine control dynamics that are updated based on the error states obtained by real-time sensor measurements. We define a CBF for a safety requirement on the unmodeled agents based on the adaptive dynamics and error states, and reformulate the safety-critical control problem as the abovementioned sequence of QPs. Then, we determine a set of events that trigger the QPs and ensure safety when solving them. We also derive a condition that guarantees the satisfaction of a CBF constraint between events. The proposed framework can also be used for state convergence guarantees for systems with unknown dynamics based on CLFs. We illustrate the effectiveness of the proposed framework on a robot control problem, an adaptive cruise control problem and a traffic merging problem using autonomous vehicles. We also compare the proposed event-driven method with the classical time-driven approach.

Model-Free Safety-Critical Control for Robotic Systems

Whole-body Dynamic Collision Avoidance with Time-varying Control Barrier Functions

Safe Tracking Control for Free-Flying Space Robots Via Control Barrier Functions

A General Safety Framework for Learning-Based Control in Uncertain Robotic Systems

Safety‐critical Control for Robotic Systems with Uncertain Model Via Control Barrier Function

Robust Tracking Control of Robot Manipulators with Actuator Faults and Joint Velocity Measurement Uncertainty

Adaptive Admittance Control for Safety-Critical Physical Human Robot Collaboration

Multi-Layered Safety-Critical Control Design for Robotic Systems Via Control Barrier Functions

Toward model-free safety-critical control with humans in the loop

Safety-critical control for autonomous systems: Control barrier functions via reduced-order models

Robust Safety-Critical Control for Dynamic Robotics

Safety-Critical Optimal Control for Robotic Manipulators in A Cluttered Environment

Reinforcement Learning for Safe Robot Control using Control Lyapunov Barrier Functions

Safe Control for Nonlinear Systems With Stochastic Uncertainty via Risk Control Barrier Functions

A Two-Layer Adaptive Safety-Critical Control Framework for Uncertain Robot Contact Tasks*

Generative Modeling of Residuals for Real-Time Risk-Sensitive Safety with Discrete-Time Control Barrier Functions

Modular Safety-Critical Control of Legged Robots

Learning for Safety-Critical Control with Control Barrier Functions

Adaptive Safety-critical Control with Uncertainty Estimation for Human-robot Collaboration

Event-Triggered Control for Safety-Critical Systems With Unknown Dynamics