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

Abstract

To satisfy the safety constraints, this paper proposes a multi-layered safety-critical control framework for robotic systems based on control barrier functions (CBFs), which is composed of the safe kinematic control layer and the inverse dynamics control layer. Both layers incorporate the CBF constraints to ensure safety, and the corresponding quadratic programs (QPs) are formulated. Through real-time optimization control, the robot can satisfy the safety constraints while tracking the desired trajectory. The proposed method is validated on the two-link planar manipulator with trajectory tracking and obstacle avoidance tasks.