Safety Critical Control of a Quadrupedal Robot Using MPC-CBF

Quadrupedal robots have been deployed in numerous challenging environments, such as nuclear facility inspections and industrial surveys. These unstructured and dynamic environments expose the robots to significant safety risks. Therefore, this paper proposes a novel design for a safety-critical control system for quadrupedal robots, integrating Model Predictive Control (MPC) with Control Barrier Functions (CBF) to enhance overall system safety. A comprehensive simulation environment has been established, and results demonstrate that the proposed MPCCBF controller outperforms the conventional MPC controller. Specifically, while the MPC controller alone provides adequate tracking of a desired trajectory in a trot gait, integrating CBF enhances obstacle avoidance and improves stability. This combination allows for navigation through more challenging environments while maintaining accurate trajectory tracking.