Comparative Analysis of Energy Transfers and Performance in Safety-Critical Control Using Control Barrier Functions
Arturo Maiani, Federico Califano, Lorenzo Govoni, Antonio Pietrabissa
AI summary
Problem
Safety-critical control for robots relies on Control Barrier Functions (CBFs), but existing designs offer conflicting trade-offs between stability guarantees and task performance. This paper investigates how different CBF formulations affect energy transfers and closed-loop performance during safety-critical maneuvers.
Approach
The authors theoretically analyze the power injection and passivity properties of energy-based and exponential CBFs, then empirically compare their performance through simulations and physical experiments on robotic manipulators navigating obstacle avoidance tasks.
Key results
- Proves energy-based CBFs preserve closed-loop passivity but restrict large state reconfigurations due to negative power injection
- Demonstrates exponential CBFs enable smoother control and better obstacle avoidance despite lacking formal stability guarantees
- Identifies conditions where exponential CBFs can create undesired stationary points near safety boundaries
- Validates theoretical trade-offs through simulations on a 3R planar robot and real-world experiments on a 7-DoF KUKA LWR manipulator
Why it matters
Guides roboticists and control engineers in selecting appropriate CBF designs based on whether stability preservation or task performance is prioritized in safety-critical applications.
Abstract
Control barrier functions (CBFs) are used in safety-critical control strategies, implementing a modification of a nominal control action to achieve invariance of a subset of the state space representing safe operating conditions. In this paper we perform a comparative study involving existing safety-critical CBF designs, including energy-based CBFs and Exponential CBFs. The analysis, performed both theoretically and on a benchmark obstacle avoidance task, provides insights into how these CBFs affect energy transfers and the overall performance of the closed-loop system, highlighting benefits and limitations of each approach. To validate our analysis, we conduct software simulations on a 3R planar robot and a 7-DoF robotic manipulator, complemented by experimental evaluations on a physical robotic platform.