This research paper introduces the application of Collision Cone Control Barrier Functions (C3BFs) to enable safe legged locomotion in dynamic environments. The key highlights are:
The authors present the Quadratic Program (QP) formulation of C3BFs, referred to as C3BF-QP, which serves as a protective filter layer atop a reference controller to ensure the robots' safety during operation.
The C3BF-QP approach is demonstrated on both Quadruped and Bipedal legged robots, showcasing its effectiveness in avoiding collisions with vertical and horizontal obstacles, including static and moving obstacles.
The C3BF-QP controller is seamlessly integrated with baseline controllers, such as Convex MPC for Quadrupeds and ZMP Walking Preview Controller for Bipeds, highlighting its versatility and potential for real-world deployment.
The model-free nature of the C3BF-QP formulation, which primarily considers acceleration, enables its easy integration with various legged robot platforms.
Simulation results conducted on PyBullet demonstrate the C3BF-QP controller's ability to navigate legged robots through complex environments while ensuring collision avoidance.
The authors conclude by discussing plans to implement the C3BF-QP controller on real-world legged robots and explore its application in cluttered environments and confined spaces.
To Another Language
from source content
arxiv.org
Ключові висновки, отримані з
by Manan Tayal,... о arxiv.org 03-29-2024
https://arxiv.org/pdf/2309.01898.pdfГлибші Запити