A Review of Central Pattern Generator-Based Approaches for Quadruped Robot Locomotion

Resumo

This paper reviews locomotion control approaches for quadruped robots, with a specific focus on methods based on Central Pattern Generators (CPGs). CPGs are bio-inspired neural circuits that generate rhythmic motor commands, offering a computationally efficient solution for controlling complex robotic movements. The study explores fundamental concepts of quadruped locomotion, including gait types and parameters and details the application of CPGs, highlighting the use of different oscillators, such as Hopf and Van der Pol, and their integration with Inverse Kinematics. The review emphasizes the potential of hybrid control techniques, which combine CPGs with learning-based methods to enhance the robot's adaptability and robustness on various terrains. The paper concludes that CPGs, especially when combined with hybrid control strategies, present a promising and efficient solution for achieving stable and adaptive locomotion in quadruped robots.