Central Pattern Generator Driven Self-Modulated Trot Gait Control for Quadruped Robot Locomotion

Abstract

Central Pattern Generator (CPG) has had a significant impact on contemporary scientific research, particularly in quadruped robotics. Initially, CPG primarily controlled joint angles in quadruped robots, enabling basic locomotion patterns. Later, their application was extended to mapping within the robot's workspace, improving control precision. Unassisted quadruped robots face challenges in maintaining height due to gravity, resulting in deviations from predefined foot trajectories and an inability to achieve the desired duty factor during trot gait. This research aims at autonomous adjustment of gait cycles and duty factor using CPG feedback and simple parameter calculations. This paper presents a simplified mechanism that uses the foot position measured at the moment of a foot contact with the ground solely for self-modulated trot gait, while maintaining a predefined period and duty factor. The modulation of CPG involves a complex network of feedback loops and inter-parameter coupling to fine-tune gait patterns. This research used a collaborative simulation approach combining Matlab and RecurDyn to analyze the locomotion of a quadruped robot. At varying speeds, cycle times and leg lift heights, the quadruped robot consistently maintains the predetermined duty factor and period with minimal error.