Multi-Phase Joint-Angle Trajectory Generation Inspired by Dog Motion for Control of Quadruped Robot

Area of Science:

• Robotics
• Control Systems
• Biomechanics

Background:

• Quadruped robots offer advantages in speed, payload, and stability over humanoids, but complex locomotion and gait transitions remain challenging.
• Existing trajectory generation methods often lack sufficient degrees of freedom for seamless gait transitions or fail to account for speed-dependent gait changes.

Purpose of the Study:

• To propose a multi-phase joint-angle trajectory generation algorithm for quadruped robots.
• To enable smooth and continuous transitions between various gait phases (walk, amble, trot, canter, gallop) across different locomotion speeds.

Main Methods:

• Expressing quadruped joint angles using cyclic basis functions with manipulated inputs.
• Formulating a synchronization function to ensure proper leg sequencing and ground contact.
• Utilizing fuzzy logic to manage discrete gait phase classifications and ensure continuous joint-angle trajectories during transitions.

Main Results:

• The proposed algorithm successfully generates joint-angle trajectories for multiple gait phases.
• Fuzzy logic integration facilitates smooth and continuous transitions between discrete gait phases.
• Simulation studies validate the effectiveness of the multi-phase trajectory generation method.

Conclusions:

• The developed algorithm provides a robust solution for complex gait transitions in quadruped robots.
• The approach enhances locomotion control by enabling seamless adaptation to varying speeds and gait requirements.
• This research contributes to more versatile and adaptable quadruped robot locomotion systems.