MGL4QGR: Multi-Gait Locomotion of Quadruped Guide Robot via Teacher-Student Learning

To address the locomotion control issue of quadruped guide robots across multiple terrains, a teacher-student framework is adopted to learn the multi-gait locomotion policy. Specifically, the interaction forces between the individual and the guide robot are treated as privileged information for the teacher's policy, and the multi-gait stylized unit serves as proprioceptive information for the student's policy. The trained agent policy is then deployed in a zero-shot and no-tuning manner. The experimental results demonstrate that the quadruped guide robot can efficiently climb 25-degree slopes and 20-cm high steps and exhibit excellent omnidirectional tracking capability in trotting, walking, and pacing gaits.