Terrain-Aware Ego-Planner for Local Path Planning and Navigation of Quadruped Robots

Abstract

Autonomous navigation in unknown environments remains a key challenge for quadruped robots in complex terrains. This paper proposes a path planning and navigation framework based on the Ego-Planner algorithm incorporating terrain height thresholds that account for quadruped locomotion adaptability. The planner formulates trajectory generation as a continuous optimization task with online refinement based on locally perceived environmental information, enabling effective traversal of uneven regions.