A Path Planning Method for Quadruped Robot Based on 3D Elevation Diagram

Abstract

Quadruped robots, due to their superior flexibility and accessibility, have attracted considerable interest and are progressively being incorporated into real-world manufacturing processes. Given the vertical motion capabilities of these robots and the high complexity of 3D environments, path planning for them poses a significant challenge. This paper introduces a path planning framework for quadruped robots that utilizes elevation grid maps. Initially, the improved RRT-Connect algorithm and heuristic random state search are employed to accomplish global path planning tasks on the pre-existing elevation grid map. Subsequently, local trajectory planning is executed using the enhanced TEB algorithm. The framework is then tested on an actual robot platform, with the results demonstrating its efficiency and effectiveness in accomplishing path planning tasks in simulation and real-world conditions.