Adaptive gait for a quadruped robot on 3D path planning

Abstract

A legged walking robot is able to not only move on irregular terrain but also change its posture. For example, the robot can pass under above head obstacles with crouching. The purpose of this research is to realize efficient path planning with a quadruped robot. Therefore, the path planning is expected to extended in three dimensions because of the mobility of the quadruped robot. Even tough these advantages, some issues of the quadruped robot that are instability and workspace limitation complicate realizing the application. Especially, a deadlock is serious problem because it causes the walk stop and less controllability. The deadlock is brought by concurrence of instability and workspace limitation. In order to improve the deadlock, we propose the new free gait pattern for a quadruped robot, called TFG (trajectory following gait). The TFG intends to obtain high controllability like a wheel robot without the deadlock. The TFG additionally allows changing posture, during the walk, which is one of merits of a walking robot. In this paper, experimental result shows that the TFG improves the problems of a quadruped robot and is valuable for three dimensional path planning.