Combined Potential Function and Graph Search Approach for Free Gait Generation of Quadruped Robots

The ability of a mobile robot to transport itself over various terrains is an important goal in developing of walking robots. This project aims to design a generic algorithm which would be able to plan the steps of a given walking quadruped robot over irregular terrain, i.e., given the kinematic structure of the robot and point of destination or desired navigation route, the algorithm will plan the set of steps required to perform the navigation task while keeping the robot kinematics constraints, balance constraints and motion characteristics that can be determined by the operator. The algorithm is based on a combination of two main mathematical tools, one is the graph search method, [1] and the other is the potential function, [2], [3]. This integration of these two tools allows dealing with the local minima problem. In this work a theoretical background is presented. In addition, the algorithm developed under this project is presented in details. Simulations results demonstrate the algorithm in a planner and irregular terrains (Figure 2). A preliminary experiment result of walking in a planner ground is presented (Figure 3).