Abstract:

 The hexapod mobile robot, which can bear the heavy load and traverse in the unknown environment of the lunar surface, is a necessary device for lunar exploration. Although the hexapod robot walking over a rugged terrain can be accomplished by employing the ground contact and attitude information, and can make a reflex action when encountering the small obstacles, however, the obstacle avoidance based on visual information is an important issue in the field of the legged locomotion when faced with the insurmountable obstacles. For an electrically-driven and radially-symmetrical hexapod robot, the terrain modeling for the lunar surface is based on the information from the laser range finder. In this paper, a virtual trunk body model is presented to achieve the autonomous obstacle avoidance strategy in the unknown environment, and the real-time optimal feasible directions and minimum distance are acquired to avoid the obstacles. The path trajectory of the trunk body and foot trajectories are planned to walk toward the feasible direction. The experimental result shows that the hexapod robot can track the yaw angle and in real time accurately from the obstacle avoidance method to achieve the obstacle avoidance autonomously in the unknown environment. The key techniques will lay a foundation for the research of the heavy-duty multi-legged robots for lunar exploration.

Key words:  Hexapod robot, Lunar exploration, Terrain modeling, Obstacle avoidance