Efficient Motion Planning for Hybrid Locomotion of Wheeled-Legged Planetary Rovers

Wheeled-legged rovers combine the flexibility of legged robotic systems and the energy efficiency of wheeled mobile systems. This paper presents a framework that allows energy-efficient and highly maneuverable locomotion for wheeled-legged rovers. Firstly, terrain geometry is first obtained by the perception system. It is modeled and classified as small relatively regular terrains. Then, the configuration topology of the rover and motion characteristics of end-effectors are utilized to express complex gait patterns and redundant motion combinations. During in-situ exploration, the decision-making process could be simplified by establishing the mapping relationship between terrain classes and locomotion strategies. Finally, trajectories are generated to achieve planned motions while maintaining stability and low energy consumption. The approach is applied to the Terrain-Adaptive Wheeled-Legged (TAWL) rover. Experiments show that the rover is able to autonomously choose appropriate gaits and has great maneuverability and stability when navigating through challenging terrains. The swing amplitude is reduced by about 83.3% compared to static walking gaits. And the maximum climbable step height has increased by 38.9%.