Real-time gait planning method for six-legged robots to optimize the performances of terrain adaptability and walking speed

Abstract

The terrain adaptability and the walking speed are two key performances of legged robots. However, methods to optimize these performances are seldom developed, especially in real-time application. This paper proposes a novel gait planning method for six-legged robots to optimize both the terrain adaptability and the walking speed. First, a six-legged robot with parallel mechanisms and the typical tripod gait are introduced. Then, the optimization problem is addressed by minimizing the displacements of the active joints and solved by applying the pseudo-inverse. Further, the performance improvements are analyzed by comparing the optimized gait with the non-optimized gait. Finally, experiments are conducted on the prototype to validate the method. The results show that the maximum AoS (angle of the slope) and the maximum SHR (step to body height ratio) are improved by 48.8% and 96.0%, respectively. The maximum SLR (speed to body length ratio) is improved by 17.1%. More importantly, all essential calculations are fulfilled within 1ms, which proves the feasibility of the method in real-time control systems.