Motion Planning and Control with Environmental Uncertainties for Humanoid Robot

Abstract

Humanoid robots are typically designed for static environments, but real-world applications demand robust performance under dynamic, uncertain conditions. This paper introduces a perceptive motion planning and control algorithm that enables humanoid robots to navigate and operate effectively in environments with unpredictable kinematic and dynamic disturbances. The proposed algorithm ensures synchronized multi-limb motion while maintaining dynamic balance, utilizing real-time feedback from force, torque, and inertia sensors. Experimental results demonstrate the algorithm's adaptability and robustness in handling complex tasks, including walking on uneven terrain and responding to external disturbances. These findings highlight the potential of perceptive motion planning in enhancing the versatility and resilience of humanoid robots in uncertain environments. The results have potential applications in search-and-rescue missions, healthcare robotics, and industrial automation, where robots operate in unpredictable or dynamic conditions.