A Real-time Planning and Control Strategy of Humanoid Robot Walking on Irregular Terrain

Improving environmental adaptability is a key to the humanoid robot research. A strategy of real-time planning and control was proposed in the paper based on F/T information feedback. The robot lands it's whole sole on the ground by adopting ankle's moment control, calculates the parameters of gait and plans current step based on stability criterion. During walking, posture is real-time regulated according to the information of ZMP to realize the stability. In the end, the walking process was simulated and experiments were developed to confirm this strategy