The stepping over an obstacle for the humanoid robot with the consideration of dynamic balance

摘要

This paper presented a fuzzy decentralized controller for a small size humanoid robot (HR) to stepping over the known obstacle and keeping balance when the force disturbance occurs. The proposed HR possessed 55 cm height, 3.7 kg weight, and 21 degree-of-freedom (DOF). This paper contains motion control, posture capture, and dynamic balance control. The first part designed a standard motion of stepping over the obstacle by human-machine interface. The second part set a posture captured sensors combining with two gyro sensors and one accelerometer. Based on SOPC (System on a Programmable Chip), it operated with parallel processing. This suggested sensors captured the responses of the pitch axis and roll axis during the stepping over the obstacle. On the other hand, the Kalman filter was applied to improve the measure accuracy. The fuzzy decentralized control (FDC) to keep the dynamic balance was then considered as the third part. During the steeping over the obstacle, the HR was punched by a standard hammer to verify the effectiveness of the proposed FDC. Finally, the corresponding experiments of steeping over obstacle with punched disturbance confirmed the suggested result.