Sensor Data Fusion for Body State Estimation in a Bipedal Robot and Its Feedback Control Application for Stable Walking

Sensors

Ching-Pei Chen¹  Jing-Yi Chen¹  Chun-Kai Huang¹  Jau-Ching Lu¹  Pei-Chun Lin² 

[1] Department of Mechanical Engineering, National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei 10617, Taiwan;

关键词: body state estimator;    sensor fusion;    extended Kalman filter;    bipedal robot;    ZMP;    preview control;   

DOI  :  10.3390/s150304925

来源: mdpi

【 摘 要 】

We report on a sensor data fusion algorithm via an extended Kalman filter for estimating the spatial motion of a bipedal robot. Through fusing the sensory information from joint encoders, a 6-axis inertial measurement unit and a 2-axis inclinometer, the robot’s body state at a specific fixed position can be yielded. This position is also equal to the CoM when the robot is in the standing posture suggested by the detailed CAD model of the robot. In addition, this body state is further utilized to provide sensory information for feedback control on a bipedal robot with walking gait. The overall control strategy includes the proposed body state estimator as well as the damping controller, which regulates the body position state of the robot in real-time based on instant and historical position tracking errors. Moreover, a posture corrector for reducing unwanted torque during motion is addressed. The body state estimator and the feedback control structure are implemented in a child-size bipedal robot and the performance is experimentally evaluated.

【 授权许可】

CC BY   
© 2015 by the authors; licensee MDPI, Basel, Switzerland.