Compliant Motion Control of a Walk-Assistant Robot Based on Force and Gait Information

• Main Authors: Wu, Shang-Yang, 吳尚陽
• Other Authors: Song, Kai-Tai
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/x3wh9b

碩士 === 國立交通大學 === 電控工程研究所 === 103 === This thesis presents a new method to provide compliant motion of a walk assistant robot based on user status and motion intent. A laser scanner is used to detect the user's gait information and a force/torque sensor detects desired motion direction. The compliance controls will provide the walking assistance. The robot velocity is regulated by user gait information and the robot motion direction is adjusted by force/torque sensor information. These two compliant controllers allow the walking assist robot to generate passive behavior and provide walking assistance effectively. This thesis develops a method to combine autonomous obstacle avoidance and compliant motion by using shared-control approach. With this method, a user can walk in a complex environment and prevent from collision with obstacles. Experimental verification of the proposed design show that a user can walk with the robot to reach the destination. Experimental results show that a combination of laser scanner and force/torque sensor can make the robot to comply with abnormal gait such as festinating gait, lame and other gait. Furthermore, autonomous obstacle avoidance allows a user to move around safely even without noticing obstacles.