A Walking-Assistant Robot Controlled by Tactile-Slip Sensation

摘要

A walking-assistant robot guided by the intention and power-driven is presented, its purpose is to provide physical support and walking assistance for the elderly to meet their needs of walking autonomy, friendliness, and maintaining the ability of walking and taking care of themselves. Tactile and slip sensor is selected as the human interface to perceive the users walking intent, and the sensor is also used to detect the user's slip trend. And the paper researches the feature representation and extraction method of tactile and slip signal for driving control pattern recognition. An improved classification and identification method combining K-means in clustering and K-nearest neighbor algorithm in classification is proposed. The paper introduces the overall design schemes of tactile and slip drive control system of walking-assistant robot, perception system, motion control system. Finally the feasibility and effectiveness of the entire system are verified by experiment.