Real-time fall and overturn prevention control for human-cane robotic system

A Intelligent Cane Robot(iCane) has been proposed to assist the elderly walking in daily life. As a nursing-care robot, the safety and dependability are the most important issues that should be investigated. For preventing the human subject from falling, the angle of human body and the acceleration of center of gravity(COG) should be less than some threshold. Although in a human-in-the-loop system, the human subject is regarded a uncontrollable object. However, while the user is falling over, the cane robot can move to a appropriately position and support the user for balance. As the prerequisite condition that the cane robot support the human balance, the stability of cane robot should be ensured firstly. According Newton-Euler Law, a dynamic model is proposed to present the stability of human-cane robotic system. A impedance control are used to achieve position, posture and force control of iCane for fall prevention. The simulation and experimental results show the performance of fall prevention by using iCane.