Modeling and tuning of a subject-loaded mobile gait rehabilitation system

Patients with neurological disorders, such as stroke survivors, can be treated with physical rehabilitation to regain motor control and function. Conventional therapy techniques are labor intensive and non-standardized. This is especially true in gait rehabilitation. A natural and tunable rehabilitation gait system (NaTUre_gaits) has been developed to assist gait rehabilitation with body-weight support locomotion training (BWSLT), gait control (GC), balance control (BC) and pelvic control (PC). All of those components are attached to a mobile platform (MP) with a total weight of more than 200kg. In order to coordinate the system with the motion of GC/BC/PC and to provide BWS for the subject throughout the walking test, the mobile system needs to move adaptively and steadily against those dynamic disturbance and heavy load. Motivated by reducing the tuning time, an identification and control method are proposed. Based on the relay response during operation, the system model is obtained and then PD controller with feed forward feedback is designed. Trials with/without subject are performed to demonstrate the effectiveness and merit of the proposed modeling and tuning method.