Assessment of Wearable Lower Limb Rehabilitation Robot Dynamics: A Study on ROM Training Performance

Abstract

A wearable lower limb rehabilitation robot (WLLR) is one of the effective ways to assist a stroke patient who has abnormal gait. The accuracy of the dynamic model is important as it relates to the performances of the robot to track the trajectory. Error in modelling can be magnified futher resulting to poor tracking performance. However, the development of the dynamic model for WLLR is challenging because the structure is highly non-linear and heavily coupled. In this paper, a mathematical modelling for an improved design of wearable lower limb rehabilitation robot (WLLR) is presented. The Lagrangian formulation was utilized to derive the dynamic model of the hip and knee joint. Simple PID (Ziegler-Nichols) based controller was developed in order to verify the developed dynamic model. MATLAB SimMechanic software was used to simulate the WLLR motion behaviour to imitate a real environment. The results demonstrate the successful tracking of desired ranges of motion (ROM) by WLLR joints with low rise time and steady-state error. Thus, the developed dynamic model is acceptable and can be utilized for future improvement of the controller systems for WLRR.