Parameter Estimation of a Mathematical Model for Lower-Limb Exoskeletons with Paralysed Humans

Abstract

The development of dynamic models for rehabilitation lower-limb exoskeletons (RLLEs) integrated with human interaction (RLLE–Patient) has been proposed by numerous researchers. However, due to the nonlinear characteristics of these systems, accurate modelling remains a critical challenge, requiring approaches that are both simple and effective in representing actual scenarios. Therefore, this paper proposes a method for developing and controlling an RLLE-Patient model that accurately replicates real rehabilitation conditions. The RLLE and human models are initially designed in CATIA and transferred to ADAMS, while the direct current motors driving the RLLE joints are modelled in MATLAB Simulink, with motor parameters identified using a parameter estimation tool. A co-simulation between ADAMS and MATLAB is employed to create a comprehensive RLLE-Patient model. Additionally, a virtual Proportional-Integral-Derivative (PID) controller is developed to regulate the dynamic motion of the RLLE, and its performance is evaluated through simulations. The results demonstrate that this method effectively facilitates the design, tuning, evaluation, and improvement of the PID controller’s performance in RLLE applications. Further optimization of the control system can be achieved by integrating the dynamic characteristics into the RLLE-Patient model.