Title: Error-based ADRC approach of lower knee exoskeleton system for rehabilitation

Authors: Nasir Ahmed Alawad; Amjad Jaleel Humaidi; Ahmed Sabah Al-Araji

Addresses: Faculty of Engineering, Department of Computer Engineering, Mustansiriyah University, BP, 10052, Iraq ' Department of Control and System Engineering, University of Technology, BP, 10066, Iraq ' Department of Computer Engineering, University of Technology, BP, 10066, Iraq

Abstract: In this study, active disturbance rejection control (ADRC) has been designed to control the exoskeleton system for rehabilitation at knee level and to replace the exercises made by physicians with systematic training devices. The time derivative of reference input and feed-back signals is an evitable in most ADRC schemes. To alleviate the burden due to derivative actions, the idea of proposing error-based ADRC (EADRC) has been introduced. In a conventional ADRC scheme, the extended state observer (ESO) is the core element of the controller to estimate both the states of the system and the exerted disturbance. The EADRC utilises the estimates in the error sense rather than the actual states. The EADRC technique is compared to traditional ADRC and the numerical results showed that the proposed EADRC outperforms the conventional ADRC in terms of tracking errors, noise and load rejection capabilities for the system subjected to noise and load uncertainties.

Keywords: exoskeleton system; ADRC; active disturbance rejection control; robustness; stability; knee rehabilitation.

DOI: 10.1504/IJAAC.2025.142972

International Journal of Automation and Control, 2025 Vol.19 No.1, pp.1 - 17

Received: 09 Aug 2023
Accepted: 16 Dec 2023
Published online: 02 Dec 2024 *