Passive and active rehabilitation control of human upper-limb exoskeleton robot with dynamic uncertainties. Issue 11 (8th August 2018)

• Record Type: Journal Article
• Authors: Brahmi, Brahim; Saad, Maarouf; Luna, Cristobal Ochoa; Archambault, Philippe S.; Rahman, Mohammad H.
• Is Part Of: Robotica. Volume 36:Issue 11(2018)
• Page Start: 1757
• Page End: 1779
• Publication Date: 2018-08-08
• Subjects: Backstepping integral control, -- Time delay estimation, -- Passive and active rehabilitation, -- Time delay error; Robots -- Periodicals; 629.89205
• DOI: 10.1017/S0263574718000723
• Languages: English
• ISSNs: 0263-5747

Abstract

SUMMARY: This paper investigates the passive and active control strategies to provide a physical assistance and rehabilitation by a 7-DOF exoskeleton robot with nonlinear uncertain dynamics and unknown bounded external disturbances due to the robot user's physiological characteristics. An Integral backstepping controller incorporated with Time Delay Estimation (BITDE) is used, which permits the exoskeleton robot to achieve the desired performance of working under the mentioned uncertainties constraints. Time Delay Estimation (TDE) is employed to estimate the nonlinear uncertain dynamics of the robot and the unknown disturbances. To overcome the limitation of the time delay error inherent of the TDE approach, a recursive algorithm is used to further reduce its effect. The integral action is employed to decrease the impact of the unmodeled dynamics. Besides, the Damped Least Square method is introduced to estimate the desired movement intention of the subject with the objective to provide active rehabilitation. The controller scheme is to ensure that the robot system performs passive and active rehabilitation exercises with a high level of tracking accuracy and robustness, despite the unknown dynamics of the exoskeleton robot and the presence of unknown bounded disturbances. The design, stability, and convergence analysis are formulated and proven based on the Lyapunov–Krasovskii functional theory. Experimental results with healthy subjects, using a virtual environment, show the feasibility, and ease of implementation of the control scheme. Its robustness and flexibility to deal with parameter variations due to the unknown external disturbances are also shown.