Adaptive dynamic feedback control of parallel robots with unknown kinematic and dynamic properties. (July 2022)

• Record Type: Journal Article
• Authors: Harandi, M. Reza J.; Khalilpour, S. Ahmad; Taghirad, Hamid D.
• Is Part Of: ISA transactions. Volume 126(2022)
• Page Start: 574
• Page End: 584
• Publication Date: 2022-07
• Subjects: Adaptive control -- Parallel robot -- Unknown kinematic and dynamic -- Motion control -- Dynamic feedback; Engineering instruments -- Periodicals; Engineering instruments; Periodicals; Electronic journals; 629.805
• DOI: 10.1016/j.isatra.2021.08.026
• Languages: English
• ISSNs: 0019-0578

Abstract

Abstract: Uncertainties in dynamic and kinematic parameters are unavoidable components in the control of robotic manipulators. Although calibration is a well-known method to reject this issue, it is time-consuming, some parameters may be altered slowly, and therefore, it is not applicable to some special cases such as deployable cable-driven robots. This paper addresses an adaptive dynamic feedback controller in which the adaptation laws together with new states could remedy these shortcomings and may be appropriately used in deployable cable-driven robots. For this purpose, the Jacobian matrix and its determinant are expressed in regressor form. Additionally, a non-singular sliding surface is considered for the trajectory tracking error. The fast finite-time feasible trajectory tracking is ensured by Lyapunov direct method using an appropriate design of adaptation laws of unknown parameters together with dynamical matrices in the presence of external disturbance. A 4RP R (revolute–prismatic–revolute) redundant rigid body and a fully actuated 3-DOF cable-driven robot are considered to verify the proposed method and also compare the results with state-of-art by simulation and experiment. Highlights: Adaptive terminal controller for uncertain parallel robots is designed. Asymptotic stability is ensured by the Lyapunov method. The effect of input constraints in redundant robots is investigated. The results are verified through experiment on a cable driven parallel robot.