Observer based output feedback repetitive learning control of robotic manipulators

Autor: Kadriye Merve Dogan, Enver Tatlicioglu, Erkan Zergeroglu, Kamil Cetin
Jazyk: angličtina
Rok vydání: 2024
Předmět:
Zdroj: IET Control Theory & Applications, Vol 18, Iss 15, Pp 1958-1967 (2024)
Druh dokumentu: article
ISSN: 1751-8652
1751-8644
DOI: 10.1049/cth2.12706
Popis: Abstract This work tackles the tracking control problem of robotic manipulators where the robot dynamics contains uncertain parameters and joint velocity measurements are not available. Specifically when the robotic manipulator is required to perform a periodic task repetitively, as in most industrial applications, a repetitive learning controller is proposed that does not require joint velocity measurements and can compensate the uncertainties of the robot dynamical parameters and additive disturbances caused due to the periodic joint motion. The proposed solution is achieved via the use of a novel learning component in the controller design in conjunction with a novel model‐free joint velocity observer design. The stability of the closed‐loop system and the convergence of both the joint position tracking error and the joint velocity observation error to the origin are guaranteed via Lyapunov based arguments. Experimental results performed on a 2 degree of freedom robot manipulator are presented to demonstrate the performance of the proposed observer–controller couple.
Databáze: Directory of Open Access Journals
Externí odkaz: