Observer-Based Adaptive Fuzzy Event-Triggered Control for Mechatronic Systems With Inaccurate Signal Transmission and Motion Constraints

• Published in: IEEE/ASME transactions on mechatronics Vol. 27; no. 6; pp. 5208 - 5221
• Main Authors: Yang, Tong, Zhang, Xinlin, Fang, Yongchun, Sun, Ning, Iwasaki, Makoto
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.12.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptation models; Adaptive control; Artificial muscles; Canonical forms; Control systems design; Controllers; Error signals; Event triggered control; Feedback control; Fuzzy control; Hysteresis; Mathematical models; Mechatronics; motion control; Observers; Output feedback; Parameters; Robot arms; Signal transmission; Stability analysis; State feedback; Sun; Tracking errors; Upper bound; Vehicle dynamics; vibration control
• ISSN: 1083-4435; 1941-014X
• DOI: 10.1109/TMECH.2022.3175969

This article presents an adaptive output feedback controller for a class of Euler-Lagrange (EL) mechatronic systems subject to time-delay, input hysteresis, and motion constraints; simultaneously, the event-triggered mechanism is introduced to decrease communication costs. During controller design, the proposed method not only relaxes additional requirements on the model structures by other methods (e.g., parameters linearly appearing in dynamics or strict triangular/cascade normal forms), but also estimates uncertain dynamics and approximation errors online, instead of utilizing "large-gain" or discontinuous terms to suppress their impacts. Regarding the inaccurate signal transmission of EL systems, this article presents a modified fuzzy observer with an elaborately designed auxiliary term, to recover unmeasurable variables and simultaneously deal with state time-delay. Meanwhile, the Prandtl-Ishlinskii model is employed to imitate input hysteresis, whose unknown parameters are also estimated online to improve tracking accuracy. By Lyapunov-based stability analysis, it is proven that the error signals are always limited within preset constraints and asymptotically converge to zero. As far as we know, for EL systems, this article proposes the first controller to address such comprehensive effects of unreliable state feedback, limited workspace, and input/output nonlinearities in practice, and also eliminate tracking errors with a theoretical guarantee . More importantly, the utilized event-triggered mechanism further improves the practicability of this article. Finally, based on a pneumatic artificial muscle (PAM)-actuated robot manipulator, the performance of the proposed controller is validated via hardware experiments.