Predefined-time control for 3-DOF parallel robotic manipulators using an orthogonal feedforward neural network

Citations

SCOPUS

0

초록

Parallel robotic manipulators (PRM) are increasingly employed to replace humans in repetitive, classification, and complex tasks across industrial, domestic, and advanced manufacturing applications such as 3D printing and CNC machining, where high accuracy is required. To address the control challenges, a novel predefined-time stability (PTS) theorem with a continuous piecewise formulation is introduced, guaranteeing predefined-time convergence of system states without singularities, while ensuring robustness against uncertainties and reducing chattering. This paper develops a predefined-time control scheme for PRMs, enhanced by an orthogonal feedforward neural network (OFFNN). A key feature of the proposed design is that the tracking errors are guaranteed to converge to the sliding mode (SM) surface within a user-defined predefined time, independent of the initial conditions. The OFFNN is employed to approximate unknown nonlinear dynamics, enabling rapid and precise compensation of lumped uncertainties while mitigating chattering, preserving the robustness required for predefined-time convergence, and improving overall tracking performance. By exploiting the PTS theorem, the control method ensures guaranteed convergence within the predefined time, strong robustness against uncertainties, reduced chattering, and high tracking precision. The closed-loop stability is rigorously established using Lyapunov theory. Finally, numerical simulations and performance comparisons are presented to validate the PTS and the effectiveness of the proposed controller for a 3-DOF PRM. © 2026 Elsevier Inc.

키워드

Lyapunov theory; Orthogonal feedforward neural network; Parallel robotic manipulators; Predefined-time stability; Sliding mode control
제목
Predefined-time control for 3-DOF parallel robotic manipulators using an orthogonal feedforward neural network
저자
Vo, Anh Tuan; Truong, Thanh Nguyen; Nguyen, Ngoc Hoai An; Hong, Ic-Pyo
DOI
10.1016/j.ins.2026.123639
발행일
2026-10
유형
Article
저널명
Information Sciences
권
753
페이지
1 ~ 26