Zeroing neural-dynamics approach and its robust and rapid solution for parallel robot manipulators against superposition of multiple disturbances

Dechao Chen, Yunong Zhang, Shuai Li

Research output: Journal article publicationJournal articleAcademic researchpeer-review

37 Citations (Scopus)

Abstract

This paper proposes a zeroing neural-dynamics (ZND) approach as well as its associated model for solving the real-time kinematic control problem of parallel robot manipulators. Unlike existing works relying on the plausibly impractical assumption that neural network models are free of external disturbances, the proposed model features the suppression of multiple disturbances in addition to its nonlinear processing and control. Theoretical analyses prove that the ZND approach and its associated model inherently possess robustness. In addition, by using an appropriate activation function, the rapid convergence performance of the corresponding ZND model is further achieved. Simulation studies and comprehensive comparisons substantiate the effectiveness, robustness and superiority of the proposed ZND approach as well as its associated model for solving the real-time kinematic control problem of parallel robot manipulators against the superposition of multiple disturbances. Moreover, results of extensive tests verify that the processing of the ZND model can be accelerated by using an appropriate activation function.

Original languageEnglish
Pages (from-to)845-858
Number of pages14
JournalNeurocomputing
Volume275
DOIs
Publication statusPublished - 31 Jan 2018

Keywords

  • External disturbances
  • Neural network models
  • Parallel robot manipulators
  • Real-time kinematic control problem
  • Robust solution
  • Zeroing neural-dynamics approach

ASJC Scopus subject areas

  • Computer Science Applications
  • Cognitive Neuroscience
  • Artificial Intelligence

Cite this