Nonlinear Systems Identification and Control Using Uncertain Rule-based Fuzzy Neural Systems with Stable Learning Mechanism

Ching Hung Lee*, Yi Han Lee, Chih Min Lin

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

This paper proposes an uncertain rule-based fuzzy neural system (UFNS-S) with stable learning mechanism for nonlinear systems identification and control. The proposed UFNS-S system not only preserves the ability of handling uncertain information but also performs less computational effort. The sinusoidal perturbations are adopted to combine with the fuzzy term sets of UFNS-S. For training the UFNS-S systems on system identification and control applications, the gradient descent method with adaptive learning rate is derived. This guarantees the convergence of UFNS-S by choosing adaptive learning rates which enhance the convergent speed. This provides a simple way for choosing the learning rates for training the UFNS-S which also guarantees convergence and faster learning. Finally, the effectiveness and performance of the proposed approach is illustrated by several examples, computational complexity analysis, nonlinear system identification, and tracking control of two-link robot manipulator system.

Original languageEnglish
Pages (from-to)470-488
Number of pages19
JournalInternational Journal of Fuzzy Systems
Volume19
Issue number2
DOIs
StatePublished - 1 Apr 2017

Keywords

  • Fuzzy neural system
  • Lyapunov theorem
  • Robot manipulator
  • Rule-based
  • System identification
  • Uncertainty

Fingerprint Dive into the research topics of 'Nonlinear Systems Identification and Control Using Uncertain Rule-based Fuzzy Neural Systems with Stable Learning Mechanism'. Together they form a unique fingerprint.

Cite this