Generalized predictive control based on self-recurrent wavelet neural network for stable path tracking of mobile robots: Adaptive learning rates approach

Sung Jin Yoo, Yoon Ho Choi, Jin Bae Park

Research output: Contribution to journalArticlepeer-review

128 Citations (Scopus)

Abstract

In this paper, a generalized predictive control (GPC) method based on self-recurrent wavelet neural network (SRWNN) is proposed for stable path tracking of mobile robots. Since the SRWNN has a self-recurrent mother wavelet layer, it can well attract the complex nonlinear system although the SRWNN has less mother wavelet nodes than the wavelet neural network. Thus, the SRWNN is used as a model identifier for approximating on-line the states of the mobile robot. In our control system, since the control inputs, as well as the parameters of the SRWNN identifier are trained by the gradient descent method with the adaptive learning rates (ALRs), the optimal learning rates which are suitable for the time-varying trajectory of the mobile robot can be found rapidly. The ALRs for training the parameters of the SRWNN identifier and those for learning the control inputs are derived from the discrete Lyapunov stability theorem, which are used to guarantee the convergence of the GPC system. Finally, simulation results are provided to demonstrate the effectiveness of the proposed control strategy.

Original languageEnglish
Pages (from-to)1381-1394
Number of pages14
JournalIEEE Transactions on Circuits and Systems I: Regular Papers
Volume53
Issue number6
DOIs
Publication statusPublished - 2006 Jun

Bibliographical note

Funding Information:
Recently, there has been considerable interest in the decay D + --'11' n + ~1 due to conflicting measurements of the branching ratio relative to the decay D~--. 0n + \[1 -4\]. A precise measurement of this decay rate is of interest since a large D~+ -,q'n + branching ratio poses problems for current models, which predict values of BR(D~ + ~'n +)/BR(D + ~0n +) of less than 2 \[5 ,6 \]. Among the mechanisms offered as possible explanations for larger values are final-state interactions and even the presence of a nearby scalar resonance \[5 ,7 \]. Several experiments have searched for evidence of the decay D~+ ~l\]'n +. An indication for a signal was reported by Mark II \[1 \], with a value of 4.8 + 2.1 for the ratio BR(D + --,Tl'n +)/BR(D + --.0n +). Subsequently, NA14' also claimed evidence for this channel, with a value of 5.0 + 1.8 + 1.2 for the same ratio \[3\]. However, in considerable disagreement with these results are the upper limits reported by Mark III and E691. These experiments have determined BR(D + ~l\]'n + )/BR(D + ~0n + ) to be less than 1.9 \[2\] and less than 1.7 \[4\] respectively, both at the t Supported by the German Bundesministerium f'tir Forschung und Technologie, under contract number 054DO51P. 2 Supported by the German Bundesministerium f'tir Forschung und Technologie, under contract number 054ER 12P. 3 Supported by the German Bundesministerium f'tir Forschung und Technologie, under contract number 054HD24P. 4 McGill University, Montreal, Quebec, Canada H3C 3J7. 5 University of Toronto, Toronto, Ontario, Canada M5S IA7. 6 Carleton University, Ottawa, Ontario, Canada K1S 5B6. 7 Supported by the Natural Sciences and Engineering Research Council, Canada. 8 Supported by the German Bundesministerium f'dr Forschung und Technologie, under contract number 054KA 17P. 9 Supported by Alexander yon Humboldt Stiftung, Bonn, FRG. 1o Supported by Raziskovalna skupnost Slovenije and the Inter-nationales Biiro KfA, Jiilich. 11 Supported by the Swedish Research Council. 12 Supported by the US Department of Energy, under contract DE-AS09-80ER 10690. at References in this paper to a specific charged state are to be interpreted as implying the charge-conjugate state also.

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Generalized predictive control based on self-recurrent wavelet neural network for stable path tracking of mobile robots: Adaptive learning rates approach'. Together they form a unique fingerprint.

Cite this