Publication detail

Advantages of Identification Based on Neural Network Approach

Václav Veleba

Original Title

Advantages of Identification Based on Neural Network Approach

Type

conference paper

Language

English

Original Abstract

The parameters estimation of the dynamic plant with great ratio of its time constant to sampling periods is considered. It is shown, that a neural network applied to on-line identification process produces more stable solution in the rapid sampling domain. Considering this results, we apply neural network as on-line estimator in an adaptive controller. Simple heuristic synthesis based on modified Z-N 1 method are discussed, that gives numerically stable parameters of the PID discrete controller.

Key words in English

Rapid Sampling, Quantization, Neural Network, Training Set, Levenberg-Marquardt Minimization, Discrete PID Controller, RLS Identification Method

Authors

Václav Veleba

RIV year

2005

Released

1. 1. 2005

Publisher

Ing. Zdeněk Novotný Csc., Ondráčkova 105, Brno

Pages from

76

Pages to

80

Pages count

5

BibTex

@inproceedings{BUT14547,
  author="Václav {Veleba}",
  title="Advantages of Identification Based on Neural Network Approach",
  booktitle="Advantages of Identification Based on Neural Network Approach",
  year="2005",
  number="1",
  pages="5",
  publisher="Ing. Zdeněk Novotný Csc., Ondráčkova 105, Brno"
}