Modelling and Identification of a Servo-mechanism using a Neural Network

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F70883521%3A28140%2F10%3A63509037" target="_blank" >RIV/70883521:28140/10:63509037 - isvavai.cz</a>
Result on the web
—
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
Modelling and Identification of a Servo-mechanism using a Neural Network
Original language description
The subject of this contribution is the use of neural networks for modeling and identification of processes. In the paper, a feedforward neural network is used. One of the advantages is that the resultant model depends on the setting of neural network weights. In the contribution, the Backpropagation algorithm was employed for this setting. Network topology plays also an important role and depends on the solved problem. Purpose of this paper is to propose a suitable topology of a neural network for a particular process and to compare the neural network model with real laboratory equipment as well. The Amira DR300 system depicted in the figure below was chosen as the laboratory model. It is a servo-mechanism which consists of a permanently excited DC-motor and a tacho-generator with the incremental encoder as a sensor of output signal. Input-output data were measured on this model and were further used for the neural network learning. Consequently the neural network describes the system
Czech name
—
Czech description
—

Publication year
2010
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection
Proceedings of XXXVth. Seminary ASR10 Instruments and Control
ISBN
978-80-248-2191-7
ISSN
—
e-ISSN
—
Number of pages
4
Pages from-to
—
Publisher name
VŠB-Technická univerzita Ostrava
Place of publication
Ostrava
Event location
Ostrava
Event date
Jan 1, 2010
Type of event by nationality
EUR - Evropská akce
UT code for WoS article
—

Similar results(10)