Neural Network-Based Identification and Adaptive Controller Optimized by Quadratic cost function

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26220%2F10%3APU88119" target="_blank" >RIV/00216305:26220/10:PU88119 - isvavai.cz</a>
Result on the web
—
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
Neural Network-Based Identification and Adaptive Controller Optimized by Quadratic cost function
Original language description
The algorithm for adaptive control has been designed with on-line identification based on a neural network and Lavenberq-Marquardt learning algorithm. For identification the three layer neural network with logistic activation functions in hidden layer and linear network were used. Synthesis of the linear adaptive LQ controller utilized by plant transfer function obtained from the learned neural network where 2th order linear approximation is getting in its operation point and was used for setting of thecontroller's parameters. Disturbance cancelling is realized using summing of control error. All algorithms were implemented into B&R PLC and presented results were achieved in electronic model regulating.
Czech name
—
Czech description
—

Project
<a href="/en/project/GA102%2F09%2F1680" target="_blank" >GA102/09/1680: Control Algorithm Design by Means of Evolutionary Approach</a><br>
Continuities
Z - Vyzkumny zamer (s odkazem do CEZ)

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

Similar results(10)