Homenko M. Optimization of control processes reversible converter with embedded neural controller

Українська версія

Thesis for the degree of Candidate of Sciences (CSc)

State registration number

0411U003460

Thesis Registration Form

0411U003460.pdf

Applicant for

Specialization

  • 05.09.12 - Напівпровідникові перетворювачі електроенергії

12-05-2011

Specialized Academic Board

Д.26.002.19

Essay

The dissertation is devoted to the development of theoretical and practical principles of reversible DC drive control based on artificial neural networks. The application of artificial neural network together with state space methods were justified for finite transient duration without overshoot obtaining. The method of training set forming, the algorithm of neural controller structure choice were developed and digital controller with different duration of quantization steps were proposed as a reference for neural network learning, that allowed us to obtain optimum performance for the transition process for any angle task.

Read more

Files

aref.pdf

dis.pdf

Similar theses

0524U000012

Andrii V. Obrubov

Methods and models of research of electromagnetic processes of resonance power converters

0521U101290

Yaskiv Volodymyr

High-efficiency semiconductor power converters based on high-frequency magnetic amplifiers

0421U101796

Marunia Yuliia V

The power hybrid filters for decentralized power supply systems.

0521U100895

Ivakhno Volodymyr

Non-reversible two-stage DC/DC converters with separated commutation

0521U100829

Verbytskyi Ievgen Volodymyrovych

Spectral analysis of processes and synthesis of semiconductor converters control laws on the basis of a Fourier series of several variables