Identification of nonlinear dynamical systems using multilayer perceptrons
Doğrusal olamayan devingen düzgelerin çok tabakalı algılayıcılar kullanılarak tanıyımı
- Tez No: 23508
- Danışmanlar: PROF. DR. MÜBECCEL DEMİREKLER
- Tez Türü: Yüksek Lisans
- Konular: Elektrik ve Elektronik Mühendisliği, Electrical and Electronics Engineering
- Anahtar Kelimeler: Neural networks, multUayer perceptrons, system identification, back propagation, learning rate, induction generator, Neural networks, multUayer perceptrons, system identification, back propagation, learning rate, induction generator
- Yıl: 1992
- Dil: İngilizce
- Üniversite: Orta Doğu Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 128
Özet
ABSTRACT IDENTIFICATION OF NONLINEAR DYNAMICAL SYSTEMS USING MULTILAYER PERCEPTRONS VOYVODAOĞLU, Tansel M.S. 1n Electrical and Electronics Engineerlng Supervisor: Prof.Dr. Mübeccel Demirekler September, 1992, 115 pages in thls study, the dynamical systems represented by nonlinear dlfference eguations are identlfied by uslng neural networks. Back propagation tralning algorithm 1s used to train multilayer perceptrons for the representatlon of nonlinear functions. W1thin this scope, neural networks and back propagation trainlng algorithm are critically rev1ewed especially from system 1dentificat1on point of view. Furthermore, computer simulatlon tests are carrled out by using the proposed 1dent1fication models to demonstrate the important propertles of the algorithm. Apart from those, different versions of the back propagation training algorithm are used in identification of the slmpUfied model of a wind türbine driven self-excited induction generator. The slmulation results reveal that multilayer perceptrons trained by back propagation method can be used effectively in 11irepresentation of nonlinear functlons in finite intervals. However, it 1s also observed that back propagatlon tralning algorithm has a very s1ow convergence rate and does not a!ways guarantee the global minimum of the performance measure.
Özet (Çeviri)
ABSTRACT IDENTIFICATION OF NONLINEAR DYNAMICAL SYSTEMS USING MULTILAYER PERCEPTRONS VOYVODAOĞLU, Tansel M.S. 1n Electrical and Electronics Engineerlng Supervisor: Prof.Dr. Mübeccel Demirekler September, 1992, 115 pages in thls study, the dynamical systems represented by nonlinear dlfference eguations are identlfied by uslng neural networks. Back propagation tralning algorithm 1s used to train multilayer perceptrons for the representatlon of nonlinear functions. W1thin this scope, neural networks and back propagation trainlng algorithm are critically rev1ewed especially from system 1dentificat1on point of view. Furthermore, computer simulatlon tests are carrled out by using the proposed 1dent1fication models to demonstrate the important propertles of the algorithm. Apart from those, different versions of the back propagation training algorithm are used in identification of the slmpUfied model of a wind türbine driven self-excited induction generator. The slmulation results reveal that multilayer perceptrons trained by back propagation method can be used effectively in 11irepresentation of nonlinear functlons in finite intervals. However, it 1s also observed that back propagatlon tralning algorithm has a very s1ow convergence rate and does not a!ways guarantee the global minimum of the performance measure.
Benzer Tezler
- Dinamik algılayıcı öğrenme algoritması ile kenar saptamanın öğrenilmesi
Learning of edge detection using recurrent perceptron learning algorithm
FİLİZ YOSMA TAŞKIN
Yüksek Lisans
Türkçe
1995
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiDOÇ.DR. CÜNEYT GÜZELİŞ
- Destek vektör regresyonu ile PID kontrolör tasarımı
Design of PID controller via support vector regression
KEMAL UÇAK
Yüksek Lisans
Türkçe
2012
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiKontrol ve Otomasyon Mühendisliği Ana Bilim Dalı
YRD. DOÇ. DR. GÜLAY ÖKE
- State of charge estimation of lithium-ion batteries using machine learning approach
Makine öğrenmesi yaklaşımı kullanılarak lityum iyon pillerin şarj durumu tahmini
OSMAN ALPER ALTUN
Yüksek Lisans
İngilizce
2024
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektrik Mühendisliği Ana Bilim Dalı
PROF. DR. EMİNE AYAZ
- Asenkron makina kontrolü için yapay sinir ağı tabanlı rotor akışı gözlemcisi
Başlık çevirisi yok
ASLI AYLA ÇAKIRGÖZ
Yüksek Lisans
Türkçe
1998
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektrik Mühendisliği Ana Bilim Dalı
PROF. DR. EMİN TACER
- Yapay sinir ağları ve uygulamaları
Başlık çevirisi yok
NURCİHAN ÜNALDI
Yüksek Lisans
Türkçe
1993
Elektrik ve Elektronik MühendisliğiHacettepe ÜniversitesiElektrik-Elektronik Mühendisliği Ana Bilim Dalı
DOÇ. DR. TURHAN ÇİFTÇİBAŞI