Genetik algoritma tabanlı PID kontrolör simülatörü tasarımı
Files
Date
2013-03-13
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Uludağ Üniversitesi
Abstract
Sistem denetimi alanında PID (oransal-integral-türevsel) kontrolörler önemli yer tutmaktadır. Sistemin zaman veya frekans bölgesi verilerinden faydalanılarak; sistem özelliklerine uygun, belirli ölçütler altında, bu tür kontrolörlerin tasarımı için Ziegler-Nichols, Cohen-Coon, Chien Hrones Reswick (CHR), Wang-Juang-Chan gibi değişik yöntemler mevcuttur. Bilgisayar ve yapay zekâ alanındaki gelişmelere paralel olarak genetik algoritmaların da kontrol sistemlerindeki uygulamaları artmaktadır. Bu çalışmada, PID kontrolörlerin tasarımları hem klasik yöntemlerle hem de genetik algoritmalarla gerçekleştirilerek karşılaştırmalı sonuçlar analiz edilmiştir. Bu doğrultuda, eğitim amaçlı da kullanılabilecek grafiksel arayüz programı tasarlanmıştır. Program ile transfer fonksiyonları girilen sistemler için uygun P, PI ve PID kontrolör katsayıları hem seçilen klasik yöntemlerle hem de genetik algoritmalarla hesaplanmakta; sisteme ait birçok parametre ve cevap eğrileri hem sayısal hem de grafiksel olarak karşılaştırmalı sunulabilmektedir.
PID (Proportional Integral and Derivative) controllers take an important place in the field of system controlling. Various methods such as Ziegler-Nichols, Cohen-Coon, Chien Hrones Reswick (CHR) and Wang-Juang-Chan are available for the design of such controllers benefiting from the system time and frequency domain data. These controllers are in compliance with system properties under certain criteria suitable to the system. Genetic algorithms have become widely used in control system applications in parallel to the advances in the field of computer and artificial intelligence. In this study, PID controller designs have been carried out by means of classical methods and genetic algorithms and comparative results have been analyzed. For this purpose, a graphical user interface program which can be used for educational purpose has been developed. For the definite (entered) transfer functions, the suitable P, PI and PID controller coefficients have calculated by both classical methods and genetic algorithms and many parameters and responses of the systems have been compared and presented numerically and graphically.
PID (Proportional Integral and Derivative) controllers take an important place in the field of system controlling. Various methods such as Ziegler-Nichols, Cohen-Coon, Chien Hrones Reswick (CHR) and Wang-Juang-Chan are available for the design of such controllers benefiting from the system time and frequency domain data. These controllers are in compliance with system properties under certain criteria suitable to the system. Genetic algorithms have become widely used in control system applications in parallel to the advances in the field of computer and artificial intelligence. In this study, PID controller designs have been carried out by means of classical methods and genetic algorithms and comparative results have been analyzed. For this purpose, a graphical user interface program which can be used for educational purpose has been developed. For the definite (entered) transfer functions, the suitable P, PI and PID controller coefficients have calculated by both classical methods and genetic algorithms and many parameters and responses of the systems have been compared and presented numerically and graphically.
Description
Keywords
PID, Genetik algoritma, Ziegler-Nichols yöntemi, Cohen-Coon yöntemi, Chien Hrones Reswick yöntemi, Wang-Juang-Chan yöntemi, Genetic algorithm, Ziegler-Nichols method, Cohen-Coon method, Chien Hrones Reswick method, Wang-Juang-Chan method
Citation
Vatansever, F. ve Şen, D. (2013). "Genetik algoritma tabanlı PID kontrolör simülatörü tasarımı". Uludağ Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi, 18(2), 7-18.