Incremental social learning-based differential evolution algorithms for large-scale convex and nonconvex economic power dispatch problems

Serdar Özyön

Makale detayları

Incremental social learning-based differential evolution algorithms for large-scale convex and nonconvex economic power dispatch problems

Yazarlar:

Serdar Özyön

Yayıncı:

Nigde Omer Halisdemir University

Ana Başlıklar:

Elektrik Tesisleri

Dergi:

Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 3, pp.907-931

Belge Tipi:

Makale

Yayın Tarihi:

7/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001354707

pISSN:

2564-6605

Anahtar kelimeler:

Özet:

 The economic power dispatch problem, which is one of the important problems of power systems engineering, is a problem that must be solved quickly for the optimum operation of the existing system. This problem is defined in the literature as the ability to meet the loads in the system at the least cost by all power generation units under different constraints. In many different studies, solutions to the problem have been sought for many different power systems. Since power systems differ from each other in terms of structure and constraints, the problem remains current. This study solves the economic power dispatch problem for both convex and non-convex fuel cost functions for a large-scale power system with modified IEEE 118 bus 54 generators. Transmission line losses were considered in the sample power system, and AC load flow analysis was performed using the Newton-Raphson method to calculate these losses. In both problems with different structures, the incremental social learning structure was integrated into the differential evolution algorithm in five different ways to calculate the minimum fuel cost values, and new incremental social learning-based differential evolution algorithms were developed and applied. By comparing the numerical values obtained in solutions made with six different algorithms, the most appropriate method specific to the problems was determined. All results are evaluated using tables, graphs, and figures. In this study, transmission line losses are realistically taken into account with Newton-Raphson based AC load flow and incremental social learning structures are integrated into the classical differential evolution algorithm to improve solution quality and algorithm stability.

 The economic power dispatch problem, which is one of the important problems of power systems engineering, is a problem that must be solved quickly for the optimum operation of the existing system. This problem is defined in the literature as the ability to meet the loads in the system at the least cost by all power generation units under different constraints. In many different studies, solutions to the problem have been sought for many different power systems. Since power systems differ from each other in terms of structure and constraints, the problem remains current. This study solves the economic power dispatch problem for both convex and non-convex fuel cost functions for a large-scale power system with modified IEEE 118 bus 54 generators. Transmission line losses were considered in the sample power system, and AC load flow analysis was performed using the Newton-Raphson method to calculate these losses. In both problems with different structures, the incremental social learning structure was integrated into the differential evolution algorithm in five different ways to calculate the minimum fuel cost values, and new incremental social learning-based differential evolution algorithms were developed and applied. By comparing the numerical values obtained in solutions made with six different algorithms, the most appropriate method specific to the problems was determined. All results are evaluated using tables, graphs, and figures. In this study, transmission line losses are realistically taken into account with Newton-Raphson based AC load flow and incremental social learning structures are integrated into the classical differential evolution algorithm to improve solution quality and algorithm stability.

 The economic power dispatch problem, which is one of the important problems of power systems engineering, is a problem that must be solved quickly for the optimum operation of the existing system. This problem is defined in the literature as the ability to meet the loads in the system at the least cost by all power generation units under different constraints. In many different studies, solutions to the problem have been sought for many different power systems. Since power systems differ from each other in terms of structure and constraints, the problem remains current. This study solves the economic power dispatch problem for both convex and non-convex fuel cost functions for a large-scale power system with modified IEEE 118 bus 54 generators. Transmission line losses were considered in the sample power system, and AC load flow analysis was performed using the Newton-Raphson method to calculate these losses. In both problems with different structures, the incremental social learning structure was integrated into the differential evolution algorithm in five different ways to calculate the minimum fuel cost values, and new incremental social learning-based differential evolution algorithms were developed and applied. By comparing the numerical values obtained in solutions made with six different algorithms, the most appropriate method specific to the problems was determined. All results are evaluated using tables, graphs, and figures. In this study, transmission line losses are realistically taken into account with Newton-Raphson based AC load flow and incremental social learning structures are integrated into the classical differential evolution algorithm to improve solution quality and algorithm stability.

 Güç sistemleri mühendisliğinin önemli problemlerinden biri olan, ekonomik güç dağıtımı problemi, mevcut sistemin optimum işletimi için hızlı bir şekilde çözülmesi gereken bir problemdir. Bu problem literatürde, sistemdeki yüklerin farklı kısıtlar altında bütün güç üretim birimleri tarafından en az maliyetle karşılanması şeklinde tanımlanır. Farklı birçok çalışmada, farklı birçok güç sistemi için probleme çözüm aranmıştır. Güç sistemleri yapı ve kısıtlar bakımından birbirlerinden farklı olduklarından problem güncelliğini korumaktadır. Bu çalışmada, modifiye edilmiş IEEE 118 bara 54 generatörlü büyük ölçekli bir güç sistemi için hem konveks yakıt maliyeti fonksiyonları için hem de konveks olmayan yakıt maliyeti fonksiyonları için probleminin çözümü yapılmıştır. Örnek güç sisteminde iletim hattı kayıpları dikkate alınmış ve bu kayıpların hesaplanabilmesi için Newton-Raphson metodu kullanılarak AC yük akışı analizi yapılmıştır. Farklı yapıdaki her iki probleminde, minimum yakıt maliyeti değerlerinin hesaplanabilmesi için diferansiyel gelişim algoritmasına artırımlı sosyal öğrenme yapısı beş farklı şekilde entegre edilerek yeni artırımlı sosyal öğrenme tabanlı diferansiyel gelişim algoritmaları geliştirilmiş ve uygulanmıştır. Altı farklı algoritma ile yapılan çözümlerde elde edilen sayısal değerler karşılaştırılarak, problemlere özgü en uygun metot belirlenmeye çalışılmıştır. Bütün sonuçlar farklı tablo, grafik ve şekillerle verilmiş ve değerlendirilmiştir. Çalışmada iletim hattı kayıpları Newton-Raphson temelli AC yük akışı ile gerçekçi şekilde hesaba katılmış, ayrıca klasik diferansiyel gelişim algoritmasına artımlı sosyal öğrenme yapıları entegre edilerek çözüm kalitesi ve algoritma kararlılığı artırılarak literatüre katkı sağlanmıştır.

Makale detayları

Incremental social learning-based differential evolution algorithms for large-scale convex and nonconvex economic power dispatch problems

Yazarlar:

Serdar Özyön

Yayıncı:

Nigde Omer Halisdemir University

Ana Başlıklar:

Elektrik Tesisleri

Dergi:

Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 3, pp.907-931

Belge Tipi:

Makale

Yayın Tarihi:

7/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001354707

pISSN:

2564-6605

AC load flow analysis

convex and non-convex economic power dispatch

large-scale power systems

incremental social learning

differential evolution algorithmAC yük akış analizi

konveks ve konveks olmayan ekonomik güç dağıtımı

büyük ölçekli güç sistemleri

artırımlı sosyal öğrenme

diferansiyel gelişim algoritması

Sayı

Makale detayları

Yazarlar: Serdar Özyön

Yayıncı: Nigde Omer Halisdemir University

Ana Başlıklar: Elektrik Tesisleri

Dergi: Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 3, pp.907-931

Belge Tipi: Makale

Yayın Tarihi: 7/2025

Belge Dili: Türkçe

Document ID (Org No.): 2001354707

pISSN: 2564-6605

Sayı

Yazarlar:

Serdar Özyön

Yayıncı:

Nigde Omer Halisdemir University

Ana Başlıklar:

Elektrik Tesisleri

Dergi:

Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 3, pp.907-931

Belge Tipi:

Makale

Yayın Tarihi:

7/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001354707

pISSN:

2564-6605