Optimizing and comparison of market chain product distribution problem with different genetic algorithm versions

Burak Gülmez

Makale detayları

Optimizing and comparison of market chain product distribution problem with different genetic algorithm versions

Yazarlar:

Burak Gülmez

Yayıncı:

Osmaniye Korkut Ata Üniversitesi

Ana Başlıklar:

Endüstri Mühendisliği

Dergi:

Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi Vol.6 Issue 1, pp.180-196

Belge Tipi:

Makale

Yayın Tarihi:

10/2023

Belge Dili:

Türkçe

Document ID (Org No.):

2001284864

pISSN:

2687-3729

Anahtar kelimeler:

Özet:

 In this study, it is aimed to carry out product distributions with the lowest distance and lowest cost for a market chain in Kayseri. Since there are too many market branches, the distribution of products in different orders affects the result quite a lot. This problem has been defined as the traveling salesman problem. When the traveling salesman problem is large, it cannot be solved by integer linear programming in polynomial time. It is therefore an NP-hard problem type. Therefore, a genetic algorithm was used for the solution. The genetic algorithm does not guarantee the optimum solution, but it can obtain acceptable solutions. It also obtains these solutions in a short time. The solution obtained is at an acceptable level, even if it is not optimal. In this study, a route was created with a genetic algorithm for 61 markets in Kayseri. In addition to the classical genetic algorithm, genetic algorithm variations that have emerged in recent years have been used As a result of all these algorithms, approximately 80 kilometers has been obtained to visit all the markets. When the solution obtained was examined, it was observed that it was a very good route.

 In this study, it is aimed to carry out product distributions with the lowest distance and lowest cost for a market chain in Kayseri. Since there are too many market branches, the distribution of products in different orders affects the result quite a lot. This problem has been defined as the traveling salesman problem. When the traveling salesman problem is large, it cannot be solved by integer linear programming in polynomial time. It is therefore an NP-hard problem type. Therefore, a genetic algorithm was used for the solution. The genetic algorithm does not guarantee the optimum solution, but it can obtain acceptable solutions. It also obtains these solutions in a short time. The solution obtained is at an acceptable level, even if it is not optimal. In this study, a route was created with a genetic algorithm for 61 markets in Kayseri. In addition to the classical genetic algorithm, genetic algorithm variations that have emerged in recent years have been used As a result of all these algorithms, approximately 80 kilometers has been obtained to visit all the markets. When the solution obtained was examined, it was observed that it was a very good route.

 In this study, it is aimed to carry out product distributions with the lowest distance and lowest cost for a market chain in Kayseri. Since there are too many market branches, the distribution of products in different orders affects the result quite a lot. This problem has been defined as the traveling salesman problem. When the traveling salesman problem is large, it cannot be solved by integer linear programming in polynomial time. It is therefore an NP-hard problem type. Therefore, a genetic algorithm was used for the solution. The genetic algorithm does not guarantee the optimum solution, but it can obtain acceptable solutions. It also obtains these solutions in a short time. The solution obtained is at an acceptable level, even if it is not optimal. In this study, a route was created with a genetic algorithm for 61 markets in Kayseri. In addition to the classical genetic algorithm, genetic algorithm variations that have emerged in recent years have been used As a result of all these algorithms, approximately 80 kilometers has been obtained to visit all the markets. When the solution obtained was examined, it was observed that it was a very good route.

 Bu çalışmada Kayseride bulunan bir market zinciri için ürün dağıtımlarının en düşük mesafe ve en düşük maliyet ile gerçekleştirilmesi hedeflenmiştir. Market şubeleri çok fazla olduğu için ürün dağıtımlarının farklı sıralamalar ile yapılması, sonucu oldukça etkilemektedir. Bu problem gezgin satıcı problemi şeklinde tanımlanmıştır. Gezgin satıcı problemi büyük boyutlu olduğunda polinom zaman içerisinde saf tam sayılı doğrusal programlama ile çözülememektedir. Bundan dolayı NP-zor bir problem türüdür. Bu yüzden çözüm için genetik algoritma kullanılmıştır. Genetik algoritma optimum çözümü garanti etmez fakat kabul edilebilir çözümler elde edebilir. Ayrıca bu çözümleri kısa bir zaman içerisinde elde eder. Elde edilen çözüm optimum olmasa bile kabul edilebilir seviyededir. Bu çalışmada Kayserideki 61 adet market için genetik algoritma ile bir rota oluşturulmuştur. Klasik genetik algoritmaya ilave olarak son yıllarda çıkan genetik algoritma varyasyonları kullanılmıştır. Tüm bu algoritmalar sonucunda bütün marketleri dolaşmak için yaklaşık 80 kilometrelik bir mesafe elde edilmiştir. Elde edilen çözüm incelendiğinde gayet iyi bir rota olduğu gözlemlenmiştir.

Makale detayları

Optimizing and comparison of market chain product distribution problem with different genetic algorithm versions

Yazarlar:

Burak Gülmez

Yayıncı:

Osmaniye Korkut Ata Üniversitesi

Ana Başlıklar:

Endüstri Mühendisliği

Dergi:

Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi Vol.6 Issue 1, pp.180-196

Belge Tipi:

Makale

Yayın Tarihi:

10/2023

Belge Dili:

Türkçe

Document ID (Org No.):

2001284864

pISSN:

2687-3729

Traveling salesman problem

Genetic algorithm

Product distribution

Gezgin satıcı problemi

Ürün dağıtımı

Genetik algoritma

Sayı

Makale detayları

Optimizing and comparison of market chain product distribution problem with different genetic algorithm versions

Yazarlar: Burak Gülmez

Yayıncı: Osmaniye Korkut Ata Üniversitesi

Ana Başlıklar: Endüstri Mühendisliği

Dergi: Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi Vol.6 Issue 1, pp.180-196

Belge Tipi: Makale

Yayın Tarihi: 10/2023

Belge Dili: Türkçe

Document ID (Org No.): 2001284864

pISSN: 2687-3729

Traveling salesman problem

Genetic algorithm

Product distribution

Gezgin satıcı problemi

Ürün dağıtımı

Genetik algoritma

Sayı

Yazarlar:

Burak Gülmez

Yayıncı:

Osmaniye Korkut Ata Üniversitesi

Ana Başlıklar:

Endüstri Mühendisliği

Dergi:

Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi Vol.6 Issue 1, pp.180-196

Belge Tipi:

Makale

Yayın Tarihi:

10/2023

Belge Dili:

Türkçe

Document ID (Org No.):

2001284864

pISSN:

2687-3729