Capacity Balancing and Variation Minimization in Unrelated Parallel Machine Scheduling with Sequence-Dependent Setup Times: A Mathematical Model and Matheuristic Approach

Merhad Ay

Makale detayları

Capacity Balancing and Variation Minimization in Unrelated Parallel Machine Scheduling with Sequence-Dependent Setup Times: A Mathematical Model and Matheuristic Approach

Yazarlar:

Merhad Ay

Yayıncı:

Gazi University

Ana Başlıklar:

Endüstri Mühendisliği | Üretimde Optimizasyon

Dergi:

Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji Vol.13 Issue 3, pp.919-931

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001353671

pISSN:

2147-9526

Anahtar kelimeler:

Özet:

 In this study, a scheduling problem for unrelated parallel machines with sequence-dependent setup times, specifically in the context of tire production, is addressed. To solve this problem, a mathematical model and genetic algorithm-based matheuristic approaches have been developed. While the mathematical model provides accurate and effective solutions for small and medium scale problems, it exhibits limitations in time-constrained applications due to increased solution times for large-scale problems. In the study, two alternative matheuristic algorithms employing different crossover operators are proposed, and their ability to achieve near-optimal solutions for large-scale problems in short periods has been tested. Additionally, the classical random mutation operator was modified into a constrained random mutation operator tailored to the problem. Experimental results demonstrate that the proposed matheuristic algorithms significantly reduce solution times compared to the mathematical model, with the MA1 algorithm showing superior performance in terms of solution quality. This study offers substantial advantages in solving real-world problems by improving both solution time and quality. In the proposed matheuristic algorithm, the problem-specific chromosome structure, the initialization method for the initial population, and the constrained random mutation operator provide significant contributions to the literature for solving similar problems.

 In this study, a scheduling problem for unrelated parallel machines with sequence-dependent setup times, specifically in the context of tire production, is addressed. To solve this problem, a mathematical model and genetic algorithm-based matheuristic approaches have been developed. While the mathematical model provides accurate and effective solutions for small and medium scale problems, it exhibits limitations in time-constrained applications due to increased solution times for large-scale problems. In the study, two alternative matheuristic algorithms employing different crossover operators are proposed, and their ability to achieve near-optimal solutions for large-scale problems in short periods has been tested. Additionally, the classical random mutation operator was modified into a constrained random mutation operator tailored to the problem. Experimental results demonstrate that the proposed matheuristic algorithms significantly reduce solution times compared to the mathematical model, with the MA1 algorithm showing superior performance in terms of solution quality. This study offers substantial advantages in solving real-world problems by improving both solution time and quality. In the proposed matheuristic algorithm, the problem-specific chromosome structure, the initialization method for the initial population, and the constrained random mutation operator provide significant contributions to the literature for solving similar problems.

 In this study, a scheduling problem for unrelated parallel machines with sequence-dependent setup times, specifically in the context of tire production, is addressed. To solve this problem, a mathematical model and genetic algorithm-based matheuristic approaches have been developed. While the mathematical model provides accurate and effective solutions for small and medium scale problems, it exhibits limitations in time-constrained applications due to increased solution times for large-scale problems. In the study, two alternative matheuristic algorithms employing different crossover operators are proposed, and their ability to achieve near-optimal solutions for large-scale problems in short periods has been tested. Additionally, the classical random mutation operator was modified into a constrained random mutation operator tailored to the problem. Experimental results demonstrate that the proposed matheuristic algorithms significantly reduce solution times compared to the mathematical model, with the MA1 algorithm showing superior performance in terms of solution quality. This study offers substantial advantages in solving real-world problems by improving both solution time and quality. In the proposed matheuristic algorithm, the problem-specific chromosome structure, the initialization method for the initial population, and the constrained random mutation operator provide significant contributions to the literature for solving similar problems.

 Bu çalışmada, sıra bağımlı hazırlık sürelerine sahip özdeş olmayan paralel makineler için lastik üretimi özelinde bir çizelgeleme problemi ele alınmış ve bu probleme yönelik bir matematiksel model ile genetik algoritma tabanlı matsezgisel yöntemler geliştirilmiştir. Matematiksel model, küçük ve orta ölçekli problemlerde etkin ve doğru çözümler sunarken, büyük ölçekli problemlerde çözüm süresinin artması nedeniyle zaman kısıtlı uygulamalarda sınırlamalar göstermektedir. Çalışmada iki farklı çaprazlama operatörünün kullanıldığı iki alternatif matsezgisel algoritma önerilmektedir ve bu algoritmaların büyük problemlerde kısa sürelerde optimuma yakın sonuçlar elde edebileceği test edilmiştir. Ayrıca, mutasyon aşamasında klasik rastgele mutasyon operatörü, probleme özgü olarak modifiye edilerek sınırlandırılmış rastgele mutasyon operatörü olarak kullanılmıştır. Deneysel sonuçlar, önerilen matsezgisel algoritmaların matematiksel modele kıyasla çok daha kısa sürelerde çözüm sağladığını ve MA1 algoritmasının çözüm kalitesi açısından iyi performans sergilediğini göstermektedir. Çalışma, çözüm süresini ve kalitesi açısından gerçek hayat problemlerinin çözümünde önemli avantajlar sunmaktadır. Önerilen matsezgisel algoritmada, probleme özgü olarak tasarlanan kromozom yapısı, başlangıç popülasyonu oluşturma yöntemi ve sınırlandırılmış rastgele mutasyon operatörü, benzer problemlerin çözümüne yönelik literatüre önemli katkılar sunmaktadır.

Makale detayları

Capacity Balancing and Variation Minimization in Unrelated Parallel Machine Scheduling with Sequence-Dependent Setup Times: A Mathematical Model and Matheuristic Approach

Yazarlar:

Merhad Ay

Yayıncı:

Gazi University

Ana Başlıklar:

Endüstri Mühendisliği | Üretimde Optimizasyon

Dergi:

Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji Vol.13 Issue 3, pp.919-931

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001353671

pISSN:

2147-9526

Parallel Machine Scheduling

Matheuristic

Optimization

Capacity BalancingParalel Makine Çizelgeleme

Matsezgisel

Optimizasyon

Kapasite Dengeleme

Sayı

Makale detayları

Yazarlar: Merhad Ay

Yayıncı: Gazi University

Ana Başlıklar: Endüstri Mühendisliği | Üretimde Optimizasyon

Dergi: Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji Vol.13 Issue 3, pp.919-931

Belge Tipi: Makale

Yayın Tarihi: 9/2025

Belge Dili: İngilizce

Document ID (Org No.): 2001353671

pISSN: 2147-9526

Sayı

Yazarlar:

Merhad Ay

Yayıncı:

Gazi University

Ana Başlıklar:

Endüstri Mühendisliği | Üretimde Optimizasyon

Dergi:

Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji Vol.13 Issue 3, pp.919-931

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001353671

pISSN:

2147-9526