A performance analysis for minimum spanning tree based genetic algorithm approach to design and optimize distribution network problem
Yükleniyor...
Dosyalar
Tarih
2006
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Erişim Hakkı
info:eu-repo/semantics/openAccess
Özet
Supply chain management encompasses a complex business rela­tions network that contains synchronized inter and intra organizational efforts to optimizing forward flow of materials and reverse flow of knowledge through the overall production and distribution system from raw material suppliers to the last customers. Thus, one of the vital issues in supply chain management is design and optimization of a large network consisting of various business enti­ties such as suppliers, manufacturers, distribution centers and customers. In the literature, this design problem is generally defined as non polynomial-hard (NP-hard) and genetic algorithm approach is one of the popular solution methods for the problem. From this point of view, in this study, performance analysis for minimum span­ning based genetic algorithm approach to a design and optimization of distri­bution networks problem (which is a combination of multiple-choice Knapsack problem with the capacitated location-allocation problem) in supply chain man­agement is made and optimum combination of genetic operators is searched. 162 experimentations are made by a genetic algorithm based program written in Borland Delphi 6.0. The program is run 972 times and as a result of experi­mentations, optimum combination is determined as elitist tournament /partially matching (PMX) for the problem. Furthermore, there is no statistical differ­ence at p <0.05 significance level between two mutation operators, which are analyzed. Also, 50x1500 is found as optimum for the combination of population size-generation number.
Açıklama
Anahtar Kelimeler
Matematik, Genetic Algorithms, Performance Analysis.
Kaynak
Türkiye Klinikleri Psikiyatri Dergisi
WoS Q Değeri
Scopus Q Değeri
Cilt
7
Sayı
1
