Altı sigma ve yapay sinir ağlarının tekstil sektöründe karşılaştırmalı bir uygulaması
Loading...
Files
Date
2010
Authors
Yılmaz, Ümit
Journal Title
Journal ISSN
Volume Title
Publisher
Uludağ Üniversitesi
Abstract
İkibinsekiz yılının son aylarında dünyayı saran ekonomik durgunluk ve enerji birim giderlerinde meydana gelen artış, işletmeleri kalitelerini düşürmeden maliyet minimizasyonuna itmiş ve bu doğrultuda iyileştirme çalışmalarına yön vermiştir.Bu çalışmada, tekstil sektöründe faaliyetini sürdüren bir firmanın üretim sürecinde meydana gelen hataların minimize edilmesi amacıyla altı sigma uygulaması gerçekleştirilmiştir. Kalite kontrol sürecinde tespit edilen hataların üretim sürecinde tekrarlanmaması amacıyla ikibinsekiz yılında gerçekleşen vardiya tipi, işletme nemi, işletme sıcaklığı, dış sıcaklık, makine tipi, tezgah hızı, dokumacı yaşı, renk tipi, yüzey tipi, ip tipi, üretim miktarı, ebat tipi, en sayısı, gramaj, atkı sayısı, hav oranı ve hata miktarı değerleri derlenmiştir. Sonrasında regresyon analizi ile belirtilen verilerin hata oluşumuna etkileri değerlendirilerek regresyon denklemine ulaşılmıştır. Elde edilen regresyon denklemine bağlı olarak optimum girdi değerlerinin bulunabilmesi için GAMS programı kullanılmıştır. Daha sonra veriler girdi ile çıktı değişkenleri arasındaki ilişkiyi incelemek amacıyla yapay sinir ağları ile analiz edilmiştir. Regresyon analizinden GAMS programı kullanılarak optimum girdi değerlerinin oluşumu sonrasında ulaşılan performans kriterleriyle yapay sinir ağları yöntemleriyle elde edilen performans kriterleri karşılaştırılarak yöntemlerin etkinliği incelenmiştir.Sonuç olarak, istenilen değerlere yüzde yüz oranında ulaşılırken, işletme iyileşme yolunda önemli bir yol katetmiştir.
Economic recession covering all the world in the last few months of two thousand eight and the increase in the energy unit expenses caused business enterprises to make expenditure minimization with no loss in quality and caused efforts for improvement in terms of this matter.In this work, Six Sigma application was fulfilled with the aim to minimize the defects that occurred during the production process of a firm maintaining its business in the textile sector. With the aim of not repeating the defects determined during quality control process in the process of production, the elements that was fulfilled in the year two thousand eight such as shift type, plant humidity, plant temperature, outdoor temperature, machine type, machine speed, weaver age, colour type, surface type, yarn type, quantity of production, dimension style, number of width, weight in grams, number of weft, fluff proportion, and amount of defects were arranged. Afterwards, regression equation was obtained by evaluating the effects of data determined through regression analysis on the emerging of defects. GAMS program was used to be able to find optimum input rates depending on regression equation obtained before. Then, in order to examine the relationship between input and output variables, data was analyzed with artificial neural networks. By comparing the performance criteria that was reached after the formation of optimum input rates through the use of GAMS program of regression analysis with the performance criteria that was obtained with the artificial neural network method, the efficiency of methods were examined.Consequently, while reaching the intended rates in proportion of 100%, the business enterprise got far in terms of improvement.
Economic recession covering all the world in the last few months of two thousand eight and the increase in the energy unit expenses caused business enterprises to make expenditure minimization with no loss in quality and caused efforts for improvement in terms of this matter.In this work, Six Sigma application was fulfilled with the aim to minimize the defects that occurred during the production process of a firm maintaining its business in the textile sector. With the aim of not repeating the defects determined during quality control process in the process of production, the elements that was fulfilled in the year two thousand eight such as shift type, plant humidity, plant temperature, outdoor temperature, machine type, machine speed, weaver age, colour type, surface type, yarn type, quantity of production, dimension style, number of width, weight in grams, number of weft, fluff proportion, and amount of defects were arranged. Afterwards, regression equation was obtained by evaluating the effects of data determined through regression analysis on the emerging of defects. GAMS program was used to be able to find optimum input rates depending on regression equation obtained before. Then, in order to examine the relationship between input and output variables, data was analyzed with artificial neural networks. By comparing the performance criteria that was reached after the formation of optimum input rates through the use of GAMS program of regression analysis with the performance criteria that was obtained with the artificial neural network method, the efficiency of methods were examined.Consequently, while reaching the intended rates in proportion of 100%, the business enterprise got far in terms of improvement.
Description
Keywords
Altı Sigma, Hata, Regresyon analizi, Tekstil, Yapay sinir ağları, Six sigma, Defect, Regression analysis, Textile, Artificial neural networks
Citation
Yılmaz, Ü. (2010). Altı sigma ve yapay sinir ağlarının tekstil sektöründe karşılaştırmalı bir uygulaması. Yayınlanmamış yüksek lisans tezi. Uludağ Üniversitesi Fen Bilimleri Enstitüsü.