M. ZIARATI, D. STOCKTON, O. N. UÇAN, E. BİLGİLİ

Design and Development of Material and Information Flow Supply Chains Using Genetic Cellular Networks

Son yıllarda, geri yayıhm tekniğine dayanan yapay sinir ağı (Ziarati and Ucan, January 2001) modeli ile gerçek bir firmanın malzeme tedarik zincirinde geleceğe dönük malzeme talep miktarı tahmin edilebilmiştir. Yapay sinir ağlarının hızlı olması, büyük miktardaki verinin ele alınabilmesi, malzeme akış diagramlarında geleceğe yönelik tahminlerde potensiel bir model olmalarını sağlamaktadır. Bu makale, (Ziarati and Ucan, January 2001) makalesinin geliştirilmiş biçimidir. Burada yapay sinir ağ (YSA) yapısı yerine Genetik Hücresel Yapay Sinir Ağ (HYSA) modeli konulmuştur. Söz konusu yaklaşım daha az parametre ile kestirim yapabilmekte ve dolayısıyla hızlı değişimli gerçek tedarik zincir problemlerine daha hızlı uyum sağlamaktadır. Önerilen modelin, tedarik zinciri problemlerinde, gerek eğitim sürecinin kısaltılmasında gerekse malzeme istek kestirimde üstün başarım göstermesi beklenmektedir.

In a recent paper by authors (Ziarati and Ucan, January 2001) a Back Propagation-Artificial Neural Network (BP-ANN) was adapted for predicting the required car parts quantities in a real and major auto parts supplier chain. It was argued that due to the learning ability of neural networks, their speed and capacity to handle large amount of data, they have a potential for predicting components requirements and establishing associated scheduling throughout a given supply chain system. This paper should be considered a continuation of the first paper as the neural network approach introduced in this paper replaces the BP-ANN by a new method viz., Genetic Cellular Neural Network (GCNN). The latter approach requires by far less stability parameters and hence better suited to fast changing scenarios as in real supply chain applications. The model has shown promising outcomes in learning and predicting material demand in a supply chain, with high degree of accuracy.

PDF

___

CHUA, L. 0., YANG, L. (1988). “Cellular Neural Networks: Theory”, IEEE Trniis. Circuit and S ysteitis, V35, pp.1257-1272.

DAVIS, L., (1991). Hnitdhook of Genetic Al goritluns, New York: Van Nostrand Reinhold.

HOLLAND, J.H. (1975). “Outline for a logical theory of adaptive systems: J. Assoc.” Computer, v.3, pp.297-314.

HOLLAND, J.H. (1975). Adaptation in neural and artiffCial s ystenis, Ann Arbor, MI: University of the Michigan Press

KOZEK, T., ROSKA, T., CHUA, L.O. (1988). “Genetic Algorithms for CNN template Learning”, IEEE Trans. Circuit aii‹f S ysteiiis, V40, pp.392-402.

STOCKTON, D.T., QUINN, L. (1993). “Identifying Economic Order Quantities Using Genetic Algorithms” liiternationnl lournnl of Oyeratioiis rind Production Manngeinent, v.3, n.11.

UÇAN, 0. et al. (2001). “Separation of Bouguer anomaly map using cellular neural network”, Jotii tial of Applied Geophysics 46, pp.129-142.

WANG, Q. (2000, November), Intf›roving the Cost Model Developitieiit Process Using Neural Networks, Thesis, De Monfort University.

ZIARATI, M., UCAN, O.N. (2001, January). “Optimisation of Economic Order Quantity Using Neural Networks Approach”, Dogu› f/itiversi Jotirnnl Number, No: 3, pp.128-140.

ZIARATI, R. (1994, May). “Factories of the Future”, Invited paper, EUROTECNET Conference, Germany

ZIARATI, R., KHATAEE, A. (1994, April). “Integrated Business Information System (IBIS) — A Quality Led Approach”, Keynote Address. SheMet 94, Belfast University Press, Ulster, UK