İMALAT SİSTEMLERİNIN TASARIMINDA NÖROTİK TAVLAMA BENZETİMİ YAKLAŞIMININ KULLANILMASI

Year 2015, Volume: 16 Issue: 1, 35 - 42, 11.01.2016

Tarık Çakar , Serdar Özer

Abstract

Üretim sistemlerinin tasarımında simülasyon yöntemi kullanılması kaçınılmaz yöntemlerden birisidir. Fakat elenmesi gereken pek çok çözüm alternatifi varsa simülasyon metodunun yanında kullanılması gereken başka metotlarda vardır. Bu metotlardan en çok kullanılanları başta yapay sinir ağları, sonra genetik algoritmalar, tavlama benzetimi, parçacık sürü optimizasyonu gibi arama algoritmalarıdır. Bu çalışmada benzetim çalışması sonucu elde edilen imalat sistemine ait performans ölçütleri kullanılarak Yapay Sinir Ağı (YSA) eğitilmiştir. Yani YSA ya istenilen performans ölçüleri girildiğinde, YSA her bir makine merkezindeki olması gereken makine sayısını vermektedir. Bunun yanında imalat sistemlerinin performansını etkileyen en önemli faktör sistemde kullanılan öncelik kurallarıdır. Bu sebeple simülasyon işlemini yaparken makine merkezlerindeki makine sayılarının değiştirilmesinin yanı sıra farklı öncelik kuralları da kullanılarak sistemin performans ölçütleri elde edilmiştir. Kullanılan öncelik kuralları SPT, EDD, CR, FCFS olmuştur. Her bir öncelik kuralı için farklı bir YSA eğitilmiştir. Çözüm elde etmek için istenilen performans ölçütleri eğitilen bu altı farklı YSA ya verilir. Elde edilen sonuçlar tekrar simüle edilerek performans ölçüleri elde edilir. İstenilen performans ölçülerine en yakın performans değerleri çözüm olarak alınır. Alınan bu çözümden daha iyi bir çözüm olup olmadığının araştırılması için ise Tavlama Benzetimi yaklaşımı kullanılır. Eğitilen YSA’ ların %0.5 hata verme olasılığı vardır. Bu yüzden az bir olasılıkla da olsa gözden kaçan iyi bir çözüm varsa bunu yakalamak için Tavlama Benzetimi yaklaşımı kullanılmıştır. Alınan 100 farklı çözümün 12 tanesinde YSA dan elde edilen çözümden daha iyi bir çözüm elde edilmiştir.   

Keywords

Tavlama benzetimi, yapay sinir ağları, imalat hücresi tasarımı

USING NEURO-SIMULATED ANNEALING APPROACH TO DESIGN MANUFACTURING SYSTEMS

Abstract

The usage of simulation methods is one of indispensable methods in the design of production systems. However, if there are some other solution alternatives to be eliminated, there should be some other solution methods to be used with simulation method. The most used methods for this purpose are mainly artificial neural networks (ANN), then genetic algorithms, simulated annealing, particle swarm optimization. In this study, an artificial neural network has been trained by using obtained performance criterion belonging production system as a result of simulation studies. In other words, when the performance criteria are entered to ANN, ANN gives the number of machines that should be in each machine center. Additionally, the most important factor affecting the per-formance of production systems is the priority rules used in the system. Therefore, in addition to changing the number of machines during the implementation of the simulation process in the machine centers performance criteria of the system have been obtained by using various priority rules. The used priority rules have been SPT, EDD, CR, FCFS. For each priority rule different neural network has been trained. To obtain the solution, re-quired performance criteria are given to these trained four neural networks. The performance criteria are obtained by re-simulation of the obtained results. The nearest performance values to the performance criteria are taken as solution. The simulated annealing approach is used to investigate whether there is a better solution than this tak-en solution. There is error possibility of the trained neural networks as %0.5. Therefore, simulated annealing ap-proach has been used to capture better solution missed if there is. %12 of the solutions among the obtained dif-ferent solutions have been found better than the solutions obtained by using ANN.

Keywords

Simulated Annealing, Artificial Neural Networks, Manufacturing Cell Design

References

• CAKAR, T., CIL, I., “Artificial Neural Networks for design of manufacturing systems and selection of priorıty rules”, International Journal of Computer Integrated Manufacturing, vol.17, 3: 195-211, 2004.
• CAKAR T. , YAZGAN, HR, KOKER, R., “Parallel robot Scheduling with genetic algorithms”, Parallel robot manipulators, New Developments. In: Ryu J-H (Ed), Vienna, I-Tech Education and Publishing, 2008, pp. 153-170.
• CHEN, W.H., SRIVASTAVA, B., “Simulated annealing procedures for forming machine cells in group technology”, 75: 100-111, 1994.
• CHRYSSOLOURIS, G., LEE, M. , “Use of meural networks for design of manufacturing systems”, Manufacturing Review, vol.3, 3: 187-194, 1990.
• DEFERSHA, F., CHEN,M., “A comprehensive mathematical model for the design of cellular manufacturing systems”, International Journal of Production Economy, 103: 767-783, 2006.
• DEFERSHA, F., CHEN,M., “Machine cell formation using a mathematical model and a genetic- algorithm-based heuristic”, International Journal of Production Research, vol.44, 12: 2421-2444, 2006.
• DOGUC, U., Doguc, “Esneklik imalat sistemlerinde makina sayılarının ve teslim tarihinin belirlenmesinde yapay sinir ağlarının kullanılması”, PhD dissertation, Endustri Mühendisliği Bölümü, Sakarya Üniversitesi, Sakarya, 2001.
• GUPTA,Y. GUPTA, A. KUMAR, A., SURDARAM, C., “A genetic algorithm-based approach to cell composition and layout design problems”, International Journal of Production Research, 34:.447-482, 1996.
• KHOO, L.P., LEE, S.G., YEN, X.F., “Multi-objective optimization of machine cell layout using genetic algorithm”, International Journal of Computer Integrated Manufacturing, 16: 140-155, 2003.
• KIRKPATRICK, S, GELATT, CD, VECCHI, MP., “Optimization by simulated annealing”, Science, vol.220, 4598: 671– 680, 1983.
• KIOONS, S.A., GULGAK, A.A., BEKTAS, T. “Integrated cellular manufacturing systems design with production planning and dynamic system reconfiguration”, Europan Journal of Operational Research, 192: 414-428, 2009.
• KOKER R., “A neuro-simulated annealing approach to the inverse kinematics solution of redundant robotic manipulators”, Engineering With Computers, Vol. 29, 4: 507-515, 2013.
• MALAKOOTI, B., YANG, Z., “Multiple criteria approach and the generation of efficient alternatives for machine-part family formation in group technology”, IIE Transaction, 34:, 837-846, 2002.
• MANSOURI, S.A., MOATTAR-HUSSEINI, S.H., ZEGORDI, S.H., “A genetic algorithm for multiple objective dealing with exceptional elements in cellular manufacturing”, Production Planning Control, vol.14, 5:437-446, 2000.
• MAHDAVI, I., ALAEI, A., PAYDAR, M.M., SOLIMANPOUR, M., “Designing a mathematical model for dynamic cellular manufacturing systems considering production planning and worker assignment”, Computational Mathematic with Applications, 3: 74-82 , 2010.
• RHEAULT, M., DROLET, J., ABOULNOUR, G., “ Physically reconfigurable virtual cells: a dynamic model for a highly dynamic environment”, Computers and Industrial Engineering, vol.29, 1-4: 221-225, 1995.
• SCHLUNZ, E.B., VAN HUUREN, J.H., “An investigation into the effectiveness of simulated annealing as a solution approach for the generator maintenance scheduling problem, Electrical Power and Energy Systems, 53: 166–174, 2013.
• SHAFER, S.M., “A gola programming approach to cell formulation problems”, Journal of Operational Management, 10: 28-43, 1991.
• SOLIMANPOU, M., SAEEDI, I., MAHDAVI, I., “Solving cell formation problem in cellular manufacturing using ant-colony-based optimization”, International Journal of Advanced Manufacturing Technology, vol.50, 9-12: 1135-1144, 2010.
• SOFIANOPOULOU, S., “Manufacturing cells design with alternative process plans and or replicate machines”, International Journal of Production Research, vol.37, 3: 707-720., 1999.
• SU, C-T., HSU, C-M., “Multi-objective machine-part cell formation through parallel simulated annealing”, International Journal of Production Research, 36: 2185-2207, 1998.
• VENUGOPAL, V., NARENDRAN, T,T, “A genetic algorithm approach to the machine-component grouping problem with multiple objectives ”, Computers and Industrial Engineering, 22: 469-480, 1992.
• VENUGOPAL, V., NARENDRAN, T,T, “Cell formation in manufacturing systems through simulated annealing: an experimental evaluation” , Europan Journal of Operational Research, 63:409-422, 1992.
• WEI, J.C., GAITHER, N., “A capacity constrained multi-objective cell formation method”, Journal of Manufacturing Systems, 9: .222-232, 1990.
• YASUDA, K., HU,L., YIN, Y., “A grouping genetic algorithm for the multi-objective cell formation problem”, International Journal of Production Research, vol.43, 4: 829-853, 2005.