Statistical and visualization data mining tools for foundry poduction

• Autorzy: Perzyk M.
• Języki publikacji: EN

Abstrakty

EN

In recent years a rapid development of a new, interdisciplinary knowledge area, called data mining, is observed. Its main task is extracting useful information from previously collected large amount of data. The main possibilities and potential applications of data mining in manufacturing industry are characterized. The main types of data mining techniques are briefly discussed, including statistical, artificial intelligence, data base and visualization tools. The statistical methods and visualization methods are presented in more detail, showing their general possibilities, advantages as well as characteristic examples of applications in foundry production. Results of the author's research are presented, aimed at validation of selected statistical tools which can be easily and effectively used in manufacturing industry. A performance analysis of ANOVA and contingency tables based methods, dedicated for determination of the most significant process parameters as well as for detection of possible interactions among them, has been made. Several numerical tests have been performed using simulated data sets, with assumed hidden relationships as well some real data, related to the strength of ductile cast iron, collected in a foundry. It is concluded that the statistical methods offer relatively easy and fairly reliable tools for extraction of that type of knowledge about foundry manufacturing processes. However, further research is needed, aimed at explanation of some imperfections of the investigated tools as well assessment of their validity for more complex tasks.

Słowa kluczowe

EN

casting quality management; data mining; visualization tools; statistical methods; foundry process parameters

PL

odlewnictwo; zarządzanie jakością w odlewnictwie; metody statystyczne; parametry procesów odlewniczych

• Wydawca: Komisja Odlewnictwa Polskiej Akademii Nauk Oddział w Katowicach
• Czasopismo: Archives of Foundry Engineering
• Rocznik: 2007
• Tom: Vol. 7, iss. 3
• Strony: 111--116
• Opis fizyczny: Bibliogr. 24 poz., rys., wykr.

Twórcy

autor

Perzyk M.

• Zakład Odlewnictwa ITMat, Politechnika Warszawska, ul. Narbutta 85, 02-524 Warszawa, Poland,

Bibliografia

• [1] D. Braha (ed.): Data Mining for Design and Manufacturing - Methods an Applications, Kluwer Academic Publ., Dordrecht, Boston, London, 2001.
• [2] H. Sadoyan, A. Zakarian, P. Mohanty, Data mining algorithm for manufacturing process control, Int. J. Advanced Manuf. Technol. vol. 28, No. 3-4 (2006) 342-350.
• [3] T. Demski, Statistics and data mining in practice, StatSoft, Warszawa - Kraków, 2004 (in Polish).
• [4] M. Perzyk, Data mining in foundry production, Research in Polish Metallurgy at the Beginning of XXI Century, Committee of Metallurgy of the Polish Academy of Sciences, ed. K. Świątkowski, Kraków, 2006.
• [5] X. Guo, Implementing Six Sigma in Foundry Industry, AFS Transactions, vol. 110 (2002), 199-210.
• [6] S. Kannan, J. E. Thixton, System Approach to Casting Defect Analyses and Reduction: Hydrogen Gas Defect in Iron Castings, AFS Transactions, vol. 112 (2004), 115-119.
• [7] P.L. Barker, B. Bidassie, Using Statistical Tools to Detect and Improve Core Shift: A Case Study, AFS Transactions, vol. 112 (2004), 121-130.
• [8] M. Perzyk, A. Kochański, Prediction of ductile cast iron quality by artificial neural networks, Journal of Materials Processing Technology, 109/3 (2001), 305-307.
• [9] M. Perzyk, J. Kozłowski, Comparison of statistical and neural networks-based methods in analysis of significance and interaction of manufacturing processes parameters, Computer Methods in Materials Science, vol. 6, No. 2 (2006), 81-93.
• [10] K. Hatanaka, T. Tanaka, H. Kominami, Breakout forecasting system based on multiple neural networks for continuous casting in steel production, Fuijtsu Scientific and Technical Journal, vol. 29 (1993), 265-270.
• [11] K. Terashima, Y. Maesa, H. Namura, Learning-control of mould hardness in blow molding, Proc. 60th World Foundry Congress, Hague 1993.
• [12] W. Chen, G. Duan, C. Ou, Neural network applied to predicting molten steel temperature profile from converter to continuous casting, Iron and Steel (China), vol. 32 (1997), 30-32.
• [13] Y. Otsuka, M. Konishi, Neural network models and its applications to iron and steel making processes, Journal of the Iron and Steel Institute of Japan, vol. 77 (1991), 1539-1543.
• [14] P.F. Bartelt, N.G. Bliss, J.S. Moberley, Application of artificial intelligence to power input control in the modern foundry. AFS Transactions, vol. 103 (1995), 221-225.
• [15] M.B. Brady, N.G. Bliss, H.D. Phillips, Integrated computer controls for the modern foundry meltshop. Proc. 51st Electric Furnace Conference, Washington, 1993, 117-120.
• [16] E.D. Larsen, D.E. Clark, H.B. Smart, K.L. Moore, Intelligent control of cupola melting. AFS Transactions, vol. 103 (1995), 215-219.
• [17] G. Wang, T.Y. Huang, Application of artificial neural networks in foundry industry. Proc. Third Asian Foundry Congress, Kyongju, South Korea, 1995, 424-431.
• [18] P.F. Bartelt, M.R. Grady, D. Dibble, Application of intelligent techniques for green sand control. AFS Transactions vol. 104 (1996), 635-642.
• [19] T. Watanabe, K. Omura, M. Konishi, K. Watanabe, K. Furukawa, Mold level control in continuous caster by neural network model, ISIJ International, vol. 39 (1999), 1053-1060.
• [20] S. Calcaterra, G. Campana, L. Tomesani, Prediction of Mechanical properties in spheroidal cast iron by neural networks, J. Mater. Proc. Technol., vol. 104 (2000), 74-80.
• [21] A. Faessler, M. Loher, Quality control in die casting with neural networks, Proc. Int. Symp. On Neuro-Fuzzy Systems, IEEE USA, Laussanne, 1996, 147-153.
• [22] P.K.D.V. Yarlagadda, E.C.W. Chiang, Neural network system for prediction of process parameters in pressure die casting, J. Mater. Proc. Technol., vol. 89-90 (1999), 583-590.
• [23] J.H. Zietsman, S. Kumar, J.A. Meech, I.V. Samarsekera, J.K. Brimacombe, Taper design in continuous billet casting using artificial neural networks, Ironmaking and Steelmaking, vol. 25 (1998), 476-483.
• [24] M. Perzyk, R. Biernacki, Modeling of manufacturing processes by learning systems: The naive Bayesian classifier versus artificial neural networks, J. Mater. Proc. Technol., vol. 164-165 (2005), 1430-1435.