|
Vol: 55(69) No: 4 / December 2010 Dynamic High Order Neural Networks. An Application for Inter-stand Tension Modelling in a Steel Cold Mill Eugen Arinton “Dunărea de Jos” University, Faculty of Electrical Engineering and Electronics, Stiintei-2, 800146 Galaţi, Romania, phone: (40) 236-470 905, e-mail: Eugen.Arinton@ugal.ro Sergiu Caraman “Dunărea de Jos” University, Faculty of Electrical Engineering and Electronics, Stiintei-2, 800146 Galaţi, Romania, e-mail: Sergiu.Caraman@ugal.ro Keywords: dynamic neural networks, high order neurons, system identification, cold rolling process Abstract The paper presents a particular type of multi-layer neural networks that can be used for modelling non-linear dynamic processes. The neurons of these networks are characterized by non-linearly pre-processed inputs. Dynamic properties can be obtained by adding a filter to the neuron. The networks, built with this type of neurons, have a structure that develops during the training process in such a way that fits the complexity of the modelled system. Applications of these networks for the system identification of a complex industrial process are discussed in the final part. References [1] C.M. Bishop, Neural Networks for Pattern Recognition, Oxford University Press, Oxford, 1995. [2] G.F. Bryant (Ed.), Automation of tandem mills, Iron and Steel Institute, 1973 [3] J.L. Castro, C.J. Mantas and J.M. Benitez, "Neural networks with a continuous squashing function in the output are universal approximators", Neural Networks, Vol. 13, pp. 561--563, 2000. [4] K. Dutton, S. Thompson and B. Barraclough, The Art of Control Engineering, Addison-Wesley, 1997. [5] S.J. Farlow (Ed.), Self-organizing Methods in Modelling: GMDH-type Algorithms, Marcel Dekker Inc., New York, 1984. [6] M.M. Gupta, L. Jin and N. Homma, Static and Dynamic Neural Networks: From Fundamentals to Advanced Theory, Wiley-IEEE Press, 2003. [7] M.T. Hagan and M. B. Menhaj, “Training feedforward networks with the marquardt algorithm”, IEEE Trans. Neural Networks, Vol. 5, No. 6, pp. 989—993, 1994. [8] B. Hammer, “On the approximation capability of recurrent neural networks”, Neurocomputing, Vol. 31, pp. 107—124, 2000. [9] S.S. Haykin, Neural Networks: A Comprehensive Foundation (2nd Edition), Prentice Hall, 1998. [10] J. Korbicz and J. Kus, "Diagnosis of sugar factory processes using GMDH neural networks", Proc. 4th IFAC Symp. Fault Detection Supervision and Safety for Technical Processes, Safeprocess 2000, Budapest, Hungary, June 14-16, Vol. 1, pp. 343--347, 2000. [11] J. Korbicz, K. Patan, A. Obuchowicz, “Neural network fault detection system for dynamic processes”, Buletin of the Polish Academy of Sciences. Technical Sciences, 2001, Vol. 49, No. 2, pp. 301-321. [12] E.B. Kosmatopoulos, M.M. Polycarpou, M.A. Christodoulou and P.A. Ioannou, “High-order neural network structures for identification of dynamical systems”, IEEE Trans. on Neural Networks, Vol. 6, No. 2, pp. 422—431, 1995. [13] G.G. Lorentz, Approximation of Functions, Holt, Rinehart and Winston, New York, 1996. [14] K.S. Narendra and K. Parthasarathy, “Identification and control of dynamical systems using neural networks”, IEEE Trans. Neural Networks, Vol. 1, No.1, pp. 4—27, 1990. [15] O. Nelles, Nonlinear System Identification: From Classical Approaches to Neural Networks and Fuzzy Models, Springer Verlag, Berlin, 2000. [16] D.T. Pham, L. Xing, Neural Networks for Identification, Prediction and Control., Springer Verlag, London, 1995. [17] W.L. Roberts, Cold Rolling of Steel, Marcel Dekker, Inc. New York and Basel, 1978. [18] F. Scarselli and A.C. Tsoi, “Universal approximation using feedforward neural networks: A survey of some existing methods, and some new results”, Neural Networks, Vol. 11, No. 1, pp. 15—37, 1998. [19] E. Arinton and S. Caraman, “Modelling the Inter-Stand Tension of a Steel Cold Mill Based on Dynamic High Order Neural Networks”, Proceedings of ICCC-CONTI 2010, Timisoara, Romania, May 27-29, 2010, Paper on the CD. |