Identificación Inteligente de un Proceso Fermentativo Usando el Algoritmo GMDH Modificado

F. Hernández, F. Herrera

Resumen

En este trabajo se aborda, de manera particular, un método para el diseño del algoritmo conocido como Group Method of Data Handling, GMDH, típico con lazo recurrente. Una modificación en una de sus fases de entrenamiento permite ampliar el número de variables utilizadas en cada capa y con ello el área de regresión. Consecuentemente se puede obtener una estructura optimizada en sí misma de mayor complejidad, posibilitando la aparición de lazos recurrentes en las capas intermedias. Lo anterior permite una reducción del error en la modelación de procesos no lineales de lento comportamiento, como el crecimiento celular en biorreactores. El modelo se probó en una fermentación tipo feed-batch de la levadura Pichia pastoris. La estabilidad y capacidad de generalización es demostrada. El método propuesto es comparado con el GMDH típico recurrente y con otras estructuras de redes neuronales clásicas.

Palabras clave

redes neuronales; recurrente; algoritmo genético; modelación; fermentación

Texto completo:

PDF

Referencias

Bagheri A., Nariman-Zadeh N., Babaei M., and Jamali A., 2007. Polynomial Modeling of the Controlled Rack-Stacker Robot Using GMDH-type Neural Networks and Singular Value Decomposition. International Journal of Nonlinear Sciences and Numerical Simulation, 8(3), 301-310.

Balanza García José, Cano I. J. M., López C. J., Alvarez C. A., Luna M. M., 1998. Estudio y Desarrollo de Sensores Software Basados en Sistemas Neuro-Difusos. Aplicación en Procesos Petroquímicos. Departamento de Ingeniería de Sistemas y Automática de la UPCT. TIC99-0446-C02-01.

Drchal, J., 2006. Evolution of Recurrent Neural Networks., Czech Technical University, Faculty of Electrical Engineering, Prague.

Ferreira, C., 2001. Gene Expression Programming in Problem Solving. Complex Systems.

Ferreira, C., 2004. Designing Neural Networks Using Gene Expression Programming. Paper presented at the 9th Online World Conference on Soft Computing in Industrial Applications.

Happel, B. L. M. and Murre, J. M. J., 1990. Structure Identification of Non Linear Dynamic Systems - A survey on input/output Approaches. Automática, 26(4), 651-667.

Holland, J. H., 1975. Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence. University of Michigan Press, (second edition: MIT Press, 1992).

Ivakhnenko A. G., K. V. V., Tetko I. V., Luik A.I., Ivakhnenko G.A., Ivakhnenko N.A., 1999. Self-Organization of Neural networks with Active Neurons for Bioactivity of Chemical Compounds Forecasting by Analogues Complexing GMDH Algorithm. Paper presented at the Poster for the ICANN'99 Conference.

Ivakhnenko A.G., 1971. Polynomial theory of complex systems. IEEE Transactions on Systems, Man, and Cybernetics, SMC-1(1).

Ivakhnenko A.G., D. W., Ivakhnenko G.A., 1998. Inductive Sorting-Out GMDH Algorithms with Polynomial Complexity for Active Neurons of Neural Network. from http://come.to/GMDH

Ivakhnenko G.A., 2001. Inductive Self-Organizing Algorithm for Maximum Electrical Load Prediction. International Centre of Informational Technologies and Systems of the National, Ac. Sci. Ukrania, Kyiv.

Jenzsch, M., Simutis, R., Lübbert, A., 2006. Optimization and Control of Industrial Microbial Cultivation Processes. Eng. Life Sci., 6(2), 117-124.

Känsäkoski, M., Kurkinen, Marika, von Weymarn, Niklas, Niemelä, Pentti, Neubauer, Peter, Juuso, Esko, Eerikäinen, Tero, Turunen, Seppo, Aho, Sirkka & Suhonen, Pirkko., 2006. Process analytical technology (PAT) needs and applications in the bioprocess industry. VTT Technical Research Centre of Finland, 60, 99.

Kondo, T., 2003. Revised GMDH-type neural networks with radial basis functions and their application to medical image recognition of stomach. Systems Analysis Modeling Simulation, 43(10), 1363-1376.

Kondo, T. and Ueno, J., 2006. Revised GMDH-Type Neural Network Algorithm With a Feedback Loop Identifying Sigmoid Function Neural Network. International Journal of Innovative Computing, Information and Control, 5(2), 985—996.

Kondo, T. and Ueno, J., 2009. Medical Image Recognition of Abdominal Multi-Organs by Rbf Gmdh-Type Neural Network. International Journal of Innovative Computing, Information and Control, 5(1), 225-240.

Leiva, G. A. (2006). Redes Neuronales como Herramienta para la Automatización de Sistemas Complejos. Paper presented at the EVIC2006.

Mark S. Voss, Xin Feng., 2002. A new methodology for emergent system identification using Particle Swarm Optimization (PSO) and the Group Method of Data Handling (GMDH). Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2002.

McNeil B., Harvey L. M., 1990. Fermentation a Practical Approach (1ra ed.). Oxford: Oxford University Press.

Miroslav Šnorek, P. K., 2006. Inductive Modelling World Wide the State of the Art. Report of investigation, Dept. of Computer Science and Engineering, Karlovo nam.

Mutasem Hiassat, N. M., 2004. An evolutionary method for term selection in the Group Method of Data Handling. Automatic Control & Systems Engineering, University of Sheffield, 11-14.

Nariman-Zadeh, N., Darvizeh, A., Ahmad-Zadeh, R., 2003. Hybrid Genetic Design of GMDHType Neural Networks Using Singular Value Decomposition for Modelling and Prediction of the Explosive Cutting Process. Journal of Engineering Manufacture, 217, 779-790.

Nariman-Zadeha N., Darvizeha A., Jamali A., Moeinib A., 2005. Evolutionary design of generalized polynomial neural networks for modeling and prediction of explosive forming process. Paper presented at the 13th International Scientific Conference on Achievements in Mechanical and Materials Engineering.

Nariman-zadeha N., Jamali A., 2007. Pareto Genetic Design of GMDH-type Neural Networks for Nonlinear Systems: Department of Mechanical Engineering, University of Guilan.

Pappas S. Sp., Ekonomou L., 2006. Comparison of Artificial Intelligence Methods for Predicting the Time Series Problem. Proceedings of the 6th WSEAS International Conference on Simulation, Modelling and Optimization, 22-24.

Passoni, L. I., 2005. Modelos en Bioingeniería: Caracterización de Imágenes Estáticas y Dinámicas. Tesis del Doctorado en Ingeniería, Universidad Nacional de Mar del Plata.

Ramsey, A., 1994. Assessment of the modeling abilities of Neural networks. U. of Massachusetts, US.

Reed G., Rehm J., Puhler A., Stadler P., 2000. Biotechnology, Measuring modelling and control. VHC, 4, 181.

Royce, P. N., 1993. A discussion of recent developments in fermentation monitoring and control from a practical perspective. Critical reviews in biotechnology, 13, 117-149.

Sánchez, P. A., 2008. Redes Neuronales Recurrentes en el Modelado de la Tasa de Cambio Colombiana. III Congreso Colombiano de Computación. Abril 23-25 Medellín.

Solla, S. A., 1989. Learning and Generalization in Layered Neural Network: the Contiguity Problem. Neural Networks: from Models to Aplications. In L. Personnas and G. Dreyfus Paris: I.D.S.E.T.

Yuan J.Q., Vanrolleghem P. A., 1999. Rolling learning-prediction of product formation in bioprocesses. Journal of Biotechnology (Elsevier Science B.V.), 69, 47–62.

Abstract Views

337
Metrics Loading ...

Metrics powered by PLOS ALM


 

Citado por (artículos incluidos en Crossref)

This journal is a Crossref Cited-by Linking member. This list shows the references that citing the article automatically, if there are. For more information about the system please visit Crossref site

1. Integration of dynamic rainfall data with environmental factors to forecast debris flow using an improved GMDH model
Hui Zhang, Xiangnan Liu, Erli Cai, Gang Huang, Chao Ding
Computers & Geosciences  vol: 56  primera página: 23  año: 2013  
doi: 10.1016/j.cageo.2013.02.003



Creative Commons License

Esta revista se publica bajo una Licencia Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Universitat Politècnica de València     https://doi.org/10.4995/riai

e-ISSN: 1697-7920     ISSN: 1697-7912