covid
Buscar en
Revista Iberoamericana de Automática e Informática Industrial RIAI
Toda la web
Inicio Revista Iberoamericana de Automática e Informática Industrial RIAI Desarrollo de un Sensor Virtual basado en Modelo NARMAX y Máquina de Vectores d...
Información de la revista
Vol. 11. Núm. 1.
Páginas 109-116 (enero - marzo 2014)
Compartir
Compartir
Descargar PDF
Más opciones de artículo
Visitas
3170
Vol. 11. Núm. 1.
Páginas 109-116 (enero - marzo 2014)
Open Access
Desarrollo de un Sensor Virtual basado en Modelo NARMAX y Máquina de Vectores de Soporte para Molienda Semiautógena
Development of a Software Sensor based on a NARMAX-Support Vector Machine Model for Semi-Autogenous Grinding
Visitas
3170
Gonzalo Acuñaa,
Autor para correspondencia
gonzalo.acuna@usach.cl

Autor para correspondencia.
, Millaray Curilemb, Francisco Cubillosc
a Departamento de Ingeniería Informática, Universidad de Santiago de Chile, USACH, Av. Ecuador 3659, Santiago, Chile
b Departamento de Ingeniería Eléctrica, Universidad de la Frontera, UFRO, Av. Francisco Salazar 01146, Temuco, Chile
c Departamento de Ingeniería Química, Universidad de Santiago de Chile, USACH, Av. Ecuador 3659, Santiago, Chile
Este artículo ha recibido

Under a Creative Commons license
Información del artículo
Resumen
Texto completo
Bibliografía
Descargar PDF
Estadísticas
Resumen

La estimación de estados, en procesos complejos como el proceso de molienda semiautógena (SAG) en la minería del cobre, es una tarea difícil debido a las dificultades para medir directamente ciertas variables relevantes en línea y tiempo real. En este trabajo se amplía una comparación, iniciada en trabajos anteriores de estos mismos autores, entre modelos dinámicos NARX y NARMAX construidos con el uso de Redes Neuronales Artificiales (RNA) y Máquinas de Vectores de Soporte (SVM), cuando actúan como estimadores de una de las variables de estado más importantes para la operación de molienda SAG. Para lograr esta comparación se propone una metodología simple y original para desarrollar modelos NARMAX confeccionados con SVM. Los resultados muestran la potencia predictiva de los modelos NARMAX, que incorporan los errores de predicción en tiempos anteriores para predecir la evolución futura del proceso y la ventaja de aquellos elaborados mediante SVM por sobre los confeccionados con RNA. NARMAX-SVM presenta un MSE significativamente inferior al de todos los otros modelos. En términos del proceso de molienda, se proporciona una herramienta útil para la estimación en línea y tiempo real de una variable que permite controlar y optimizar el proceso y que no puede ser medida mediante instrumentos fácilmente disponibles.

Palabras clave:
Redes Neuronales Artificiales
Máquinas de Vectores de Soporte
NARX
NARMAX
Proceso de Molienda
Sensor Virtual.
Abstract

State estimation in complex processes such as the semi- autogenous grinding process (SAG) in copper mining is an important and difficult task due to difficulties for real-time and on-line measuring of some relevant process variables. This paper extends a comparison, initiated in previous work of the same authors, between NARX and NARMAX dynamic models built using Artificial Neural Networks (ANN) and Support Vector Machines (SVM), when acting as estimators of one of the most important state variables for SAG milling operation. To accomplish this comparison we propose a simple and original methodology to develop NARMAX models with SVM. The results show that SVM-NARMAX models outperform SVM- NARX models because they incorporate previous prediction errors in order to improve prediction of the future evolution of the process. Advantages of SVM over those RNA models are also highlighted. NARMAX-SVM has a significantly lower MSE than all other models. In terms of the milling process, it provides a useful tool for estimating important state variables that are not easily available on-line and in real time thus aiding control and monitoring of the process.

Keywords:
Artificial Neural Network
Support Vector Machine
NARX
NARMAX
Grinding Process
Software Sensor
Referencias
[Acuña and Curilem, 2009]
G. Acuña, M. Curilem.
Comparison of neural networks and support vector machine dynamic models for state estimation in semiautogeneous mills,.
Lecture Notes in Computer Science LNCS, 5845 (2009), pp. 478-487
[Canu et al., 2005]
Canu, S., Grandvalet, Y., Guigue, V., Rakotomamonjy, A, 2005, SVM and Kernel Methods Matlab Toolbox. Perception Systèmes et Information, INSA de Rouen, Rouen, France.
[Chai et al., 2005]
M.-I. Chai, S. Song, S. Li, N-n.
A review of some main improved models for neural network forecasting in time series.
In: Proceedings of the IEEE Intelligent Vehicles Symposium, (2005), pp. 866-868
[Curilem et al., 2011]
Curilem, M., Acuña, G., Cubillos, F. and Vhymeister, E, 2011, Neural networks and support vector machine models applied to energy consumption optimization in semiautogenous grinding, Chemical Engineering Transactions, 25: 761-766, Dot: 10 3303/CET1125127.
[Frohlich and Zell, 2005]
H. Frohlich, A. Zell.
Efficient parameter selection for support vector machines in classification and regression via model-based global optimization.
In: Proceedings of the IEEE International Joint Conference on Neural Networks, IJCNN 2005., (2005), pp. 1431-1436
[Gao and Joo Er, 2005]
Y. Gao, M. Joo Er.
NARMAX time series model and prediction: feedforward and recurrent fuzzy neural network approaches.
Fuzzy Sets and Systems, 150 (2005), pp. 331-350
[Gonzaga et al., 2009]
J. Gonzaga, L. Meleiro, C. Kiang, R. Maciel Filho.
ANN-based soft- sensor for real-time process monitoring and control of an industrial polymerization process.
Computers and Chemical Engineering, 33 (2009), pp. 43-49
[Guo et al., 2008]
Guo, G., Wu, X., Zhuo, S., Xu, P., Cao, B. 2008, Prediction state of charge of Ni-MH battery pack using support vector machines for hybrid electric vehicles, IEEE Vehicle Power and Propulsion Conference (VPPC), September3-5, 2008, Harbin, China.
[Hornstein and Parlitz, 2004]
A. Hornstein, U. Parlitz.
Bias Reduction for Time Series Models Based on Support Vector Regression.
International Journal of Bifurcation and Chaos, 14 (2004), pp. 1947-1956
[Leontaritis and Billings, 1985]
I.J. Leontaritis, S.A. Billings.
Input-Output Parametric Models for Non-Linear Systems; Part 1: Deterministic Non-Linear Systems; Part 2: Stochastic Non-Linear Systems.
International Journal of Control, 45 (1985), pp. 303-344
[Ljung, 1987]
L. Ljung.
System Identification: Theory for the user.
Prentice-Hall, (1987),
[Magne et al., 1997]
L. Magne, W. Valderrama, J. Pontt.
Conceptual Vision and State of the Semiautogenous Mill Technology. In: Revista Minerales.
Instituto de Ingenieros de Minas de Chile., 52 (1997),
[Martínez-Ramón et al., 2006]
M. Martínez-Ramón, J. Rojo-Alvarez, G. Camps-Valls, J. Muñoz-Marí, A. Navia-Vásquez, E. Soria-Olivas, A. Figueiras-Vidal.
Support vector machines for nonlinear kernel ARMA system identification.
IEEE Transactions on Neural Networks, 17 (2006), pp. 1617-1622
[Norgaard, 2003]
Norgaard, M., 2003, The NNSYSID Toolbox, for use with Matlab, http://www.iau.dtu.dk/research/control/nnsysid.html.
[Salazar et al., 2009]
J. Salazar, L. Magne, G. Acuña, F. Cubillos.
Dynamic modelling and simulation of semi-autogenous mills.
Minerals Engineering, 22 (2009), pp. 70-77
[Sapankevych and Sankar, 2009]
N. Sapankevych, R. Sankar.
Time series prediction using support vector machines: a survey.
IEEE Computational Intelligence Magazine, (2009), pp. 24-38
[Schölkopf et al., 2000]
B. Schölkopf, A. Smola, R.C. Williamson, P.L. Bartlett.
New support vector algorithms.
Neural Computation, 12 (2000), pp. 1083-1121
[Suárez and Gómez, 2011]
A. Suárez, Z. Gómez.
Softsensor trained using the concept of instrumental variables and applied in the measurement of temperature in a Teniente converter.
Revista Iberoamericana de Automática e Informática Industrial, 8 (2011), pp. 54-63
[Suykens et al., 2002]
Suykens, J.A.K., Van Gestel, T, De Brabanter, J., De Moor, B., Vandewalle, J. 2002, Least Squares Support Vector Machines, World Scientific, Singapore.
[Vapnik, 1995]
V. Vapnik.
The Nature of Statistical Learning Theory.
Springer Verlag, (1995),
[Werbos, 1990]
P.J. Werbos.
Backpropagation Through Time: What It Does and How to do it.
Proc. IEEE, 78 (1990), pp. 1150-1560
[Wolmarans and Morgan, 2009]
E. Wolmarans, P. Morgan.
Milling circuit selection for the Nkomati 375 ktpm concentrador.
SIAMM-Journal of the South African Institute of Mining and Metallurgy, 109 (2009), pp. 653-664
[Woolley et al., 2010]
J. Woolley, P. Agarwal, J. Baker.
Modeling and prediction of chaotic systems with artificial neural networks.
Internacional Journal for Numerical Methods in Fluids, 63 (2010), pp. 989-1004
[Yan et al., 2004]
W. Yan, H. Shao, X. Wang.
Soft sensing modeling based on support vector machine and Bayesian model selection.
Computers and Chemical Engineering, 26 (2004), pp. 1489-1498
Copyright © 2011. EA
Opciones de artículo