For predicting the key technology index of electroslag remelting (ESR) process (the melting rate and cone purification coefficient of the consumable electrode), a radial basis function (RBF) neural network soft-sensor model optimized by the artificial fish swarm algorithm (AFSA) is proposed. Based on the technique characteristics of ESR production process, the auxiliary variables of soft-sensor model are selected. Then the AFSA is adopted to train the RBF neural network prediction model in order to realize the nonlinear mapping between input and output variables. Simulation results show that the model has better generalization and prediction accuracy, which can meet the online soft sensing requirement of ESR process real-time control. 1. Introduction Electroslag remelting (ESR) process is an advanced smelting method to make purified steels based on rudiment steel in order to reduce impurity and get the high-quality steel which is uniformity, density, and crystal in vertical [1]. The main purpose of ESR process is to purify metal and get the ingot with uniform density crystallization. The steel undergoing the ESR process has many advantages, such as high-purity, lower sulfur and inclusion of nonmetal, smooth surface of the ingot, the uniform density crystallization, and the uniform metal structure and chemical composition. The metal bars as consumable electrodes are inserted into the liquid slag. The current goes through the consumable electrodes and the slag resistance heat appears in the slag pool to melt the metallic electrode to produce metallic droplets. Then the metallic droplets undergo the physical and chemical reaction by the way of dripping in the slag pool and are cooled and recrystallized in the compendium. The quality of the electricity slag ingots depends on the proper ESR technique and the effective control methods. From the control point of view, ESR process is a typical complex controlled object, which has multivariable, distributed parameters and nonlinear and strong coupling features. During the early stage of research, the normal control methods include voltage swings, constant current, constant voltage, and descending power. The voltage swing control method is adopted to adjust the voltage swing amplitude in order to control the electrode movements and maintain the stability of slag resistance [2]. The key technology of melting speed control system and the mathematical model of the time-variant system is discussed, but not analyzing the fundamental reasons causing voltage fluctuation at the expense of the current control accuracy in
References
[1]
F. Cao, W. Wang, and J. Wang, “Intelligent optimal setting control of electroslag remelting process based on knowledge mining and case-based reasoning,” Information and Control, vol. 39, no. 1, pp. 114–119, 2010.
[2]
L. Zhang, J. Wang, and J. Wang, “Analysis and application of two principles for ESR furnace swing control,” Metallurgical Industry Automation, vol. 2, pp. 53–55, 2006.
[3]
W.-Z. Li, X.-H. Jia, J.-C. Wang, and Y.-S. Lin, “Improvement of consarc ESR furnace control system,” Journal of Iron and Steel Research, vol. 19, no. 4, pp. 99–106, 2007.
[4]
L. Zhao, J. Song, X. Liu, and H. Liu, “Intelligent control study based on genetic algorithm for electroslag remelting process,” Machinery & Electronics, vol. 5, pp. 16–19, 2008.
[5]
D. Wang, G. He, W. Song, Y. Kang, and J. Liu, “Study and application on intelligent control of the ESR process,” Industrial Heating, vol. 34, no. 6, pp. 42–45, 2005.
[6]
W. Ren, X. Zheng, L. Jiang, J. Hu, and Y. Yin, “Fuzzy adaptive PID control for electrode regulating system of electroslag furnace,” Metallurgical Industry Automation, vol. 1, pp. 15–18, 2006.
[7]
F. Cao and W. Wang, “Harmony search based particle swarm optimisation approach for optimal PID control in electroslag remelting process,” International Journal of Modelling, Identification and Control, vol. 15, no. 1, pp. 20–27, 2012.
[8]
S. Ahn, J. J. Beaman, R. L. Williamson, and D. K. Melgaard, “Model-based control of electroslag remelting process using unscented Kalman filter,” Journal of Dynamic Systems, Measurement and Control, vol. 132, no. 1, pp. 1–9, 2010.
[9]
Y. Ma, D. Huang, and Y. Jin, “Discussion about dynamic soft-sensing modeling,” Journal of Chemical Industry and Engineering, vol. 56, no. 8, pp. 1516–1519, 2005.
[10]
Y. G. Li, W. H. Gui, C. H. Yang, and Y. Xie, “Soft sensor and expert control for blending and digestion process in alumina metallurgical industry,” Journal of Process Control, vol. 23, no. 7, pp. 1012–1021, 2013.
[11]
K. Fujiwara, M. Kano, and S. Hasebe, “Development of correlation-based pattern recognition algorithm and adaptive soft-sensor design,” Control Engineering Practice, vol. 20, no. 4, pp. 371–378, 2012.
[12]
B. Li, F. Wang, and M. Shan, “Numerical analysis of electromagnetic field in an electroslag remelting process with three-phases electrodes,” in CFD Modeling and Simulation in Materials Processing, pp. 131–138, John Wiley & Sons, Hoboken, NJ, USA, 2012.
[13]
V. Weber, A. Jardy, B. Dussoubs et al., “A comprehensive model of the electroslag remelting process: description and validation,” Metallurgical and Materials Transactions B, vol. 40, no. 3, pp. 271–280, 2009.
[14]
J. Wang, L. Liu, Y. Yang, and G. Liu, “Research on electroslag remelting technology and intelligent control methods,” in Proceedings of the 25th Chinese Control and Decision Conference (CCDC '13), pp. 3374–3379, IEEE, 2013.
[15]
S. A. Iliyas, M. Elshafei, M. A. Habib, and A. A. Adeniran, “RBF neural network inferential sensor for process emission monitoring,” Control Engineering Practice, vol. 21, no. 7, pp. 962–970, 2013.
[16]
F. Fernández-Navarro, C. Hervás-Martínez, R. Ruiz, and J. C. Riquelme, “Evolutionary Generalized Radial Basis Function neural networks for improving prediction accuracy in gene classification using feature selection,” Applied Soft Computing Journal, vol. 12, no. 6, pp. 1787–1800, 2012.
[17]
S. Gholizadeh, E. Salajegheh, and P. Torkzadeh, “Structural optimization with frequency constraints by genetic algorithm using wavelet radial basis function neural network,” Journal of Sound and Vibration, vol. 312, no. 1-2, pp. 316–331, 2008.
[18]
S.-K. Oh, W.-D. Kim, W. Pedrycz, and B.-J. Park, “Polynomial-based radial basis function neural networks (P-RBF NNs) realized with the aid of particle swarm optimization,” Fuzzy Sets and Systems, vol. 163, no. 1, pp. 54–77, 2011.
[19]
S.-K. Oh, W.-D. Kim, W. Pedrycz, and S.-C. Joo, “Design of K-means clustering-based polynomial radial basis function neural networks (pRBF NNs) realized with the aid of particle swarm optimization and differential evolution,” Neurocomputing, vol. 78, no. 1, pp. 121–132, 2012.
[20]
X.-L. Li, Z.-J. Shao, and J.-X. Qian, “Optimizing method based on autonomous animats: fish-swarm algorithm,” System Engineering Theory and Practice, vol. 22, no. 11, pp. 32–38, 2002.
[21]
X. Li and J. Qian, “Studies on artificial fish swarm optimization algorithm based on decomposition coordination techniques,” Journal of Circuits and Systems, vol. 8, no. 1, pp. 1–6, 2003.
[22]
H.-C. Tsai and Y.-H. Lin, “Modification of the fish swarm algorithm with particle swarm optimization formulation and communication behavior,” Applied Soft Computing Journal, vol. 11, no. 8, pp. 5367–5374, 2011.
