Ill-posed Echo State Network based on L-curve Method for Prediction of Blast Furnace Gas Flow

Blast furnace gas is a significant energy resource in steel industry. Keeping stable of blast furnace gas flow is an important task for the furnace itself and the application of byproduct gas. However, owing to fluctuation and noisy of gas flow, echo state network is usually ill-posed in the prediction and it is very difficult to accurately predict the amount of gas. In this paper, in order to increase the accuracy of prediction in ill-posed echo state network model, L-curve method is used to compute the regularization parameter, which can alleviate the influence of ill-condition for ESN. Finally, to verify the effectiveness of the proposed method, the real data from blast furnace is employed in the experiments. Compared with two parameter regularization methods and four types of prediction methods, the results demonstrate that the proposed method exhibits a higher prediction accuracy for gas prediction.

[1]  Min Han,et al.  Predicting Multivariate Time Series Using Subspace Echo State Network , 2013, Neural Processing Letters.

[2]  Min Han,et al.  gamma-C plane and robustness in static reservoir for nonlinear regression estimation , 2009, Neurocomputing.

[3]  Peter Fogh Odgaard,et al.  Estimation of Uncertainty Bounds for the Future Performance of a Power Plant , 2009, IEEE Transactions on Control Systems Technology.

[4]  ProblemsPer Christian HansenDepartment.  The L-curve and its use in the numerical treatment of inverse problems , 2000 .

[5]  Helmut Hauser,et al.  Echo state networks with filter neurons and a delay&sum readout , 2010, Neural Networks.

[6]  Harald Haas,et al.  Harnessing Nonlinearity: Predicting Chaotic Systems and Saving Energy in Wireless Communication , 2004, Science.

[7]  Min Han,et al.  Support Vector Echo-State Machine for Chaotic Time-Series Prediction , 2007, IEEE Transactions on Neural Networks.

[8]  David L. Phillips,et al.  A Technique for the Numerical Solution of Certain Integral Equations of the First Kind , 1962, JACM.

[9]  P. Hansen.  The truncatedSVD as a method for regularization , 1987 .

[10]  C. Lawson,et al.  Solving least squares problems , 1976, Classics in applied mathematics.

[11]  Charles L. Lawson,et al.  Solving least squares problems , 1976, Classics in applied mathematics.

[12]  Yoshikazu Nishikawa,et al.  An optimal gas supply for a power plant using a mixed integer programming model , 1991, Autom..

[13]  Wu Yi.  Improved Echo State Network Based on Data-driven and Its Application to Prediction of Blast Furnace Gas Output , 2009 .

[14]  Gholam Ali Montazer,et al.  An improvement in RBF learning algorithm based on PSO for real time applications , 2013, Neurocomputing.

[15]  Michael A. Saunders,et al.  LSQR: An Algorithm for Sparse Linear Equations and Sparse Least Squares , 1982, TOMS.

[16]  Herbert Jaeger,et al.  The''echo state''approach to analysing and training recurrent neural networks , 2001 .

[17]  Garrison W. Cottrell,et al.  2007 Special Issue: Learning grammatical structure with Echo State Networks , 2007 .

[18]  S. F. Ghaderi,et al.  Integration of Artificial Neural Networks and Genetic Algorithm to Predict Electrical Energy consumption , 2006, IECON 2006 - 32nd Annual Conference on IEEE Industrial Electronics.

[19]  Herbert Jaeger,et al.  Reservoir computing approaches to recurrent neural network training , 2009, Comput. Sci. Rev..

[20]  Peter Tiño,et al.  Predictive Modeling with Echo State Networks , 2008, ICANN.

[21]  Chao Wu,et al.  Forecasting stock indices using radial basis function neural networks optimized by artificial fish swarm algorithm , 2011, Knowl. Based Syst..

[22]  K. V. Sudhakar,et al.  ANN back-propagation prediction model for fracture toughness in microalloy steel , 2002 .

[23]  Vicente Alarcón Aquino,et al.  A Neural Network Scheme for Long-Term Forecasting of Chaotic Time Series , 2011, Neural Processing Letters.

[24]  Jun Zhao,et al.  Improved Echo State Network Based on Data-driven and Its Application to vskip0.3 baselineskip Prediction of Blast Furnace Gas Output: Improved Echo State Network Based on Data-driven and Its Application to vskip0.3 baselineskip Prediction of Blast Furnace Gas Output , 2009 .

[25]  G. S. Mahapatra,et al.  Robust feedforward and recurrent neural network based dynamic weighted combination models for software reliability prediction , 2014, Appl. Soft Comput..

[26]  Tao Wang,et al.  A hybrid optimization-based recurrent neural network for real-time data prediction , 2013, Neurocomputing.

[27]  B. Hofmann.  Regularization for Applied Inverse and III-Posed Problems , 1986 .

[28]  Chonghun Han,et al.  A Novel MILP Model for Plantwide Multiperiod Optimization of Byproduct Gas Supply System in the Iron- and Steel-Making Process , 2003 .

[29]  Gene H. Golub,et al.  Generalized cross-validation as a method for choosing a good ridge parameter , 1979, Milestones in Matrix Computation.

[30]  G. Venayagamoorthy,et al.  Neural networks letter Effects of spectral radius and settling time in the performance of echo state networks , 2009 .

[31]  Wataru Yashiro,et al.  Solving Ill-Posed Linear Systems With Constraints on Statistical Moments , 2012, IEEE Signal Processing Letters.

[32]  Mengjie Zhang,et al.  Cooperative coevolution of Elman recurrent neural networks for chaotic time series prediction , 2012, Neurocomputing.

[33]  Kyoung-jae Kim,et al.  Financial time series forecasting using support vector machines , 2003, Neurocomputing.

[34]  O. Lepskii.  On a Problem of Adaptive Estimation in Gaussian White Noise , 1991 .

[35]  Ganesh K. Venayagamoorthy,et al.  Online design of an echo state network based wide area monitor for a multimachine power system , 2007, Neural Networks.

[36]  G. Wahba.  A Comparison of GCV and GML for Choosing the Smoothing Parameter in the Generalized Spline Smoothing Problem , 1985 .