Modelling of a post-combustion CO2 capture process using neural networks

Abstract This paper presents a study of modelling post-combustion CO 2 capture process using bootstrap aggregated neural networks. The neural network models predict CO 2 capture rate and CO 2 capture level using the following variables as model inputs: inlet flue gas flow rate, CO 2 concentration in inlet flue gas, pressure of flue gas, temperature of flue gas, lean solvent flow rate, MEA concentration and temperature of lean solvent. In order to enhance model accuracy and reliability, multiple feedforward neural network models are developed from bootstrap re-sampling replications of the original training data and are combined. Bootstrap aggregated model can offer more accurate predictions than a single neural network, as well as provide model prediction confidence bounds. Simulated CO 2 capture process operation data from gPROMS simulation are used to build and verify neural network models. Both neural network static and dynamic models are developed and they offer accurate predictions on unseen validation data. The developed neural network models can then be used in the optimisation of the CO 2 capture process.

[1]  Chengye Zhao,et al.  Melt index prediction based on fuzzy neural networks and PSO algorithm with online correction strategy , 2012 .

[2]  Eric B. Bartlett,et al.  Process modeling using stacked neural networks , 1996 .

[3]  Erdogan Alper,et al.  Steady-state rate-based modelling for CO2/amine absorption—desorption systems , 1994 .

[4]  H. M. Kvamsdal,et al.  Dynamic modeling and simulation of a CO2 absorber column for post-combustion CO2 capture , 2009 .

[5]  G. Versteeg,et al.  CO2 capture from power plants. Part I: A parametric study of the technical performance based on monoethanolamine , 2007 .

[6]  H. Yeung,et al.  Dynamic modelling and analysis of post-combustion CO2 chemical absorption process for coal-fired power plants , 2010 .

[7]  Jie Zhang,et al.  Combination of multiple neural networks using data fusion techniques for enhanced nonlinear process modelling , 2005, Comput. Chem. Eng..

[8]  Christine W. Chan,et al.  A statistical analysis of the carbon dioxide capture process , 2009 .

[9]  Meihong Wang,et al.  Demonstrating full-scale post-combustion CO2 capture for coal-fired power plants through dynamic modelling and simulation , 2012 .

[10]  D. Marquardt.  An Algorithm for Least-Squares Estimation of Nonlinear Parameters , 1963 .

[11]  Qing Zhou,et al.  Applications of three data analysis techniques for modeling the carbon dioxide capture process , 2010, CCECE 2010.

[12]  Jie Zhang,et al.  Developing Robust Neural Network Models by Using Both Dynamic and Static Process Operating Data , 2001 .

[13]  L. Cooper,et al.  When Networks Disagree: Ensemble Methods for Hybrid Neural Networks , 1992 .

[14]  Rich Caruana,et al.  Overfitting in Neural Nets: Backpropagation, Conjugate Gradient, and Early Stopping , 2000, NIPS.

[15]  Axel Meisen,et al.  Simulation of pilot plant and industrial CO2-MEA absorbers , 1993 .

[16]  Jie Zhang,et al.  Bayesian selective combination of multiple neural networks for improving long-range predictions in nonlinear process modelling , 2004, Neural Computing & Applications.

[17]  Jie Zhang,et al.  A Reliable Neural Network Model Based Optimal Control Strategy for a Batch Polymerization Reactor , 2004 .

[18]  Jie Zhang,et al.  Developing robust non-linear models through bootstrap aggregated neural networks , 1999, Neurocomputing.

[19]  R. Clemen.  Combining forecasts: A review and annotated bibliography , 1989 .

[20]  B. Efron.  The jackknife, the bootstrap, and other resampling plans , 1987 .

[21]  Meihong Wang,et al.  Post-combustion CO2 capture with chemical absorption: A state-of-the-art review , 2011 .

[22]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[23]  Huaqin Jiang,et al.  Melt index prediction by adaptively aggregated RBF neural networks trained with novel ACO algorithm , 2012 .

[24]  Eugeny Y. Kenig,et al.  Reactive absorption: Optimal process design via optimal modelling , 2001 .

[25]  Nikolett Sipöcz,et al.  The use of Artificial Neural Network models for CO2 capture plants , 2011 .

[26]  Gary T. Rochelle,et al.  Dynamic Modeling to Minimize Energy Use for CO2 Capture in Power Plants by Aqueous Monoethanolamine , 2009 .

[27]  David H. Wolpert,et al.  Stacked generalization , 1992, Neural Networks.

[28]  Heekuck Oh,et al.  Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[29]  H. Yeung,et al.  Dynamic modelling of CO 2 absorption for post combustion capture in coal-fired power plants 1 , 2008 .

[30]  Sergio Mussati,et al.  CO2 capture using monoethanolamine (MEA) aqueous solution: Modeling and optimization of the solvent regeneration and CO2 desorption process , 2012 .

[31]  Chris Bishop,et al.  Improving the Generalization Properties of Radial Basis Function Neural Networks , 1991, Neural Computation.

[32]  Ting Wang,et al.  Melt index prediction by aggregated RBF neural networks trained with chaotic theory , 2014, Neurocomputing.

[33]  David J. C. MacKay,et al.  Bayesian Interpolation , 1992, Neural Computation.

[34]  C. Kiparissides,et al.  Inferential Estimation of Polymer Quality Using Stacked Neural Networks , 1997 .

[35]  Jie Zhang,et al.  Recurrent neuro-fuzzy networks for nonlinear process modeling , 1999, IEEE Trans. Neural Networks.

[36]  Meihong Wang,et al.  Dynamic modelling of CO2 absorption for post combustion capture in coal-fired power plants , 2009 .