Predictions of heat transfer coefficients of supercritical carbon dioxide using the overlapped type of local neural network

Abstract An overlapped type of local neural network is proposed to improve accuracy of the heat transfer coefficient estimation of the supercritical carbon dioxide. The idea of this work is to use the network to estimate the heat transfer coefficient for which there is no accurate correlation model due to the complexity of the thermo-physical properties involved around the critical region. Unlike the global approximation network (e.g. backpropagation network) and the local approximation network (e.g. the radial basis function network), the proposed network allows us to match the quick changes in the near-critical local region where the rate of heat transfer is significantly increased and to construct the global smooth perspective far away from that local region. Based on the experimental data for carbon dioxide flowing inside a heated tube at the supercritical condition, the proposed network significantly outperformed some the conventional correlation method and the traditional network models.

[1]  Jules Thibault,et al.  A neural network methodology for heat transfer data analysis , 1991 .

[2]  S. Ashforth-Frost,et al.  Evaluating convective heat transfer coefficients using neural networks , 1996 .

[3]  L. Piazza,et al.  Analysis of forced convection heat transfer to supercritical carbon dioxide inside tubes using neural networks , 2003 .

[4]  Mihir Sen,et al.  Simulation of heat exchanger performance by artificial neural networks , 1999 .

[5]  Shang-Liang Chen,et al.  Orthogonal least squares learning algorithm for radial basis function networks , 1991, IEEE Trans. Neural Networks.

[6]  D. D. Bruns,et al.  WaveARX neural network development for system identification using a systematic design synthesis , 1995 .

[7]  Rodney L. McClain,et al.  Neural network analysis of fin-tube refrigerating heat exchanger with limited experimental data , 2001 .

[8]  Tanaka Hiroaki,et al.  Forced convection heat transfer to fluid near critical point flowing in circular tube , 1971 .

[9]  Luigi Piroddi,et al.  Nonlinear identification and control of a heat exchanger: a neural network approach , 1997 .

[10]  Federico Girosi,et al.  On the Relationship between Generalization Error, Hypothesis Complexity, and Sample Complexity for Radial Basis Functions , 1996, Neural Computation.

[11]  M. Mukhopadhyay.  Natural extracts using supercritical carbon dioxide , 2000 .

[12]  Tony Clifford,et al.  Fundamentals of Supercritical Fluids , 1999 .

[13]  M. Nazmul Karim,et al.  Model-predictive pH control using real-time NARX approach , 1994 .

[14]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[15]  Eckhard A. Groll,et al.  Heat transfer from supercritical carbon dioxide in tube flow: A critical review , 1998 .