Artificial neural network modelling of hydrogen storage properties of Mg-based alloys

Abstract An artificial neural network model has been created for prediction of the hydrogen storage capacity and the temperature and pressure of dehydrogenation of Mg-based alloys as a function of alloy composition. The effects of 24 chemical elements are considered in the model, which is based on a two-layer feedforward hierarchical architecture. The neural network was trained using the Levenberg–Marquardt training algorithm in combination with Bayesian regularization. The model was used to study the influence of the alloying elements on the hydrogen storage properties of MgH 2 . For almost all of the investigated alloying elements, increasing their content results in a decrease of the hydrogen storage capacity, but several elements lead to a reduction of the temperature for hydrogen desorption. A graphical user interface (GUI) has been established for the prediction of the hydrogen storage capacity, temperature and pressure of dehydrogenation for magnesium alloys as function of their chemical composition, as well as for investigation the influence of the different alloying elements on the hydrogen storage properties in magnesium alloys.

[1]  J. Mathieu,et al.  Thermodynamic and structural properties of LaNi5−yAly compounds and their related hydrides , 1979 .

[2]  Djc MacKay,et al.  Neural network analysis of steel plate processing , 1998 .

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

[4]  S. Suda,et al.  A study of the surface composition and structure of fluorinated Mg-based alloys , 1995 .

[5]  J. L. Rana,et al.  Modeling of material behavior data in a functional form suitable for neural network representation , 1999 .

[6]  D. Mackay,et al.  Bayesian Neural Network Analysis of Fatigue Crack Growth Rate in Nickel Base Superalloys , 1996 .

[7]  Martin T. Hagan,et al.  Neural network design , 1995 .

[8]  L. A. Dobrzański,et al.  Application of a neural network in modelling of hardenability of constructional steels , 1998 .

[9]  H. K. D. H. Bhadeshia,et al.  Estimation of the γ and γ' lattice parameters in nickel-base superalloys using neural network analysis , 1998 .

[10]  Wei Sha,et al.  Modelling the correlation between processing parameters and properties in titanium alloys using artificial neural network , 2000 .

[11]  E. L. Huston,et al.  Engineering properties of metal hydrides , 1980 .

[12]  Richard P. Lippmann,et al.  An introduction to computing with neural nets , 1987 .

[13]  A. Pedersen,et al.  The formation of hydride in pure magnesium foils , 1987 .

[14]  G. Sandrock,et al.  The IEA/DOE/SNL on-line hydride databases , 2001 .

[15]  Chen Bououdina,et al.  Carrying Clean Energy to the Future – Hydrogen Absorbing Materials , 2000 .

[16]  Sybrand van der Zwaag,et al.  Effects of Carbon Concentration and Cooling Rate on Continuous Cooling Transformations Predicted by Artificial Neural Network , 1999 .

[17]  H. K. D. H. Bhadeshia,et al.  Neural Networks in Materials Science , 1999 .

[18]  Mohammad Bagher Menhaj,et al.  Training feedforward networks with the Marquardt algorithm , 1994, IEEE Trans. Neural Networks.

[19]  H. K. D. H. Bhadeshia,et al.  The yield and ultimate tensile strength of steel welds , 1997 .

[20]  A. P. de Weijer,et al.  Prediction of jominy hardness profiles of steels using artificial neural networks , 1996 .

[21]  M. Pezat,et al.  The Mg2Ni0.75M0.25 alloys (M = 3d element): Their application to hydrogen storage , 1983 .

[22]  Franck Tancret,et al.  Comparison of artificial neural networks with gaussian processes to model the yield strength of nickel-base superalloys , 1999 .

[23]  D. J. C. Mackay,et al.  Estimation of Hot Torsion Stress Strain Curves in Iron Alloys Using a Neural Network Analysis , 1999 .