Neural Network Analysis of Quantitative Relationship Between Alloying Component and Initial Discharge Capacity of AB5-Based Hydrogen Storage Alloy
HE Wen-tao1，LIU Shu-mei1，YOU Wei2，LIU Ya-hui1
1.College of Materials Engineering，Shanghai University of Engineering Science，Shanghai 201620，China 2. College of Mechanical and Electrical Engineering，North China Institute of Science and Technology，Sanhe 065201，Hebei，China
Abstract：Radial-basis function artificial neural network (ANN) was developed to study the quantitative relationship between the alloying component and initial discharge capacity of AB5-based hydrogen storage alloy. “Leave-one out” method was used to train the ANN model. After being trained, ANN model was used to predict the initial discharge capacity of five samples,the prediction and the measured values distribute along the 45° diagonal line in the scatter diagram, statistical indicators were MSE=6.063, MRSE=0.0262%, VOF=1.9345, which showed that the prediction results of ANN model were accurate and reliable. The quantitative effects of alloying component on the initial discharge capacity were analyzed by ANN model. The results show that parabola relationship exists between the initial discharge capacity and La and Nd content. Ce content has a great effect on the initial discharge capacity which increases with the increase of Ce content. With the increase of Pr content, the initial discharge capacity increases slightly and leveled off finally.
Chen TP, Chen H, Liu R W. Approximation capability in C() by multilayer feedforward networks and related problems[J]. IEEE Trans NN, 1995, 6(1): 25-30.
[6]
Hecht-Nielsen. Neurocomputing[M]. Massachussetes: Addison Wesley Publishing Co Inc, 1991: 59.
[1]
Willems J J G. Metal Hydride Electrodes, Stability of LaNi5 related compounds[ J]. Philips J Res, 1984, 39(1):1.
[9]
Guo J, Li C H, Liu H L, et al. Effect of rare earth compositions on the Hydrogen storage properties of AB5-type alloys and pattern recognition analysis[J]. J Electrochem Soc, 1997, 144 (7): 2276.
[11]
Narayan V, Abad R, Bhadeshia H K D H. Estimation of hot torsion stress strain curves in iron alloys using neural network analysis[J]. ISIJ International, 1999, 39 (10): 999-1005.
Sung-Sau So, Martin Karplus. Evolutionary optimization in quantitative structure-activity relationship: an application of genetic neural networks[J]. Journal of Medicinal Chemistry, 1996, 39 (7): 1524.
[14]
Scarselli F, Tsoi A C. Universal approximation using feedforward neural networks：a survey of some existing methods, and some new results[J]. Neural Networks, 1998, 11(1): 15-37.
2014, Vol. 21 
