Title :
The Prediction of Sn-Ag Solder Properties Based on BP Algorithm of Artificial Neutral Network
Author :
Sun, Li ; Qi, Fangjuan ; Zhezhe Hon ; Xiao, Yuehua
Author_Institution :
Shijiazhuang Railway Inst., Shijiazhuang
Abstract :
In this paper, the prediction of properties of the lead-free solders were focused on by using the BP neural network. Then different algorithm and parameter of BP neural network has been used in the training process, and the results have been analyzed for obtaining the optimum algorithm and parameter. The influence of adding In. Bi. Sb. RE Cu to Sn- Ag alloy to tensile strength, shear strength and the solidification temperature of Sn-Ag alloy was discussed in this paper. So the input variables are the contents of In. Bi, Sb. RE. Sn. Ag. Cu. and the output variable the tensile strength, shear strength and the solidification temperature of Sn-Ag alloy respectively. 15 groups of data were chosen for training of the model and 3 groups of data for prediction. The results show that the predicted data are good agreement with ones for the given experimental conditions. The accuracy and applicability for the present neural network modeling are thus verified.
Keywords :
backpropagation; electronic engineering computing; shear strength; silver alloys; solders; solidification; tensile strength; tin alloys; BP neural network; SnAg; artificial neutral network; lead-free solders properties; shear strength; solidification temperature; tensile strength; Algorithm design and analysis; Artificial neural networks; Bismuth; Copper alloys; Environmentally friendly manufacturing techniques; Input variables; Lead; Solid modeling; Temperature; Tin alloys;
Conference_Titel :
Electronic Packaging Technology, 2007. ICEPT 2007. 8th International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4244-1392-8
Electronic_ISBN :
978-1-4244-1392-8
DOI :
10.1109/ICEPT.2007.4441529
