THE PREDICTION OF THE MICROSTRUCTURE CONSTITUENTS OF SPHEROIDAL GRAPHITE CAST IRON BY USING THERMAL ANALYSIS AND ARTIFICIAL NEURAL NETWORKS

被引:0
|
作者
Glavas, Z. [1 ]
Unkic, F. [1 ]
Lisjak, D. [2 ]
机构
[1] Univ Zagreb, Fac Met, Sisak 44103, Croatia
[2] Univ Zagreb, Fac Mech Engn & Naval Architecture, Zagreb 10002, Croatia
关键词
spheroidal graphite cast iron; microstructure constituents; thermal analysis; artificial neural networks; EUTECTOID TRANSFORMATION; AS-CAST; DILATATION ANALYSIS;
D O I
暂无
中图分类号
TF [冶金工业];
学科分类号
0806 ;
摘要
This paper presents the application of articial neural networks in the production process of spheroidal graphite cast iron. Backpropagation neural networks have been established to predict the microstructure constituents (ferrite content, pearlite conent, nodule count and nodularity) of speroidal geaphite cast iron using the thermal analysis parameters as inputs. Generalization properties of the developed artificial neural netyworks are very good, which id=s confirmed by a very good accordance between the predicted and the targeted values of the microstructure constituents on a new data that was not included in the training data set.
引用
收藏
页码:247 / 253
页数:7
相关论文
共 50 条