摘要
本文试用人工神经网络总结合金钢研制数据中的规律，并应用偏相关指数表征各影响因子对合金钢性能影响的程度。计算结果表明，ＩＦ钢的延伸率、耐低温钢的无塑性变形温度预报值与实测值相当符合。根据偏相关指数绝对值的大小判别影响耐深冲钢合格率的主要因素效果亦佳。
Artificial neural network has been used to investigate the regularities of experimental data in new alloy steel exploration works, and the partial correlation index has been used to find the relative importance of various factors affecting the properties of alloy steel samples. The results of computation indicate that the elongation of IF steel, the temperature of zero plastic deformation of low-temperature steel predicted are in agreement with the experimental data. The relative importance of factors influencing the quality of ST14 steel are found, and the dominating factors found are useful for the industrial optimization of steel production.
出处
《计算机与应用化学》
CAS
CSCD
1994年第3期225-227,共3页
Computers and Applied Chemistry
关键词
神经网络
合金钢
研制
IF钢
Artificial neural network, Partial correlation index, Alloy steel
