摘要
开发了较完善的 VD炉终点钢液温度在线预报系统 ,应用神经网络方法对 VD炉终点钢液温度进行预报 ,系统在线连续预报了 76炉次 ,预报温度与实际测量温度之差在± 4℃、± 5℃、± 6℃之内的炉次分别占 67.1%、80 .3 %、89.5 %。
A neural network prediction system of molten steel temperature for VD has been developed. 76 heats were predicted by this system. The results were:the hitting rate within a tolerance of ±4 ℃,±5 ℃ and ±6 ℃ is 67.1 %,80.3 % and 89.5 % respectively. The accuracy is analyzed.
出处
《钢铁》
CAS
CSCD
北大核心
2003年第1期16-18,共3页
Iron and Steel
