摘要
分析了对转炉终点温度的影响因素,利用减法聚类自动确定模糊规则的数目,建立了模糊神经网络系统预报转炉终点温度.结果表明,该方法建立的模型能够对终点温度进行较好的预报,误差在±4℃以内的命中率可达25.49%;预报误差小于±20℃的炉数可达84.31%.
The primary influence factors for end-point temperature of BOF were analyzed. Subtractive clustering algorithm was adopted to determine the number of fuzzy rules, and then a prediction model based on fuzzy neural network for the end-point temperature （EPT） of BOF was established. The results showed that this model could predict EPT efficiently, with the hit rate of 25.49% in error range of ± 4℃, and 84.31% in±20℃.
出处
《材料与冶金学报》
CAS
2006年第4期247-249,共3页
Journal of Materials and Metallurgy
关键词
转炉
减法聚类
模糊神经网络
终点预报
温度
BOF
subtracfive clustering
fuzzy neural network
end - point prediction
temperature
