双层自校正预报方法及其在高炉铁水含硅量预报中的应用
A TWO-LEVEL SELF-TUNING PREDICTION METHOD AND ITS APPLICATION TO Si-CONTENT PREDICTION OF MOLTEN IRON IN BLAST FURNACE
-
摘要: 针对象高炉这样复杂的MISO系统,本文提出了一种双层自校正预报方法,将快时变部分的递推辨识和慢时变部分的迭代修正相结合,较好地解决了各子模型时间常数大小和时变性快慢相差较大所带来的问题,并在高炉铁水含硅量预报的实际应用中获得了满意的结果.Abstract: A two-level self-tuning prediction method is proposed for a blast furnace,a complex MISO time-varying system.Combining the recursive identification of the faster time-varying part with the iterative correction of the slower part,this method solves the problem of too big disparities in time constants andtime-varying speeds of submodels.Practice shows its effectiveness for predicting Sicontent of molten ironin a blast furnace.