论文部分内容阅读
根据目前高炉炉温预报推理规则都是由高炉专家根据经验制定的情况,提出了一种新的规则生成方法——数据挖掘获取高炉炉温预报关联规则.针对现有挖掘算法的不足,提出了一种改进的多维时间序列模糊关联规则挖掘算法,该算法基于时间子序列和子序列间隔的双重模糊化,避免了挖掘结果“时间边界锐化”的问题.该算法应用于武钢的1#高炉,挖掘效果良好.
According to the current blast furnace temperature prediction rules are based on the experience of blast furnace experts to develop a new rule generation method - data mining to obtain blast furnace temperature prediction association rules.According to the lack of existing mining algorithms, proposed An Improved Multidimensional Fuzzy Association Rules Mining Algorithm for Time Series Based on Double Fuzzification of Time Subsequence and Subsequence Interval, This algorithm avoids the problem of mining results “sharpening time boundaries.” This algorithm is applied to the 1 # Blast furnace, excavation effect is good.