面向冷轧带钢表面缺陷的加权模糊层次关联规则挖掘方法

Weighted fuzzy level association rule mining method for surface defects of cold rolled strip

为了克服传统关联规则在解决钢铁工业泛在信息多样性、模糊性和层次性方面的不足,在传统的层次型关联规则挖掘算法基础上,建立层次间的模糊隶属度关系,并引入加权扩展,设计了一种加权模糊层次关联规则算法,并将该算法应用于冷轧带钢生产过程缺陷与缺陷形成原因的关联关系挖掘中。试验表明,该算法不但可以挖掘出缺陷之间的关联关系,同时还能挖掘出不同缺陷与其对应的不同缺陷形成原因之间的关联关系,弥补了传统挖掘算法只能挖掘同种类别数据之间关系的不足。

In order to overcome the shortages of the traditional association rules in the steel industry in the field of information diversity, ambiguity and gradation. On the basis of traditional hierarchical association rules mining algorithm, introduces the weighted extension while establishing a hierarchical relation between fuzzy memberships, and proposes a weighted fuzzy hierarchy association rules algorithm. At the same time, the algorithm is applied to the association between defects and defects in the process of cold strip rolling. The experimental resuhs showed that,this algorithm can not only dig the relationship between defects, but also dig out the corresponding relationship between different defects and its different reasons. It make up for the shortages of that traditional mining algorithms can only be mining the data at same category.