动态关联规则挖掘辅助产品服务系统规划分析

Aided analysis for product service system planning based on dynamic association rule mining

数据挖掘方法可以从已有的设计知识数据库中提取用户需求和设计需求之间的关联规则,辅助产品服务系统(product service system,PSS)规划分析。Apriori算法是常用的基于静态数据库的关联规则挖掘方法。实际上,设计知识数据库是不断动态更新的,关联规则库也需要更新。针对这个问题,提出了基于Apriori的改进关联规则挖掘方法和增量式更新算法以辅助PSS规划分析。关联规则挖掘方法中将新增一个竞争集合储存有希望成为规则的项集。考虑新增数据库具有较强的创新意义,增量式更新算法将基于数据库加权策略更新规则库。以某计量泵企业的PSS规划设计为例,验证了所提方法的有效性。

Data mining method can extract the association rules between customer requirements（CRs） and design requirements（DRs） from the existing design knowledge database,to aid the product service system（PSS） planning analysis.Apriori is a common association rule mining algorithm,which is based on static database.In fact,design knowledge database updates dynamically and association rule base also needs updating.To deal with this problem,an improved association rule mining method based on Apriori and an incremental updating technique were proposed to aid PSS planning analysis.A competitive set was added in the association rule mining method to store the itemsets which are hopeful to be rules.Considering the new-added database has much more innovative meaning,the incremental updating technique updated rules base based on a weighting strategy.A case study of PSS planning for a metering pumps enterprise was presented to illustrate effectiveness of the proposed method.