关联规则挖掘在电厂设备故障监测中应用

Association rule mining in fault monitoring of power plant equipment

关联规则挖掘是数据挖掘的重要分支,其通过描述数据库中不同数据属性之间所存在的潜在关系规则,找出满足给定支持度阀值和置信度阀值多个域之间的依赖关系。随着电厂设备运行期间各种故障的发生,各状态监测点参数也会发生相应变化,利用关联规则挖掘算法,找出故障发生时故障现象与故障类别之间的关联关系,更好地对设备进行故障监测与诊断。阐述了关联规则挖掘的主要概念,对挖掘时最常用的Apriori算法进行探讨,并以汽轮机凝汽器的一种典型故障为例说明了算法的执行情况,对挖掘结果进行了解释。结果验证了所用方法的可行性与正确性。

Association rule mining,an important branch of data mining,is to find the inter-dependence among domains,satisfying the given support threshold and confidence threshold,by describing hidden relationships among different data attributes. Parameters at different monitoring points in the power plant are changing with various faults occurred during equipment operation. With the association rule mining algorithm,the relationship between fault phenomenon and fault type during fault occurrence can be set for further fault detection and diagnosis. The concept of the association rule mining and the widely used Apriori algorithm are expatiated. Taking a typical fault of the steam turbine condenser as an example,the feasibility and correctness of the monitoring method are validated by the result analysis.