基于粗糙集理论的变压器故障诊断方法

Diagnosis Method of Transformer Faults Based on Rough Set Theory

将一种基于粗糙集理论的信息熵约简方法应用于变压器故障诊断问题中。首先应用粗糙集理论将电力变压器故障历史数据进行分析统计,建立决策表,然后采用信息熵约简算法对其进行条件属性约简,求取一组最小约简知识系统,并采用粗糙集约简方法对新系统进行简化,得到一组故障诊断的最小决策规则集。方法大大减小了编码的工作量,避免了约简属性组合查询及缺少关键属性时规则匹配所带来的不便,所以运算速度也会相对加快。最后结合实例分析,证明该方法的简便及有效性。

In this paper an information entropy algorithm based on Rough Set theory was utilized to solve the problem of fault diagnosis for power transformer. First of all, based on Rough Set theory, this paper analyzed the historical data of transformers faults and set up a decision table. Then this paper used the information entropy algorithm in conditional attributes reduction to get a reduction system. Subsequently, the new system using the method of Rough Set reduction was reduced and a group of minimal decision rules were produced. Accordingly, the coding workload was greatly decreased, attributes compounding queries and the problem of rules matching in the case of lacking key attributes were avoided, and so computing speed will pick up. Finally, the case studies showed that the method is simple and effective.