136 / 2016-04-25 14:44:26

Transformer Fault Diagnosis Based on Variable Precision Rough Set

variable precision, transformer, IEC three-ratio method, fault diagnosis

Transformer fault diagnosis has important significance for the operation of the entire power system. Transformer fault diagnosis model based on variable precision rough set has been proposed for the actual situation of incompleteness of the test data and small transformer fault data sample. VPRS has the ability to deal with data classification problems without function between attributes. Firstly, conditional attribute reduction to eliminate redundant attributes, and then using the minimal decision algorithm for transformer fault diagnosis. Examples show that the method has high fault-tolerant, is able to handle a variety of complex transformer failure to overcome the shortage of IEC three-ratio method in such situations. In addition, variable precision rough set can handle transformer fault symptoms containing omissions or errors to improve fault diagnosis accuracy.