The Application Of Case-Based Reasoning In Transformer Faultdiagnosis

The power transformer is an important component of power system and itsoperating state have important meaning for the security, stability and economicoperating of power system. It is high necessary to develop conditionmonitoring and transformer fault diagnosis.This paper mainly investigates the combination of weighted fuzzy kernelclustering (WFKC) and Case-based reasoning (CBR) used in fault diagnosis.CBR, which is the emerging technology of AI, have been used successfully infault diagnosis. However, with regard of the Euclidean distance algorithm inCBR cannot reflects the relationship of sample data, and the Retrievalefficiency of case retrieval isn't high, so the CBR is limited in the faultdiagnosis of power transformer . Aimed at the problems existed in CBRalgorithm which is applied in DGA, the weighted fuzzy kernel clusteringalgorithm is firstly introduced into the Fault Diagnosis of Power Transformersto build a new fault diagnosis model．In the algorithm, firstly considering thatthe different affects of the different attributes on cluster results, so the similaritybased weighting method is used to assign weight to features of the transferredsamples, and then weighted fuzzy kernel clustering in the feature space isrealized when the transferred samples in the original space is mapped into highdimensionalfeature space. The new fault diagnosis model can adequatelyconsider that the different affects of the different attributes on cluster results andeffectively improve the clustering capability for the complex dataset, and theRetrieval efficiency is effectively improved.The paper improves the model of fault diagnosis for power transformerbased on CBR, and builds the system of fault diagnosis for power transformerbased on CBR, so the correct rate is effectively improved. The proof of a great deal instance indicated the validity and reliability of this system towardstransformer diagnosis and analysis. All the job completed in this paper haveactive and indicative meaning to the fault diagnosis and condition-basedmaintenance of transformer.