An Approach to Transformer Fault Diagnosis based on Quantum Neural Network

J.-y. Wang, G.-w. Cai, L.-f. Zheng, and C. Pan (PRC)


Quantum Neural Network, quantum intervals, power transformer, fault diagnosis


When traditional BP network is used to analyze the uncertain information and the cross data of Dissolved Gas Analysis in the transformer fault diagnosis, the results are not ideal. In this paper, an approach for transformer fault diagnosis based on Quantum Neural Network is put forward. Firstly, the proper relation between the uncertain data of the decision-making and fault pattern is established in order to decrease the uncertainty of relative types identification. Secondly, the parameter space is mapped to the fault state space rationally by updating the quantum intervals and the output error of different fault is smoothed, through which the impact of fuzzy data, incomplete information and incorrect data on accuracy is solved effectively. Finally, the proposed method has been applied to practical engineering diagnosis, and the validity, feasibility and practicability have been proved.

Important Links:

Go Back