Abstract
As for the failure diagnosis problem of the transformer of traction system of the bullet train, the model-based diagnosis method is introduced to the failure diagnosis of traction transformer of the bullet train in this paper, and the failure diagnosis scheme based on the combination of model diagnosis and fuzzy Petri network is also proposed. Two-layer model for the structure and function of traction transformer is set up, so that the minimum conflicts candidate set can be searched offline, and the minimum conflicts set can be identified online, thus the external failure diagnosis of the traction transformer can be achieved. As for the internal failure diagnosis, the model-based diagnosis is applied to the positioning of failure elements of the traction transformer; and the fuzzy Petri network is used for regional knowledge representation so as to ratiocinate and achieve the internal failure diagnosis. The analysis is implemented with the failure data of three-phase V/X wiring transaction- transformer of the bullet train as the example; and the diagnosis result is the verification for the feasibility and effectiveness of transaction-transformer failure diagnosis scheme based on the integration of model diagnosis and fuzzy Petri network.
