Fault diagnosis method for high-voltage circuit breaker energy storage mechanism

Abstract

The invention relates to a fault diagnosis method for an energy storage mechanism of a high-voltage circuit breaker. The method comprises the following steps: firstly, removing background noise from acollected acoustic signal by adopting morphology, putting forward a kurtosis and envelope similarity-based time scale alignment method to ensure the synchronism of an acoustic vibration signal, thenconstructing a two-dimensional image feature matrix for the acoustic vibration signal subjected to data expansion by utilizing a Pearson correlation coefficient, and finally training the feature matrix by utilizing a CNN (Convolutional Neural Network). A CNN model structure is improved by adopting local mean normalization and kernel function decorrelation, so that the influence of large data change in the energy storage process on the diagnosis accuracy of the circuit breaker energy storage mechanism is reduced. According to the method, the overall diagnosis accuracy reaches 98.1%, the generalization performance is good, and the method has obvious advantages compared with a traditional method.

Description

technical field [0001] The invention relates to a fault diagnosis method for an energy storage mechanism of a high-voltage circuit breaker, in particular to a fault diagnosis method for an energy storage mechanism of a circuit breaker using acoustic vibration signals to construct a CNN feature matrix. Background technique [0002] As an important control and protection device in the power system, the high-voltage circuit breaker can directly affect the safety and stability of the power system, so the fault diagnosis of the circuit breaker is very important. At present, research on circuit breakers focuses on the opening and closing process: using coil current, insulation rod displacement, and vibration signals to identify mechanical faults, the research focus is on the problems that occur in the equipment itself during the operation of the circuit breaker, and there is insufficient research on the faults in the energy storage process In-depth, lack of quantitative basis for ...

AI Technical Summary

Problems solved by technology

[0005] (1) The difference of acoustic and vibration signals is not considered, and the acoustic and vibration signals are combined mechanically, resulting in insufficient diagnostic accuracy

[0006] (2) The method of feature extraction relies on manual selection and expert knowledge, which is too subjective and easily causes the omission of fault information

Most of the above studies are based on the fault diagnosis of a single signal by 1DCNN. The diagnosis process is cumbersome and the accuracy is not high enough, and the self-learning ability of CNN is not maximized.

Images