Circuit breaker mechanical state detection method, device and terminal equipment

Abstract

The invention provides a method, device and terminal equipment for detecting the mechanical state of a circuit breaker. The method includes: obtaining a plurality of mechanical parameter samples of the circuit breaker; determining the correlation degree between every two mechanical parameter samples, and according to each mechanical parameter Construct a binary tree based on the degree of association between the samples; train a node classifier based on the mechanical parameter samples contained in each node in the binary tree, and the node classifier is used to output the fault category of the circuit breaker according to the mechanical parameters of the circuit breaker input to the node and failure probability; obtaining mechanical parameters of the circuit breaker to be detected, and inputting the mechanical parameters of the circuit breaker to be detected into a binary tree for fault detection of the circuit breaker to be detected. The circuit breaker mechanical state detection method, device and terminal equipment provided by the invention can improve the fault detection speed and accuracy of the circuit breaker.

Description

technical field [0001] The invention belongs to the technical field of fault detection, and more specifically relates to a method, device and terminal equipment for detecting the mechanical state of a circuit breaker. Background technique [0002] In the prior art, when detecting the mechanical state of the circuit breaker, the method usually used is: extracting the mechanical state parameters of the circuit breaker, and training the neural network based on the mechanical state parameters of the circuit breaker, and realizing the fault detection of the circuit breaker through the neural network . [0003] However, there is a relationship between the faults of the circuit breaker. For two different types of faults, fault A may occur at the same time as fault B, and the two can be transformed into each other; fault A may not be related to fault B. occur at the same time, the two are mutually exclusive. Correlation between faults is not considered in the prior art, so there a...

AI Technical Summary

Problems solved by technology

[0004] (1) Failure to consider the transformation relationship between fault A and fault B leads to inaccurate distinction between fault A and fault B, and the neural network is prone to misjudgment;

[0005] (2) Failure to consider the repulsive relationship between fault A and fault B leads to a decrease in the fault detection speed of the neural network when there is a large amount of data

Images