Breaker fault diagnosis method

Abstract

The invention discloses a circuit breaker fault diagnosis method. The method comprises the following steps: collecting a vibration signal and a sound signal of a circuit breaker; optimizing a parameter K and a parameter alpha in VMD decomposition by using a firework algorithm, and then acquiring a vibration signal IMF component and a dimensionless parameter of a sound signal; taking the energy entropy of the vibration signal IMF component as a vibration signal feature vector W1, and taking the dimensionless parameter of the sound signal as a sound signal feature vector W2; and taking W1 and W2as subsets of the feature vector W, and inputting the feature vector W into a support vector machine optimized by grid search for diagnosis and identification. Compared with a single-signal fault diagnosis algorithm, the circuit breaker fault diagnosis method based on vibration-sound joint analysis has obvious advantages; compared with unoptimized variational mode decomposition, the variational mode decomposition effect optimized based on the firework algorithm has a better signal decomposition effect, signals can be decomposed more completely, and the decomposition result is better.

Description

Technical field [0001] The present invention relates to a circuit device fault diagnosis method, in particular a circuit breaker fault diagnosis method. Background technique [0002] High-voltage circuit breakers serve as a protective equipment in electricity production, it is important to operate the normal operation of electrical equipment. According to research results, most of the faults of the circuit breaker are mechanical failures, so the diagnosis of mechanical failure of high-voltage circuit breaker is very important. [0003] The current research method for the mechanical fault diagnosis of high-pressure circuit breaker is divided into three types of coil current, vibration signal, and sound signal, which is dominated by a vibration signal based on the vibration signal. The circuit breaker fault diagnosis based on the coil current can only diagnose the state diagnosis of electromagnets and auxiliary switches. The vibration signal is mainly generated by the vibration of ...

AI Technical Summary

Problems solved by technology

[0005] Someone proposed a variational mode decomposition (Variational Mode Decomposition, VMD) method, compared with the ensemble empirical mode decomposition (Ensemble empirical Mode Decomposition, EEMD) and the local mean decomposition (Local Mean Decomposition, LMD) method, the VMD decomposition method The effect is good, and it can separate signals with similar frequencies. It is less affected by the sampling frequency and has a certain denoising effect. However, the selection of K and alpha parameters under different signal characteristics has a great impact on the decomposition effect.

Images