An Improved Method for Extracting Characteristic Orders of Unbonded Phase Faults

Abstract

The invention discloses an improved keyless phase fault feature order extraction method, comprising the following steps: 1) Obtaining the state information of the equipment through a vibration acceleration sensor, and preprocessing the obtained state information by using a time-frequency analysis method to obtain the instantaneous frequency ; 2) Carry out conventional integral operation on the estimated instantaneous frequency to obtain a roughly estimated instantaneous phase, and then use the Romberg integral rule to correct the roughly estimated instantaneous phase, and finally obtain an accurate estimated instantaneous phase; 3) According to the time domain and The mapping relationship in the angle domain uses the accurately estimated instantaneous phase information to resample the original signal in the angle domain; 4) uses the flexible angle domain synchronous averaging method to perform noise reduction processing on the angle domain resampled signal, and then the angle domain after noise reduction The fault characteristic order of the equipment is extracted by performing order spectrum analysis on the domain resampled signal. This method can accurately extract the fault characteristic order of the equipment under the condition of keyless phase change speed.

Description

technical field [0001] The invention belongs to the field of fault feature extraction of mechanical vibration signals and variable speed working conditions of mechanical equipment, and relates to an improved keyless phase fault feature order extraction method. Background technique [0002] Mechanical fault diagnosis is usually based on the assumption that the machinery is running smoothly, but in fact many mechanical equipments have working conditions where the speed changes. The operating characteristics of variable speed conditions cause the fault characteristics of traditional diagnostic methods to appear blurred, dynamically changed, or even annihilated. At the same time, due to the continuous existence of noise, the signal-to-noise ratio is reduced, which increases the difficulty of extracting fault characteristics. Existing diagnostic methods for stable working conditions often rely on the acquisition of frequency modulation information. In the absence of frequency mod...

AI Technical Summary

Problems solved by technology

There are two difficulties in using this method to track the order of the keyless phase: 1) directly integrating the obtained instantaneous frequency will lead to deviations in the instantaneous phase information, making it difficult to accurately guide the resampling of the original signal; 2) There is noise interference in the original signal. When the signal is re-sampled in the angle domain, the noise interference will also enter the angle domain. The commonly used keyless phase order technology lacks research in eliminating noise interference. Therefore, the extracted equipment fault characteristic order times inaccurate

Images