Unbalanced signal extraction method based on variable-sampling-length correlation filtering

Abstract

The invention discloses an unbalanced signal extraction method based on variable-sampling-length correlation filtering. The method comprises the following steps: acquiring a maximum interference frequency and an unbalanced signal frequency: acquiring a vibration signal detected by a vibration sensor, performing MUSIC (Multiple Signal Classification) power spectrum estimation on the acquired signal after digital tracking and filtering to obtain the interference frequency with a maximum amplitude, and measuring the unbalanced signal frequency through a rotating speed sensor; determining a sampling length T: selecting a reasonable sampling length according to an interference degree distribution rule, judging whether the absolute value of a primary sampling length is integer multiples as large as an unbalanced signal or not, if so, determining the absolute value as the sampling length, otherwise, determining the sampling length as an adjacent integer multiples unbalanced signal period with a small interference degree; performing band-pass filtering on the acquired signal of which the sampling length is T, and extracting the phase and amplitude of the unbalanced signal by adopting a correlation filtering method. According to the unbalanced signal extraction method, the amplitude and phase extraction errors do not change along with the signal to noise ratio and the frequency difference between the signal and the interference, and the stability of unbalanced signal extraction is ensured.

Description

technical field [0001] The invention relates to an unbalanced signal extraction method, in particular to an unbalanced signal extraction method based on variable sampling length correlation filtering. Background technique [0002] As the rotation speed of rotating machinery increases, the stability of mechanical operation is strictly required. The stability of rotation can be measured by the phase and amplitude of the unbalanced signal, and the balance is achieved by balancing the unbalanced amount, that is, dynamic balancing technology. The principle of dynamic balance technology: when the motor drives the shaft to rotate, the vibration signal of the shaft is collected by the sensor installed on the support structure of the shaft, and the vibration signal is processed by related signal processing such as noise elimination and filtering to obtain the unbalanced signal, and the phase and phase of the unbalanced signal are calculated. Amplitude, by moving the compensation devi...

AI Technical Summary

Problems solved by technology

The FFT method has spectrum leakage in the case of limited data length, and it is difficult to suppress the influence of near-frequency interference on unbalanced signals.

Correlation filtering method and sinusoidal parameter estimation method have low frequency resolution and cannot distinguish near-frequency interference

The modern spectrum estimation method has high resolution and can accurately estimate the interference frequency under short data conditions, but the amplitude and phase estimation are inaccurate

Images