A propeller shaft frequency searching method based on improved noise envelope signal recognition

Abstract

The invention relates to a propeller shaft frequency searching method based on improved noise envelope signal recognition. The method includes: enveloping an original hydrophone signal, and performingHilbert-Yellow transform on each wavelet packet signal to obtain a wavelet envelope signal; performing wavelet reconstruction to obtain an improved noise envelope signal; calculating a time correlation spectrum, take a line spectrum protruding at the peak of the low frequency band as the suspected shaft frequency; setting a step size smaller than the frequency resolution according to the demand;forming a suspected shaft frequency set with the suspected shaft frequency as the center; performing cyclic correlation calculation on a time correlation spectral signal on each suspected shaft frequency to obtain the frequency statistical amount of a cyclic modulation spectrum; and finally searching the shaft frequency of a target according to the frequency statistical amount. The method of the invention corrects axial frequency offset errors generated by the conventional cyclic modulation spectrum methods in the spectrum calculation process, improves the accuracy of the shaft frequency detection, greatly reduces the calculation amount of the loop iteration, and is convenient to implement in engineering application.

Description

technical field [0001] The invention belongs to the fields of underwater acoustic engineering, marine engineering and sonar technology, and relates to a propeller shaft frequency search method based on improved noise envelope signal identification, which is an improved noise envelope based on the combination of cyclic modulation spectrum and multiple signal classification algorithms The signal recognition method is used to extract the shaft frequency of the ship propeller and the harmonic line spectrum of each order. This method is suitable for target recognition of ship radiation noise. Background technique [0002] Theoretical research and actual measurement show that the propeller noise of a real ship is a non-stationary broadband continuous signal, and its spectrum signal is composed of continuous spectrum noise and low-frequency discrete noise. The main sound source of the continuous spectrum is the cavitation noise of the propeller. The low-frequency discrete noise is ...

AI Technical Summary

Problems solved by technology

[0004] In order to avoid the deficiencies of the prior art, the present invention proposes a propeller shaft frequency search method based on improved noise envelope signal identification, which overcomes the large amount of calculations that exist in the existing noise signal identification method when estimating the target shaft frequency. Disadvantages such as slow iterative convergence and poor estimation accuracy

Images