An optimal demodulation frequency band selection method based on Holo-Hilbert spectral analysis for propeller modulation feature recognition

The modulation spectrum of underwater propellers typically contains information on propeller structural characteristics, such as the shaft frequency and blade number. Detection of Envelope Modulation on Noise (DEMON) analysis is an effective method for underwater propeller modulation feature recognition. However, its performance is limited by the selection of bandpass filters. To address this issue, this study proposes a method for selecting the optimal demodulation frequency band (ODFB) based on Holo-Hilbert spectral analysis (HHSA). Its core idea is to utilize the full frequency representation strategy of HHSA to construct a modulation intensity marginal spectrum (MIMS) curve, which adaptively reveals the modulation intensity at each frequency point thereby provides a basis for selecting the ODFB. Simulations and experimental data analysis illustrate that, compared with classical methods such as Fast Kurtogram, Protrugram, and Cyclic Modulation Spectrum (CMS) analysis, the proposed Holo-Hilbert DEMON (HH-DEMON) method involves selection of the ODFB containing more comprehensive modulation information, with the corresponding DEMON spectrum possessing higher spectrum quality with less interference. In practical applications, the HH-DEMON method enables more accurate and effective recognition of underwater propeller shaft frequency and blade frequency.