DEMON Feature Extraction of Acoustic Vector Signal based on 3/2-D spectrum

Abstract

The vector characteristic of target radiated noise is analyzed by 3/2-D spectrum that can theoretically remove Gaussian and symmetrical distribution noise. A study is carried out to compare the performance of 3/2-D DEMON about acoustic vector signal to conventional DEMON about sound pressure and acoustic vector signal. Propeller shaft frequency of sound pressure and acoustic vector signal is detected and its performances are analyzed under the various noise backgrounds and the different SNR conditions. It is proved through simulation that the ability of DEMON feature extraction and detection of propeller shaft frequency about acoustic vector signal based on 3/2-D spectrum is stronger than sound pressure and acoustic vector signal based on power spectrum. 3/2-D spectrum has provided a way of DEMON feature extraction and detection of propeller shaft frequency for low frequency acoustic vector signal based on the method of higher-order statistics.