Extraction of shaft frequency based on the DEMON line spectrum

Abstract

Target-identification is an important technique in modern sea war. As the core device to identify targets, negative sonar analyzes noise underwater and abstracts their properties and determines its source. Shaft frequency which barely lies in the properties of the propeller differs among different kinds of ships, so this property can be regarded as a key to identify targets. This paper studied the theory and math model of the noise produced by ships. And Detection of Envelope Modulation on Noise (DEMON) is an effective technique to abstract low-frequency properties from high-frequency signals via demodulation. There are two methods to demodulate the noise, namely absolute low pass demodulation and square low pass demodulation. To improve the performance of DEMON spectrum estimation, adaptive line spectrum enhancement technology is introduced. An adaptive threshold is then set to eliminate the continuous spectrum. Then the largest common divisor algorithm is presented to calculate to shaft frequency out of the array consist of the centre frequency of these line spectrum. Simulation is conducted as described above. Two methods of demodulation is compared in this part. Brief sea experiment data processing shows that this algorithm is able to abstract the shaft frequency among the ship noise.