Underwater Transient Signal Detection And Feature Extraction Based On Vector Hydrophone

As aircraft torpedo is a menace of submarine, researching the underwater instantaneous signal produced by launching a torpedo is very important. So if we find the torpedo by detecting the underwater instantaneous signal in time, defending against the torpedo attack is easy. Given that the condition of ocean is complicated, it is a hard spot to research in this domain. Presented in this thesis are a study of the underwater instantaneous signal with vector signal processing technology and testing out theories by experiment.In the first part, basic principle of vector signal processing technology and common methods of signal analysis based on the time-frequency field are introduced. Hilbert-huang transform is particularly introduced because it has advantages in processing the non-stationary signals. In this thesis,Hilbert-huang transformis used to analysis the frequency contents of instantaneous signal. Meanwhile, EMD decomposing in Hilbert-huang transform is applied to vector signal processing, then simulation and analysis are made on DOA estimation and vector array beam forming. Through simulation and analysis, it is proved that this method can reduce the impact of the niose , effectively detect weak targets and enhance resolution of the DOA estimation, vector hydrophone is superior in acoustic signal processing to acoustic pressure hydrophone.A mathematical model is made based on the analysis of the underwater instantaneous signal, which is produced when aircraft torpedo is dropping into water.this mathematical model includes the acoustic pressure ,velocity, and acceleration of the underwater instantaneous signal. We provide a theory of underwater instantaneous signal processing by deducing the mathematical model with vector signal processing technology. Finally, a series of experiments which were designed for testing the theory, the experimental data was processed based on energy detection and short-time fourier transform. What is more, signal detection and feature extraction of acoustic signal of target were completed.