Method Of Sonar Platform’s Self-Noise And Its Computer Implementation

Sonar is an important electronic device in naval vessels, which can be used to localize and recognize the target by processing the received acoustic signals. While sonar can be distorted by ambient noise sources. Ship radiated noise, ocean ambient noise and the sonar flow noise will affect sonar’s beamforming performance. Thus they can affect sonar’s detection distance, and affect the performance on target localization and features recognition.This paper analized the main noise source among sonar background noises. The effects sonar selfnoise has on sonar is studied. The noise dmitted by the main sources are diveded into correlated noise and uncorrelated noise, the effects both noise have on beamforming performance were discussed in simulation.Since sonar self noise and ship radiated noise have similar noise sources, sonar self noise model was built based on ship radiated noise model. Bartlett power spectrum estimation, stochastic processing analysis and line spectrum extraction were used to realize the simulation of Statistical analysis modeling method of sonar self noise. Signal has the same power spectrum as given signal can be obtained with arbitrary length by this method. In this method, self noise power spectrum was supposed to be superimposed by smooth continuous spectrum and line spectrum. The smooth continuous spectrum component was simulated by AR model. The line spectrum component was simulated by a time domain method of generating single-frequency signals. The problems of frequency domain resolution and line spectrum extraction were discussed simply.Another sonar self noise simulation method was proposed in this paper. The near field naval vessels was simplified as several point sources and a reflection effect from the sea floor and sea surface in this method, furthermore, the acean ambient noise and sonar flow noise were considered in this method. Most of the sound source radiated noise simulation were done by generating known power spectrum in stochastic process. As to the sonar flow noise, only the noise originated from pressure fluctuations in turbulent boundary layer(TBL) was considered in this paper. Corcos model and Carpenter model are adopted in the simulation of pressure fluctuated beam-frequency spectrum. The function obtained by Lindermann O A was used to approximate the transferred function of the sonar dome. The power spectrum of flow noise was estimated with numerical integration method, then the time domain signal of the flow noise was simulated by use of obtained power spectrum. Signals from different sources was delayed respectively. Simulatioin of sonar array signals was completed.