The Study Of Self-noise Characteristic Of One Under-water Vehicle

The self--noise of underwater vehicles is one of the main interferesources in sonar, which is compound of three kinds of noise inc1uding themachinery noise, the propel ler noi se and the water ki net ics noise. It wi l lbe influence detect ability of sonar heavi 1y. So the study on self--noiseof underwater vehicles is an important work in sonar. It is studied inthi s thesi s that the self--noise data sampl ing in one underwater vehiclesmade in the institute of Kunming 705 is get, then the pecu1 iarity analysisof noise and some results is given out in this thesis.Modern spectra1 ana1ysis techno1ogy is called the high resolutionspectral analysi s technology because of its advantage wh ich have more hi ghspectral resolution ratio and overcome the di sadvantage of the classicalspectral analys is. In special ly, the est imat ion of power spectra1 densi ty(PSD) can be get in Burg method which estimate reflection coefficientfirstly, then get the AR parameter est imat ion usi ng the Levinson recursion.The recursion made the prognosis error power estimation minimum fordifferent orders prognosi s. The simulation results shown that it isunbiased estimation for the wide band procedure and the narrow bandprocedure in the Burg method and it approaches to the true using less datarecords in the condition of 1ower SNR.In the experiment, the self--noise samp1ing system is composed in bothof high speed signal processing system and sold state hard--disk, theunderwater vehic1es moves in high and lower ve1oci ty in the lake and thenoise data sampling are get using this samp1 ing system. the connectionof the velocity and self--noise intensity is studied. Then the spectralana1ysis in the Burg method and results is given out in high and 1owerve l oc i ty.