| With the development of human beings in the oceans,marine exploration and underwater acoustic applications,under the condition of low signal-to-noise ratio,detection of the non-cooperative underwater acoustic pulse signal is often needed.Due to the unknown of signal form,appearance time,duration,center frequency,and so on,the detection of non-cooperative underwater acoustic signal is still a difficult problem.In this thesis,the basic principles of entropy detection,instantaneous correlation integration,high-order cumulate,cyclic spectrum and short-time Fourier transform are studied firstly,and then this thesis shows a variety of underwater acoustic signal detection methods with unknown parameters theoretical analysis,through the simulation to verify the feasibility of each detection algorithm.Secondly,this thesis studies the influence of signal-to-noise ratio,signal form and signal parameters on the detection performance.The simulation and comparison of different methods under different influence factors are given.The influence of different factors on the detection performance of each method influence,analyze the robustness of various methods to test performance.The simulation results show that the short-time Fourier transform and the instantaneous correlation integral method have the least influence on the signal-to-noise ratio;and the short-time Fourier transform,the instantaneous correlation integral method and the high order cumulate method are insensitive to the signal form;and the width and bandwidth have an impact on the detection of each method.Finally,an integrated optimization detection scheme of underwater acoustic pulse signal with unknown parameters is constructed by comparative analysis,which can meet the detection ability of underwater acoustic signals with unknown parameters in various signal forms.Based on the experimental data of Qian Dao Lake's test,the performance analysis shows the feasibility,applicability and robustness of the scheme. |
PDF Full Text Request