Study On The Identification Technology Of Seismic Wave Interfacial Wave For Shallow Ship

In this thesis,we mainly study on the identification technology of seismic wave interface-wave for ship in the shallow sea.Which propagates between two different media(flow-solid),also was called Scholte wave.Scholte waves in seismic waves ofship has the characteristics of slow propagation speed,low frequency,medium amplitude,large energy and long period of vibration,itis very suitable for water target detection.Through the difference of time that the ship seismic wave signals arrive at two sensors of a certain distance away from each other to calculate the seismic velocity of the ship and to synthesize the seismic records manually.The Scholte waves are extracted by the ?-p transformation algorithm and transformed from x-t domain to ?-p domain.According to the velocity of each wave,the p value of Scholte wave is the maximum,and the other interference wave can be cut off by setting the range of p value,and the Scholte wave can be extracted.According to the polarization filtering theory,the shape of the trajectory of each particle in each wave is different.According to the unique particle trajectory of Scholte wave(elliptical track),the Scholte wave is identified.According the pseudo-Wigner distribution to extract two eigenvalues(frequency and energy)of the Scholte wave and build a two-class SVM training sample,including seven Scholte waves and their three eigenvalues(velocity,frequency and energy),the Scholte wave is analysedand classified.By simulating the signals of seismic wave interface-wave of ship,to verify the two-class SVM,the classification accuracy has reached to 95%,the Scholte wave can be well identified.