Research On The Methods Of Seismo-acoustic Signal Processing

When the low frequency acoustic wave transmits from the sea to the near shore, the energy of the acoustic wave has been greatly attenuated. But a part of energy of the acoustic wave transmits into the seabed, comes into being geoacoustics signal, then the signal propagates to the shore. So the target in the sea can be detected by detecting the low frequency geoacoustics signal. The key point of the problem is the weak low frequency signal detection with strong noise. For solving it, this paper mainly works for the noise reduction, signal detection, and power spectrogram analyse.Firstly, some of noise reduction methods are introduced. The research is doing at the following aspects: eliminating the trend, removing the transient interference, reducing noise. The results show these methods increased the SNR and they are help for the signal detection.Then, through introducing the theory of the classical copy-correlation and chaos signal detection, some research have been done for the efficient signal detection with strong noise at the background.At last, the theory of power spectrogram estimation and LOFAR spectrogram analyse are discussed. The experiments data is analyzed at the frequency spectrum by these methods.This paper attempts to do some research on the geoacoustics signal for the first time. Through studying some traditional signal detection theory, we find the methods which adapt to the detection of the geoacoustics signal. The characteristic of geoacoustics signal frequency spectrum is cognized by analyzing the power spectrum. Through processing the experiments data, it is verified that the geoacoustics signal can be detection when the source is far from the receiving station. It provides a new thought for the object detection in the distance.