Research On Lifting Algorithm Of Mult-synchrosqueezing Transform And Its Application In Seismic Signal Processing

Time-frequency analysis method plays an important role in the field of signal processing,it is an important tool for processing and analyzing signals.Time-frequency analysis method can provide information of signal in time domain and frequency domain,and clearly describes how the signal frequency changes over time.This paper has carried out a series of research on time-frequency analysis methods.Firstly,some traditional time-frequency analysis methods such as Short-Time Fourier Transform,Continuous Wavelet Transform,S-transform and three parameter Generalized S-transform are introduced and analyzed.Then a new high precision time frequency analysis method is introduced,it's Muti-sychrosqueezing transform(MSST).On the basis of Short-Time Fourier Transform,the obtained STFT spectrum is squeezed,and the divergent energy around the real frequency is squeezed to the real frequency center to obtain a high-precision time-frequency spectra.The "squeezing" process is repeated to improve the accuracy of time-frequency spectrum.Because MSST has some advantages in signal processing,it means that MSST can be applied in the field of seismic signal processing to a certain extent and provide a powerful tool for seismic reservoir identification.On the basis of MSST,this paper improves it from different directions.We put forward two new algorithms: Joint Empirical Wavelet Transform and Muti-sychrosqueezing Transform(EMSST)and Muti-Sychrosqueezing Generalized S-transform(MSSGST).We study its application in synthetic signal and seismic signal.The specific research contents and achievements are as follows:(1)In this paper,we study and compare the commonly used time-frequency analysis methods,such as Short-Time Fourier Transform,Continuous Wavelet Transform,S-transform and three parameter Generalized S-transform,the MSST algorithm is introduced and further studied by constructing appropriate synthetic signals,which lays a good foundation for the follow-up study of new algorithm and its application in seismic reservoir prediction.(2)EMSST method is proposed based on Wavelet Transform and Multi-synchrosqueezing Transform.The MSST method is introduced into seismic signal processing and according to its weak ability in processing multi-component complex seismic signal,the new method(EMSST)is proposed.The EMSST method is used to process the synthetic signal,and the results are compared with those of MSST and EWT-HT.It is found that the EMSST method has better time-frequency characterization ability.Then,the EMSST method is applied to the actual data analysis of seismic signal to further realize the purpose of seismic reservoir prediction.(3)The Multi-synchrosqueezing Generalized S-transform(MSSGST)is proposed.According to the fact that MSST is a post-processing method of short-time fourier transform,and the three parameter Generalized S-transform has a wider range of application and higher time-frequency accuracy than short-time Fourier transform.The Muti-sychrosqueezing Generalized S-transform is proposed,and its strict mathematical reasoning process is given.Then,the MSSGST method is tested by constructing appropriate synthetic signals,and compared with the results of STFT,GST,SSGST and MSST.It is concluded that the MSSGST method not only has a good time-frequency analysis accuracy,but also accurately describes the high-frequency and low-frequency components of the signal.Finally,the MSSGST method is applied to the actual data of seismic signal to test the processing effect of the seismic signal.The experimental results of the actual data show that the MSSGST method can predict the seismic reservoir very well.Compared with the STFT,GST,SSGST and MSST,the results show that the MSSGST method has higher time-frequency analysis accuracy and more accurate reservoir prediction.