The Research On Coherent Noise Suppression Of Seismic Signal Based On S Transform

The seismic signal of seismic exploration acquisition always contains a lot of coherent noise, which seriously hampers the researchers'work for correct seismic interpretation. Thus it's necessary to suppress these noises. The two primary coherent noise, Surface wave and linear interference, have aroused people's concern. The paper takes these two noises as main research object, and researching on corresponding different algorithms for these noises'suppression.The paper begins with the introduction of the noises in seismic signals and commonly used methods for noise suppression. In chapter two, classical time-frequency analysis and S transform time-frequency analysis are introduced. This chapter first introduces the research status of time-frequency analysis; then completes classical time-frequency analysis introduced by theoretical analysis and experimental simulation; finally introduces S transform through theoretical analysis and experimental simulation. In this chapter, we explain the advantages of S transform by contrasting with classical time-frequency analysis. It offers support for suppression of surface wave and linear interference.The third chapter mainly researches on suppression of surface wave in one-dimensional seismic signal. This chapter proposes an algorithm based on S transform to suppress surface wave in one-dimensional seismic signal, by designing time-frequency filter. In the end, comparing with Fourier transform and wavelet transform for surface wave suppression in the theoretical model and real seismic signal, this chapter proves that this algorithm can suppress surface wave cleaner and reserve more effective signal.The fourth chapter researches on suppression of surface wave in seismic records. This chapter proposes an algorithm based on S transform to suppress surface wave in seismic records, considering the low frequency characteristic and space-time distribution features of surface wave. Comparing with F-K method and wavelet transform of surface wave suppression in theoretical model, this chapter proves that this algorithm can suppress surface wave cleaner and reserve more effective signal. In addition, we show its wide applicability with signal-to-noise ratio as evaluation in theoretical model which has different surface wave energy intensity. In the end, Comparing with F-K method and wavelet transform of surface wave suppression in real seismic records, this chapter proves that this algorithm can suppress surface wave cleaner and reserve more effective signal, and through the analysis of spectrum and frequency-wave number diagram, we obtain a same conclusion.The fifth chapter mainly researches on suppression of linear interference in seismic records. This chapter proposes an algorithm, which combines automatic tracking and S transform. Combined with automatic tracking, a small window, correlation and S transform, this algorithm can complete linear interference suppression. Comparing with MP (matching pursuit algorithm) method and FK filtering method, it verifies that this algorithm can suppress linear interference more thoroughly without waveform distortion phenomenon. The last section proposes an improved algorithm in order to remove the weak linear interference. This algorithm completes suppression of linear interference through filtering in S domain. It is based on the first algorithm. Comparing with MP method and F-K filtering method, it verifies that this algorithm can suppress strong and weak linear interference.