Research On Seismic Signal Denosing Based On Mode Decomposition Algorithm

With the development of science and technology,the limits of human activities are getting smaller,accompanied by the increasingly complex geological structure and environment of the exploration area,which inevitably cause more complex noise in seismic signals gathered and bring great difficulties to the subsequent processing and interpretation of seismic data,and ultimately affect the judgment of oil and gas storage and geological institutions.These noises can be classified into coherent noise and incoherent noise.Incoherent noise is random noise and difficult to be removed.In this paper,the Mode Decomposition algorithm is studied to suppress random noise contained in seismic signals and improve the signal-to-noise ratio of seismic signals.The main work includes:1?The seismic signal is denoised using the empirical mode decomposition algorithm.Firstly,the one-dimensional synthetic seismic signal is denoised and then each signal of the actual seismic signal is denoised by using the algorithm.Finally,the denoised actual seismic signal is synthesized.The results show that this method can eliminate the noise in the signal and improve the signal-to-noise ratio.The denoising effect of this method is verified.2?The seismic signal is denoised using complete global empirical mode decomposition algorithm.Firstly,the one-dimensional synthetic seismic signal is denoised.Then each signal of the actual seismic signal is denoised using the complete global empirical mode decomposition algorithm.Finally,the denoised actual seismic signal is synthesized.It can be seen from the results that the complete global empirical mode decomposition algorithm solves the modal aliasing well caused by EMD(empirical mode decomposition)and the completeness of empirical mode decomposition is also maintained.3?The seismic signal is denoised using the variational mode decomposition algorithm based on correlation coefficients.Firstly,the one-dimensional synthetic seismic signal is denoised.Then each signal of the actual seismic signal is denoised using the variational mode decomposition algorithm based on a correlation coefficient.Finally the de-noised actual seismic signal is synthesized.The denoised results show that this method can select the most appropriate K value when decomposing the variational modes,which overcomes the blindness of K value selection.The effect of denoising is better than empirical mode decomposition and complete global empirical mode decomposition.