Fractional Order Complex Diffusion Algorithm For Desert Seismic Random Noise Suppression

With the development of society,the demand for oil and natural gas is increasing.At present,the output of oil and gas in our country is far from meeting the demand.Guaranteeing the development and supply of oil and gas resources has become a strategic issue of national economic development.Tarim basin is located in the south of Xinjiang,China.The oil and gas content in Tarim desert is very high.Exploring oil and gas resources in desert by seismic exploration has important theoretical significance and economic value for national economic development.Seismic exploration is an important means of oil and gas resources exploration.Because of the special geological conditions and mining environment in desert areas,the effective signal energy of underground information reflected in seismic records obtained is low,accompanied by low frequency,weak similarity and strong random noise,which greatly reduces the quality of seismic data and makes it difficult to directly recover effective signals from seismic data to identify underground structures.Therefore,it is necessary to filter desert seismic exploration data to suppress random noise and improve signal-to-noise ratio.The low-frequency characteristics of desert random noise are similar to that of effective signals.The existing methods of suppressing desert random noise usually assume that the noise has the properties of Gauss white noise,so it is difficult to obtain satisfactory suppression effect of desert random noise.Aiming at this problem,in this paper,aiming at the characteristics of desert noise,a method of suppressing random noise based on complex diffusion filter is studied,so as to realize the suppression of random noise in desert seismic exploration and the preservation of effective information of seismic signal.Based on Schrodinger equation,the complex diffusion filtering algorithm extends the diffusion filtering to the complex domain,so that the imaginary part of the filtering result is equivalent to the smooth second derivative,which can well characterize the edge and structure of the signal.Then,the imaginary part is used to control the diffusion intensity of different structures,which improves the edge-preserving property of the diffusion filtering and overcomes the ladder effect based on gradient diffusion.In this paper,complex diffusion filter is applied to desert seismic exploration data processing.The influence of diffusion function and iteration times of complex diffusion filter on filtering results is analyzed.The synthetic seismic data analysis shows that the complex diffusion filter with optimized parameters can effectively suppress desert random noise suppression.In addition,a fractional complex diffusion random noise suppression algorithm is proposed for complex shape and detail recovery of seismic exploration signals.The gradient of stochastic noise suppression model based on complex diffusion equation adopts integer order,which has insufficient ability to capture complex structural features and details of seismic signals.Therefore,based on G-L fractional calculus method,a fractional complex diffusion filter model is proposed.Theoretical analysis and data simulation results show that the fractional gradient can better characterize the structural characteristics of seismic signals.On this basis,the threshold and fractional order schemes of diffusion function are studied to adjust the diffusion degree of signal and noise regions.Fractional complex diffusion model uses imaginary part and diffusion threshold to control the diffusion degree,which ensures that the seismic signal features and texture details are not destroyed during the diffusion process,and avoids smoothing texture details when integer order diffusion occurs.In order to verify the processing effect of fractional complex diffusion filtering algorithm on desert seismic exploration data,this paper applies fractional complex diffusion to noise suppression of synthetic desert seismic exploration data and actual desert seismic exploration data,and compares it with P-M diffusion algorithm and F-K filtering algorithm in the paper.The results show that the fractional order complex diffusion filtering algorithm can suppress the low-frequency random noise more thoroughly,protect the signal details better,and seismic events with large slopes clearer.