Full Waveform Inversion Applied Research Based On The Lack Of Low Frequencies

With the rapid development of full waveform inversion research in recent years,full waveform inversion and its theoretical research has gradually improved.In terms of application,it is becoming practical.Studys on seismic waveform inversion involve seismic wave field simulation,parallel policies,gradient estimation and cost function settings and other aspects of the content.With the rapid development of computational technology and new optimization algorithms are proposed,full waveform inversion is currently used in offshore exploration data and has gradually realized the practical.Full waveform inversion technique is an effective means to build high-precision velocity model.However,due to the loss of low-frequency information,the lack of long offsets,low stability of source Wavelet and low signal to noise ratio and other complex factors,the realization of full waveform inversion based on land application is still a big challenge.Theoretical basis of Full waveform inversion is discussed in this paper,and basic workflow of traditional full waveform inversion application is given.We point out that the problems in traditional inversion process of land applications,namely low-frequency information is missing and the initial model lack of precision leads to the process of inversion does not converge.In response to these issues,we proposed joint modeling method of logging information models,starting from the model domain to realize the coupling of inversion by improving the accuracy of initial model.This method uses logging information,through a specific interpolation method for logging data information into the background velocity to form the joint model.The model contains the background velocity field of low wave number,and number of full wave logging data elements.Meanwhile,the seismic data analysis and pre-processing,using multiscale inversion strategies,reducing the nonlinearity of the problem,and got a very good testing and practical application results.After analysis of full waveform inversion problems in application and log information,we point that to improve the accuracy in the initial model is an effective ways to avoid low frequency information.By introducing the well logging information in the background velocity we can improve the model vertical resolution.Through plug value algorithm research,we proposed the joint of wells information and the background of velocity to build initial model,not only effectively improved the precision of initial model,but reduced the dependence of full waveform inversion on low frequency data,In the case of effective frequency of 6 Hz has achieved good inversion result,Overthrust model test results show that the method is effective,and used in an exploration area of Sinopec.