| With the development of oil and gas exploration and development towards greater depth and more complex geological structure areas,high-precision geophysical imaging technology needs to be developed urgently.Reverse time migration imaging can solve the problem of high and steep Angle and adapt to most types of velocity changes.At the same time,reverse-time migration technology also has some defects,such as not very good amplitude,low frequency noise interference and other shortcomings.The least square inverse time migration imaging technology introduces the idea of model matching data into the inverse time migration imaging technology,making the imaging results not only have the advantages of the inverse time migration imaging results,but also take into account less low-frequency noise,higher resolution and more accurate amplitude.This thesis mainly studies inverse time migration imaging technology and least square inverse time migration imaging technology.The main research achievements are as follows:(1)Reverse time migration imaging from 2d and 3D is realized.To solve the problem of strong low frequency noise in reverse time migration,Laplace filtering method is used to suppress low frequency noise in reverse time migration imaging results.(2)The least square inverse time migration imaging from 2d to 3D is realized.(3)The wave-field reconstruction technology based on efficient boundary storage strategy and the wave-field reconstruction technology based on Checkpointing technology are realized,and the two wave-field reconstruction technology is applied to reverse time migration and least square reverse time migration.The results show that the application of the two wave-field reconstruction technology can effectively reduce the storage in the imaging process.(4)The parallel computing technology of MPI+CUDA is introduced into the least squares reverse time migration,which effectively improves the computing efficiency of least squares reverse time migration.(4)MPI+CUDA parallel computing technology is introduced into the least square inverse time migration,which effectively improves the computational efficiency of the least square inverse time migration. |
