Study And Application Of Missing Seismic Information Restruction Method

Entering the 21 st century, the exploration and application of energy resources have affected all aspects of daily life. As the gold blood of the industry among various resources, the oil and gas play an important role for national economy, society development and defense security. The rapid development of economy boosts the demand of the oil and gas, and promotes the development of oil-gas exploration in high-tech fields. Geophysical techniques improve rapidly and new theories of seismic geology come up constantly.The complicated geological structure and hard exploration environment bring severe challenges to the series work of seismic data processing and interpretation. Due to the limitation of environment and instruments, the seismic data tend to lose many important information, such as sampling data and high-frequency, falling short on requirements of delicate seismic processing and interpretation technology. The thesis proposes the restructing lost seismic information algorithms from two aspects of seismic interpolation restruction and seismic bandwidth extension addressing the problem of incomplete sampling data and lost high-frequency. The major work is shown as follows:Firstly, the thesis improves the traditional POCS algorithm. The traditional POCS is easy to implement and achieve good restuction of seismic data, but the method computes inefficiently and converges slowly. So we improve the threshold strategy from three aspects such as descending model, threshold function and stopping criterion. In addition, we modify the procedure of the method combined with the previous research simplifying redundant steps and improve computational efficiency. Synthetic data and real data experiments demonstrated the validity and universality of the proposed algorithm. Moreover, we research the application of the proposed algorithm in higher dimension seismic data interpolation.Secondly, the thesis proposes spatial-correlation extending bandwidth algorithm on non-stationary seismic signal. The earth filtering effects caused the loss of high-frequency and vertical resolution. Additionally, seismic signal is non-stationary signal whose frequency spectrum changes as time changes. The thesis proposes an extending bandwidth algorithm which combines the S transformation and attenuation theory and correlation idea maintaining the horizontal consistency. The experiments compensate high-frequency and improve resolution.