Time-Frequency Analysis In The Application Of Glutenite Reservoir Prediction Research

Seventy-five of oil resources and ninety percent the oil production in China are distributed in the east, the Cenozoic continental fault basin. Continental down-faulted basin phase in a large amount of glutenite fan body especially in fault depression period, forming a complex glutenite reservoirs. At present a lot of oil and gas basin gradually explore the glutenite fan body reservoirs as the main exploration targets.Gravel rock reservoir has the characteristics of strong heterogeneity, multip hase superimposed, buried deep, the large difficulty of reservoir evaluation. In order to better development of hydrocarbon reservoirs, increase oil production, and enhance the accuracy of reservoir simulation, it need to not only accurately identify the nature of the reservoir but also to predict reservoir thickness and spatial distribution. so the geophysicist urgently need to explore the higher resolution and accuracy higher seismic processing and interpretat io n techniques.Seismic wave dynamics characteristics such as amplitude, frequency, phase is in a good way to depict the reservoir distribution. Conventional seismic interpretation is separate in time domain and frequency domain. And the time-frequency analysis technology combines time domain and frequency domain to describe the seismic signal, to a certain extent, which can improve the accuracy of seismic interpretation.This thesis make a scrutiny into the basic principle of time-frequency analysis technique and different time-frequency analysis methods, embarking from the theory of non-stationar y signal. The thesis analyzed theoretically the advantages and disadvantages of different time-frequency analysis methods and the relationship. Besides To better comparative analysis the advantages and disadvantages of several time-frequency analysis methods, this thesis respectively established the theory of signal and the actual signal and compares and analyzes time- frequency resolution and its influencing factors of different algorithms.Short-time Fourier transform and continuous wavelet time-frequency analysis results directly affected by the selection basis function with low resolution. The Wigner distribut io n has higher time-frequency resolution, but for the cross terms of non-stationary signal, it makes the spectrum error of the results. Sparse pulse inversion algorithm under the constraint of regularization parameter resolution higher time-frequency distribution can be obtained.Using the advantage of high resolution sparse constraint inversion algorithm and apply the algorithm to ST region gravel rock reservoir prediction effectively. The study of area frequency analysis shows that 30 hz seismic section has the most consecutive event. Buried depth is more shal ow section of the effective distribution of glutenite reservoir in near of hollow faults, and large buried depth interval effective glutenite reservoir distribution in the area of distance control depression fault distance.By the theory and application research of time-frequency analysis method,it conclusions that time-frequency analysis method to a certain extent, can improve the signalto-noise ratio of seismic data, the frequency domain of seismic data resolution is higher than time domain seismic resolution and the time-frequency analysis of frequency domain properties can be more fine depict glutenite reservoir distribution than conventional time-domain properties, which has practical implications to glutenite reservoir prediction.