Porosity prediction method based on depth domain prestack angle gather

Abstract

The invention discloses a porosity prediction method based on a depth domain prestack angle gather, which belongs to the field of an oil and gas geophysical prospecting project. The porosity prediction method comprises the following steps: (1) based on logging data, seismic data and depth domain angle domain common imaging point gather data, seismic attributes of the depth domain are extracted; (2) re-sampling porosity data are acquired; (3) based on the seismic attributes and the re-sampling porosity data, a joint sample set is built, and the joint sample set is then subjected to standard deviation standardized processing to acquire standardized seismic attributes and the corresponding standardized porosity data; (4) the joint sample set is randomly divided into a training data set and a testing data set, and an optimal seismic attribute subset is acquired; (5) based on the optimal seismic attribute subset, a prediction model is built; (6) based on the testing data, the prediction model is detected to acquire a detection prediction model; and (7) as for all optimal seismic attribute subsets in a work area, through the detection prediction model, the porosity data in the work area can be predicted.

Description

technical field [0001] The invention relates to the field of oil and gas geophysical prospecting engineering, in particular to a porosity prediction method based on pre-stack angle gathers in depth domain. Background technique [0002] With the improvement of exploration requirements, the reservoir prediction technology based on post-stack attributes can no longer meet the current exploration needs of lithology, tight sandstone and other types of oil and gas reservoirs. At present, the extraction of pre-stack seismic attributes is mainly concentrated in the time domain. In the case of complex geological structures, the reservoir depth reflected in the time domain is obtained through inaccurate velocity information, and the error is bound to be large. [0003] In the past ten years, domestic reservoir parameter prediction research methods mainly include neural network, function approximation and geostatistical methods, and their different combination methods emerge in endless...

AI Technical Summary

Problems solved by technology

The above methods are all based on pre-stack seismic attributes in the time domain to predict reservoir parameters, making it impossible to meet the increasingly high precision requirements of oil and gas exploration.

Images