Fluvial facies reservoir earthquake sensitive parameter template analysis method

Abstract

The invention relates to a fluvial facies reservoir earthquake sensitive parameter template analysis method.The method comprises the following steps that a to-be-measured position is determined, seismic attributes are extracted according to a preset time window, smooth and filtering preprocessing is performed on the seismic attributes, abnormal values in the seismic attribute data are removed, the seismic attributes are extracted, optimal sensitive seismic attribute parameters are obtained, optimization is performed, a forward modeling analysis method, a multi-side-face and multi-angle reservoir related compatibility analysis method and a decision optimization and combination descending dimension method are adopted for obtaining an optimization attribute set with the most sensitive reservoir parameter, and compatibility reduction of the seismic attributes is achieved.According to the analysis method, the reservoir sensitive seismic attribute combination is effectively built on the river equality complicated geographic condition, and a seismic sensitive parameter template drawing is drawn according to the obtained optimized seismic attributes; the seismic sensitive parameter template pattern is contracted with a known sample well or a known reservoir information coordinate template, and finally the characteristic property of a determined reservoir within a measured region is finally determined.

Description

technical field [0001] The invention relates to a template analysis method for seismic parameters, in particular to a template analysis method for seismic sensitive parameters of fluvial reservoirs. Background technique [0002] At present, in order to improve the accuracy of reservoir prediction, the attribute optimization method is gradually introduced in order to find the most sensitive and representative sensitive attributes to solve the problem of reservoir prediction. However, there is often a certain correlation between seismic attributes, and the selected single seismic sensitive attribute may only reflect a certain aspect or a certain local geological information of the work area, and may even only appear as an accidental phenomenon , the result of reservoir prediction may have multiple solutions, and it is difficult to obtain the optimal and most accurate prediction result. [0003] There are some complex relationships between seismic attributes and reservoir info...

AI Technical Summary

Problems solved by technology

[0004] When using the neural network for reservoir prediction, in order to obtain higher-precision reservoir prediction results, certain requirements are required for the samples participating in the training. On the one hand, a sufficient number of samples is required, that is, the number of known sample wells must reach a certain number; On the other hand, the distribution of sample wells in the work area should be as uniform as possible. If the distribution of sample wells is too concentrated, it may cause low prediction accuracy in areas with few wells or large distances between wells.

At the same time, it is necessary to consider the various particularities of the training samples. If all the samples participating in the learning are treated equally, it will definitely have a certain impact on the results and cause errors.

Images