Intelligent reservoir type division method based on unbalanced samples

CN115422988BActive Publication Date: 2026-06-16CHINA NATIONAL OFFSHORE OIL (CHINA) CO LTD
1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
CHINA NATIONAL OFFSHORE OIL (CHINA) CO LTD
Filing Date
2022-07-22
Publication Date
2026-06-16

AI Technical Summary

⚠Technical Problem

Existing technologies suffer from low efficiency and insufficient accuracy in intelligent classification of reservoir types for unbalanced samples, especially when the number of effective reservoir samples is far less than the number of mudstone or non-reservoir samples, where traditional methods are not ideal.

⚗Method used

Cluster analysis is used to determine the range of physical parameters of the resampled samples, and limited resampling is performed within a reasonable range. The random forest method is then used to perform hierarchical partitioning to ensure the balance of sample quantity and physical meaning, and to reduce the impact of noise.

🎯Benefits of technology

It improves the accuracy and efficiency of intelligent reservoir type classification, ensures the balance of sample quantity and the physical meaning of data, reduces the impact of noise, and achieves high-precision reservoir type identification.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The application provides a reservoir type intelligent division method based on unbalanced samples, comprising the following steps: carrying out cluster analysis by using actual logging data, determining the cluster center value of each physical parameter of different types of reservoirs, and carrying out a limited number of resampling on each sample set within the range of 30% of the center value; recombining the reservoir types with relatively less sample type number but similar parameter characteristics in the unbalanced samples as a reservoir type and encoding the samples; taking characteristic values such as longitudinal wave velocity, transverse wave velocity, density, shale content and porosity as input, and using the random forest method to divide the reservoirs layer by layer and step by step; and the like. The reservoir type intelligent division method based on unbalanced samples determines the physical parameter range of the resampled samples, recombines the reservoir types with similar parameter characteristics, encodes the samples, divides the reservoir types layer by layer and step by step, and thus effectively solves the problem of intelligent division of the reservoir types of unbalanced samples.
Need to check novelty before this filing date? Find Prior Art