Reservoir Interpretation Method Based on Bayesian Discriminant Analysis and Logistic Regression

Abstract

The invention discloses a method for explaining oil and gas layers based on Bayesian discriminant analysis and logistic regression, which comprises the following steps: 1) collecting relevant data of existing oil test conclusions that belong to the same oil and gas reservoir as the layer to be evaluated; 2) collecting 3) Eliminate the difference in magnitude and dimension between different data; 4) Use the standardized data to conduct discriminant analysis on the reservoir to obtain the coefficients and constants of each discriminant function; Substituting the transformed data of the layer and the layer to be evaluated into the discriminant function for calculation; 6) Carrying out disordered multi-category logistic regression based on the data of each layer with known oil testing conclusions, and building a logistic regression model for each layer; 7) The discriminant function value, oil level, and spectrogram shape data of the layer to be evaluated are substituted into the logistic regression model; 8) The nature of the layer to be evaluated is judged according to the principle of maximum membership. The invention can meet the requirement of precise interpretation and evaluation of oil and gas layers, and the interpretation result is more scientific.

Description

technical field [0001] The invention relates to an evaluation method of oil and gas water layers in petroleum and natural gas reservoirs, in particular to an oil and gas layer interpretation method based on Bayes discriminant analysis and logistic regression. Background technique [0002] Accurate evaluation of oil and gas reservoirs is the ultimate goal of mud logging and logging work, which requires a lot of manpower and material resources, and there are various evaluation methods. In the mud logging field, there are gas logging, rock pyrolysis logging, quantitative fluorescence logging, light Hydrocarbon logging, etc. In the field of logging, there are resistivity logging, acoustic logging, imaging logging, etc. Each method can directly detect and derive many parameters, and these data include both quantitative data and qualitative classification data , play their respective roles in oil and gas reservoir evaluation. Any technology has both advantages and blind spots. Fo...

AI Technical Summary

Problems solved by technology

[0003] 1. The current interpretation and evaluation of oil and gas reservoirs is mostly based on single data, such as gas logging data interpretation, well logging interpretation, geochemical interpretation, etc. Although each single interpretation also refers to other data, it is mostly based on experience and cognition The degree of comprehensive application varies from person to person

[0004] 2. Only quantitative data can be calculated, and qualitative data cannot participate in the calculation

[0005] 3. A single technology can only detect part of the oil and gas information, and cannot reflect the overall characteristics of oil and gas reservoirs

[0006] 4. The evaluation results are ambiguous

Images