Intelligent prediction method and system for reservoir parameters of volcanic rock oil reservoir

Abstract

The invention relates to an intelligent prediction method and system for reservoir parameters of a volcanic rock oil reservoir. The method comprises the steps that firstly data preprocessing is performed on known logging information of the volcanic rock oil reservoir according to the logging curve characteristics, and logging curve data reaching a certain threshold with the reservoir parameter correlation of the volcanic rock oil reservoir is extracted from data pretreatment results to form training set data and the remaining logging curve data forms test set data; based on the training set data, a plurality of regression algorithms are combined with a big data machine learning algorithm to automatically establish regression estimation models between the reservoir parameters of the volcanic rock oil reservoir and the logging curve characteristics, and the regression estimation models are subjected to automatic optimization of the model parameters based on the multiple regression algorithms; the test set data is used for testing and evaluating the optimized regression estimation models, and the unqualified models are subjected to automatic optimization of the model parameters untilthe regression estimation model is qualified by inspection and evaluation; and finally, the regression estimation model qualified by the inspection and evaluation is used for carrying out the intelligent prediction of the unknown reservoir parameters, and the prediction result has high precision and small errors.

Description

technical field [0001] The invention relates to the technical field of prediction of geological engineering reservoir parameters, in particular to an intelligent prediction method and system for reservoir parameters of volcanic rock reservoirs. Background technique [0002] At present, in reservoir exploration and development technology, the research on the distribution law of reservoir parameters (also called reservoir physical properties, including porosity, permeability, saturation, pore structure, etc.) is the core of reservoir description. It is also an important parameter and basis in reservoir evaluation research, so reservoir parameter prediction is of great significance in reservoir exploration and development. However, in the current reservoir exploration and development technology, the existing various reservoir parameter prediction methods have certain limitations, and generally have the problems of low accuracy and large errors. From a mathematical point of vie...

AI Technical Summary

Problems solved by technology

[0016] To sum up, the existing technologies and methods for predicting reservoir parameters of volcanic reservoirs still face key problems: the application of technical methods is single, and there is no complete, comprehensive and applicable reservoir parameter prediction method for volcanic reservoirs. The method of reasonably processing reservoir logging information, the accuracy of prediction results is still low

Images