CN-121997014-A - Method and system for analyzing master control parameters of acid fracturing effect of carbonate

CN121997014ACN 121997014 ACN121997014 ACN 121997014ACN-121997014-A

Abstract

The invention discloses a method and a system for analyzing main control parameters of a carbonate acid fracturing effect, which relate to the technical field of acidification and solve the problem that main control parameters are difficult to identify in the prior carbonate acid fracturing effect evaluation; the method comprises the steps of executing a random forest algorithm and a recursion elimination algorithm based on parameters to obtain first weights and second weights of the parameters, carrying out weighted average on the first weights and the second weights to obtain comprehensive weights of the parameters, selecting the parameters according to the comprehensive weights to construct a plurality of schemes, inputting the schemes into a BP neural network to simulate the yield to obtain the correlation between the actual yield and the predicted yield, selecting the parameters contained in the scheme with the highest correlation as main control parameters of acid fracturing effects, and analyzing the main control parameters of the acid fracturing effects of the carbonate based on the random forest, the recursion elimination and the BP neural network algorithm.

Inventors

CHEN WEIHUA
JIA YUCHENG
WANG HANCHENG
YANG MIAO
XIE CHEN
FU YAN

Assignees

中国石油天然气股份有限公司

Dates

Publication Date: 20260508
Application Date: 20241105

Claims (10)

1. The method for analyzing the master control parameters of the acid fracturing effect of the carbonate is characterized by comprising the following steps of: S1, acquiring parameter data influencing the acid fracturing effect of carbonate, preprocessing the parameter data, calculating the correlation of each parameter based on the parameter data, and filtering out the parameter with extremely strong correlation; s2, executing a random forest algorithm and a recursion elimination algorithm based on the filtered parameters to obtain a first weight and a second weight of each parameter; s3, carrying out weighted average on the first weight and the second weight of each parameter to obtain the comprehensive weight of each parameter; S4, selecting parameters according to the sequence from high to low of comprehensive weights to construct a plurality of schemes, inputting the schemes into the BP neural network to simulate the yield to obtain the correlation between the actual yield and the predicted yield, and selecting the parameters contained in the scheme with the highest correlation as the main control parameters of the acid fracturing effect.
2. The method for analyzing master control parameters of a carbonate acid fracturing effect according to claim 1, wherein S1 comprises: Selecting parameters affecting the acid fracturing effect of the carbonate from three aspects of engineering parameters, geological parameters and well information parameters; Collecting parameter data, deleting the parameter with unique value, supplementing the data with missing parameter, and identifying and correcting the data with abnormal parameter; And calculating the pearson correlation coefficient among the parameters based on the parameter data, and filtering out the parameters of which the pearson correlation coefficient is in an extremely strong correlation interval.
3. The method for analyzing master control parameters of a carbonate acid fracturing effect according to claim 1, wherein S2 comprises: Training a random forest model by taking the parameters as the characteristics, and taking the importance weight distributed by the trained random forest model for each characteristic as a first weight of the corresponding parameters; And eliminating the least important features according to the importance weights of the features, retraining the random forest model based on the remaining features until the stopping condition is met, and taking the importance weight distributed by the trained random forest model for each feature as a second weight of the corresponding parameter.
4. The method for analyzing master control parameters of a carbonate acid fracturing effect according to claim 1, wherein S3 comprises: Normalizing the first weight and the second weight of the parameter; and carrying out weighted average based on the normalized first weight and the normalized second weight to obtain the comprehensive weight of the parameter.
5. The method for analyzing master control parameters of a carbonate acid fracturing effect according to claim 1, wherein S4 comprises: selecting parameter construction schemes according to the sequence of the comprehensive weights from high to low, wherein the number of the selected parameters is gradually increased until all the parameters are selected, so as to obtain a plurality of schemes; Inputting parameter data of the scheme into a BP neural network to obtain predicted yield, and calculating the correlation between the predicted yield and the actual yield; And selecting the parameter contained in the scheme with highest correlation as the main control parameter of the acid fracturing effect.
6. The utility model provides a carbonate acid fracturing effect master control parameter analysis system which characterized in that includes: The data processing unit is used for acquiring parameter data affecting the acid fracturing effect of the carbonate rock, preprocessing the parameter data, calculating the correlation of each parameter based on the parameter data and filtering out the parameter with extremely strong correlation; the weight analysis unit is used for executing a random forest algorithm and a recursion elimination algorithm based on the filtered parameters to obtain a first weight and a second weight of each parameter; the comprehensive weight analysis unit is used for carrying out weighted average on the first weight and the second weight of each parameter to obtain the comprehensive weight of each parameter; And the main control parameter analysis unit is used for selecting parameters according to the sequence from high to low of comprehensive weights to construct a plurality of schemes, inputting the schemes into the BP neural network to simulate the yield to obtain the correlation between the actual yield and the predicted yield, and selecting the parameters contained in the scheme with the highest correlation as main control parameters of the acid fracturing effect.
7. The carbonate acid fracturing effect master control parameter analysis system according to claim 6, wherein the data processing unit specifically comprises: a parameter selection unit for selecting parameters affecting the acid pressure effect of the carbonate from three aspects of engineering parameters, geological parameters and well information parameters; The data preprocessing unit is used for collecting parameter data, deleting the parameter with unique value, supplementing the data with missing parameter, and identifying and correcting the data with abnormal parameter; And the data filtering unit is used for calculating the pearson correlation coefficient between the parameters based on the parameter data and filtering out the parameters of which the pearson correlation coefficient is in an extremely strong correlation interval.
8. The carbonate acid fracturing effect master control parameter analysis system according to claim 6, wherein the weight analysis unit specifically comprises: The first weight analysis unit is used for training a random forest model by taking the parameters as the characteristics, and taking the importance weight distributed by the trained random forest model for each characteristic as the first weight of the corresponding parameters; And the second weight analysis unit is used for eliminating the least important features according to the importance weights of the features, retraining the random forest model based on the remaining features until the stopping condition is met, and taking the importance weights distributed by the trained random forest model for each feature as second weights of corresponding parameters.
9. The carbonate acid fracturing effect master control parameter analysis system according to claim 6, wherein the comprehensive weight analysis unit specifically comprises: The weight normalization unit is used for normalizing the first weight and the second weight of the parameter; and the comprehensive weight calculation unit is used for carrying out weighted average based on the normalized first weight and the normalized second weight to obtain the comprehensive weight of the parameter.
10. The system for analyzing master control parameters of the acid fracturing effect of carbonate according to claim 6, wherein the master control parameter analyzing unit specifically comprises: The equation construction unit is used for selecting parameter construction schemes according to the sequence of the comprehensive weights from high to low, and the number of the selected parameters is gradually increased until all the parameters are selected, so that a plurality of schemes are obtained; the scheme prediction unit is used for inputting parameter data of the scheme into the BP neural network to obtain predicted yield, and calculating the correlation between the predicted yield and the actual yield; And the parameter selection unit is used for selecting parameters contained in the scheme with highest correlation as main control parameters of the acid fracturing effect.

Description

Method and system for analyzing master control parameters of acid fracturing effect of carbonate Technical Field The invention relates to the technical field of acidification, in particular to a method and a system for analyzing main control parameters of a carbonate acid fracturing effect. Background The carbonate reservoir has low permeability, poor physical properties, crack development, strong heterogeneity and low initial yield, and acid fracturing measures are the most main technical means for improving the productivity of a gas well. With the wide use of acid fracturing technology, the difficulty of acid fracturing construction is increased, the construction success rate is reduced, and the construction cost is increased day by day, so that the acid fracturing effect evaluation is also important. For acid fracturing effect evaluation, the parameters influencing the acid fracturing effect are numerous, and if the main control parameters of the acid fracturing effect can be selected, the analyzed parameters can be reduced in a dimension reduction mode while the evaluation accuracy of the acid fracturing effect is maintained. However, in the current evaluation process of the acid fracturing effect of the carbonate rock oil and gas reservoir, it is difficult to accurately identify and analyze the main control parameters affecting the acid fracturing effect. Therefore, the application provides a method and a system for analyzing master control parameters of a carbonate acid fracturing effect, which solve the problems. Disclosure of Invention The application aims to provide a method and a system for analyzing main control parameters of a carbonate acid fracturing effect, which solve the problem that main control parameters affecting the acid fracturing effect are difficult to accurately identify and analyze in the existing carbonate oil and gas reservoir acid fracturing effect evaluation process. The application provides a method for analyzing main control parameters of a carbonate acid fracturing effect, which comprises the steps of S1, obtaining parameter data affecting the carbonate acid fracturing effect, preprocessing the parameter data, calculating the correlation of each parameter based on the parameter data, filtering out the parameter with extremely strong correlation, S2, executing a random forest algorithm and a recursion elimination algorithm based on the filtered parameter to obtain a first weight and a second weight of each parameter, S3, carrying out weighted average on the first weight and the second weight of each parameter to obtain the comprehensive weight of each parameter, S4, selecting the parameters according to the sequence from high to low of the comprehensive weight, constructing a plurality of schemes, inputting the schemes into a BP neural network to simulate the yield to obtain the correlation of the actual yield and the predicted yield, and selecting the parameters contained in the scheme with the highest correlation as the main control parameters of the acid fracturing effect. In one possible implementation mode, S1 comprises the steps of selecting parameters affecting the acid pressure effect of carbonate from three aspects of engineering parameters, geological parameters and well information parameters, collecting parameter data, deleting the parameters with unique values, supplementing the data with the missing parameters, identifying and correcting the data with the abnormal parameters, calculating the Pearson correlation coefficient among the parameters based on the parameter data, and filtering out the parameters with the Pearson correlation coefficient in an extremely strong correlation interval. In one possible implementation, S2 comprises training a random forest model by taking parameters as features, taking importance weights distributed by the trained random forest model for each feature as first weights of corresponding parameters, eliminating least important features according to the importance weights of the features, retraining the random forest model based on the remaining features until stopping conditions are met, and taking the importance weights distributed by the trained random forest model for each feature as second weights of corresponding parameters. In one possible implementation, the step S3 includes normalizing the first weight and the second weight of the parameter, and carrying out weighted average based on the normalized first weight and second weight to obtain the comprehensive weight of the parameter. In a possible implementation manner, the S4 comprises the steps of selecting parameter construction schemes according to the sequence of the comprehensive weights from high to low, gradually increasing the number of the selected parameters until all the parameters are selected to obtain a plurality of schemes, inputting parameter data of the schemes into a BP neural network to obtain predicted yield, calculating the correlation between the predic