CN118607082A - Prediction method for casing deformation in fracturing process based on deep learning - Google Patents
Prediction method for casing deformation in fracturing process based on deep learning Download PDFInfo
- Publication number
- CN118607082A CN118607082A CN202411088244.XA CN202411088244A CN118607082A CN 118607082 A CN118607082 A CN 118607082A CN 202411088244 A CN202411088244 A CN 202411088244A CN 118607082 A CN118607082 A CN 118607082A
- Authority
- CN
- China
- Prior art keywords
- parameters
- parameter
- prediction model
- deformation
- fracturing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000008569 process Effects 0.000 title claims abstract description 26
- 238000013135 deep learning Methods 0.000 title claims abstract description 15
- 238000012549 training Methods 0.000 claims abstract description 18
- 238000013136 deep learning model Methods 0.000 claims abstract description 4
- 230000006870 function Effects 0.000 claims description 14
- 230000004913 activation Effects 0.000 claims description 8
- 238000011176 pooling Methods 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000012795 verification Methods 0.000 claims description 4
- 238000009933 burial Methods 0.000 claims description 3
- 238000006073 displacement reaction Methods 0.000 claims description 3
- 239000012530 fluid Substances 0.000 claims description 3
- 239000004576 sand Substances 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 230000007812 deficiency Effects 0.000 claims 2
- 238000012545 processing Methods 0.000 abstract description 7
- 238000011161 development Methods 0.000 abstract description 5
- 238000010606 normalization Methods 0.000 abstract description 3
- 238000012217 deletion Methods 0.000 description 16
- 230000037430 deletion Effects 0.000 description 16
- 230000003044 adaptive effect Effects 0.000 description 3
- 230000007423 decrease Effects 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of shale gas development, and particularly discloses a prediction method of casing deformation in a fracturing process based on deep learning, which comprises the following steps: collecting geological parameters, engineering parameters and casing deformation of a fractured gas well, and constructing an original casing data set; carrying out missing parameter value processing on input parameters in the original set of variable data; then, the correlation between the input parameters and the output parameters is analyzed and processed; finally, carrying out normalization processing to obtain a final set data set; and establishing a sleeve deformation prediction model based on the deep learning model, training and verifying the sleeve deformation prediction model through a final sleeve deformation data set to obtain a final sleeve deformation prediction model, and predicting the sleeve deformation of the fracturing gas well through the final sleeve deformation prediction model. According to the invention, the original set-change data set is processed, so that the data quality is effectively improved, a data base is provided for the constructed set-change prediction model, and the prediction precision of the set-change prediction model is effectively improved.
Description
Technical Field
The invention belongs to the technical field of shale gas development, and particularly relates to a prediction method for casing deformation in a fracturing process based on deep learning.
Background
At present, volume fracturing is used as one of core technologies for efficient development of shale gas, and when reservoir transformation is carried out, the problem of casing deformation frequently occurs, and the fracturing effect and the shale gas development process are seriously influenced, so that the method has great significance for predicting casing deformation in order to ensure smooth progress of fracturing construction.
And the factors such as geological engineering, development and the like have nonlinear, uncertainty and time variability among the factors as a result of long-term comprehensive actions when the sleeve is deformed. The traditional sleeve deformation prediction method is mainly based on an empirical formula or a mechanical model, and is difficult to achieve various influence factors, so that the prediction accuracy of sleeve deformation is low, and the cost is high.
Machine learning, which obtains valuable insight based on collected data, can help to make quick and correct decisions, has been widely used in other industries in petroleum engineering, among which is also the application of machine learning to casing deformation prediction, but its research depth is insufficient, resulting in poor prediction accuracy. Therefore, we propose a prediction method for casing deformation in a deep learning based fracturing process.
Disclosure of Invention
The invention aims to solve the technical problems of the prior art, and provides a prediction method for sleeve deformation in the fracturing process based on deep learning, which can effectively improve the precision of sleeve deformation prediction, thereby improving the safety in the fracturing construction process.
The technical scheme adopted by the invention is as follows: a prediction method of casing deformation in a fracturing process based on deep learning comprises the following steps:
Step 1: collecting geological parameters, engineering parameters and casing deformation of a fractured gas well, and constructing an original casing data set;
step 2: screening geological parameters and engineering parameters in the original set of variable data, setting a missing parameter threshold, calculating a parameter missing ratio of each geological parameter and engineering parameter, eliminating the geological parameters and engineering parameters with the parameter missing ratio being greater than the missing parameter threshold, and filling the geological parameters and engineering parameters with the parameter missing ratio being less than the missing parameter threshold and not being 0 by adopting a KNN algorithm to obtain a first-level set of variable data;
Step 3: the geological parameters and engineering parameters in the primary set of variable data are screened again, a coefficient threshold is set, the maximum mutual information coefficient of each geological parameter and engineering parameter is calculated, and the geological parameters and engineering parameters with the maximum mutual information coefficient smaller than the coefficient threshold are removed to obtain a secondary set of variable data;
Step 4: normalizing the secondary set variable data set to obtain a final set variable data set;
Step 5: randomly dividing a set-change data set according to a ratio of 3:1, wherein 75% is a training set, 25% is a testing set, and establishing a set-change prediction model based on a deep learning model, wherein the set-change prediction model comprises an input layer, a convolution layer, an activation function, a pooling layer, a convolution layer, a pooling layer, a full-connection layer, an activation function layer and an output layer which are sequentially arranged; the training set is input into the sleeve transformer prediction model to train the sleeve transformer prediction model, the test set is input into the trained sleeve transformer prediction model to be verified, and the final sleeve transformer prediction model is obtained after the verification is qualified;
Step 6: and obtaining the geological parameters and engineering parameters of the fracturing gas well, inputting the obtained geological parameters and engineering parameters into a final sleeve deformation prediction model to obtain the sleeve deformation, and predicting the sleeve deformation of the fracturing gas well through the sleeve deformation.
Preferably, the geological parameters comprise fracture or fault inclination angle, fracture or fault length, distance from fracture or fault to set transformation point, horizontal section burial depth, maximum horizontal ground stress and minimum horizontal ground stress; the engineering parameters include fracturing displacement, fracturing fluid volume, fracturing time, fracturing segment length, fracturing cluster number, reverse drainage rate and sand ratio.
Preferably, the parameter deletion ratio is a ratio of a parameter deletion sample size to a parameter total sample size.
Preferably, in the training process of the set-variant prediction model in step 5, a mean square error is used as a loss function.
Preferably, in the step 5, the adaptive adjustment learning rate is added in the training process of the set-variant prediction model.
Compared with the prior art, the invention has the beneficial effects that: the method has the advantages that the constructed original set data set is subjected to missing parameter value processing, correlation analysis processing of input parameters and output parameters and normalization processing in sequence, so that the data quality is effectively improved, and a quantized data base is provided for the establishment of a set-change prediction model; meanwhile, based on deep learning analysis of nonlinear relations between geological parameters and engineering parameters and deformation of the sleeve, prediction accuracy of deformation of the sleeve is improved, and self-adaptive adjustment learning rate is added in a training stage of the sleeve deformation prediction model, so that the sleeve deformation prediction model can adapt to changes in the training process better, and performance and generalization capability of the sleeve deformation prediction model are improved.
The method and the device effectively improve the data quality by processing the collected original set data set, provide a good data basis for the set-change prediction model constructed by deep learning, and effectively improve the prediction precision of the set-change prediction model.
Drawings
FIG. 1 is a flow chart of the present invention.
Detailed Description
The invention will be described in further detail with reference to the accompanying drawings and specific examples.
Examples:
as shown in fig. 1, the prediction method for casing deformation in the deep learning-based fracturing process provided in this embodiment includes the following steps:
step 1: collecting geological parameters, engineering parameters and casing deformation of the fractured gas well;
The geological parameters comprise a fracture or fault inclination angle, a fracture or fault length, a distance from the fracture or fault to a sleeve change point, a horizontal section burial depth, a maximum horizontal ground stress and a minimum horizontal ground stress;
the engineering parameters comprise fracturing displacement, fracturing fluid quantity, fracturing time, fracturing segment length, fracturing cluster number, reverse drainage rate and sand ratio;
Taking each geological parameter and engineering parameter as an ordinate, taking the deformation of the sleeve corresponding to each group of parameters as an abscissa, and constructing an original sleeve variable data set, wherein the geological parameter and engineering parameter are taken as input parameters, and the deformation of the sleeve is taken as output parameters;
when a fractured gas well is selected, besides the fracturing gas well data with the deformation of the sleeve, the fracturing gas well data without the deformation of the sleeve is selected, and the deformation of the sleeve in the fracturing gas well data without the deformation of the sleeve is recorded as 0;
step 2: because most of the collected geological parameters and engineering parameters have data missing, the data missing affects the accuracy of a subsequent prediction model, so that missing data in an original set of data set is required to be processed;
Setting a missing threshold value and defining a parameter missing ratio The ratio of the parameter missing sample size to the parameter total sample size is:
Wherein, In order to obtain the parameter deletion ratio,For the input parameter to miss the sample size,The total sample size is the input parameter; for example, 100 total samples of parameters of fault dip angle parameters in input parameters are obtained, the number of missing parameters is 20, namely, 20 missing samples of the parameters are obtained, and the missing ratio of the parameters is calculated to be 20%;
the missing threshold is generally set to 60% -80%, and can be adjusted according to actual conditions;
Calculating to obtain a parameter deletion ratio of each input parameter, sequencing from large to small, and eliminating input parameters of which the parameter deletion ratio is larger than a deletion threshold value; the input parameters with the parameter deletion ratio larger than the deletion threshold value indicate that the input parameters have more information deletion in the collected data set, if the input parameters with more deletion are filled, the original data is easy to be distorted, the prediction accuracy of a prediction model is affected, so that the input parameters with the parameter deletion ratio larger than the deletion threshold value are removed, and a primary data sample consisting of T effective input parameters is obtained;
and for the input parameters with the parameter deletion ratio larger than zero and smaller than the deletion threshold value, adopting a KNN algorithm to carry out data filling on the deletion parameter values of the input parameters in each data sample in the primary data sample, thereby obtaining a primary set of variable data sets, wherein the data filling of the deletion parameter values is specifically as follows:
setting the K value in a KNN algorithm, training an original set-change data set based on the KNN algorithm to obtain K neighbor samples of each original set-change data sample, and obtaining a trained KNN model;
Respectively inputting each data sample with a missing parameter value in the primary data sample into a trained KNN model to obtain K neighbor samples corresponding to each data sample, wherein the missing parameter value of the input parameter in each data sample is the average value of the corresponding input parameters in the K neighbor samples;
for example: the s-th primary data sample containing T valid input parameters =（, , ,…, ) The component corresponding to the t-th input parameter in (a)Absence, i.e.Obtaining the first-level data sample according to KNN algorithm for the missing parameter value of the t-th input parameterIs the K nearest neighbor samples of:，，，…， I.e. =（, , ,…, ) K e [1, K ], calculating a missing parameter value of the t-th input parameter using an average of the t-th input parameters of the K nearest neighbor samples：
Namely, calculating the missing parameter value of the T-th input parameter in the s-th primary data sample containing T effective input parameters;
Step 3: some of the remaining input parameters after the screening in the step 2 have weak correlation with the output parameters, and the input parameters with weak correlation with the output parameters influence the training efficiency, namely the prediction accuracy, of the subsequent prediction model, so that the input parameters with weak correlation with the output parameters need to be removed;
Setting a coefficient threshold, calculating a maximum mutual information coefficient between each input parameter and each output parameter, if the maximum mutual information coefficient is lower than the coefficient threshold, indicating that the correlation between the input parameter and the output parameter is not strong, and eliminating the input parameter of which the maximum mutual information coefficient is lower than the coefficient threshold, wherein the method specifically comprises the following steps:
sequentially selecting geological parameters and engineering parameters as input parameters, and recording as ; The deformation of the sleeve is taken as an output parameter and recorded as；
Input parameters in an m x n grid pairAnd output parametersTwo-dimensional grid division is performed to lead the grid to fall on the firstThe frequency of the data points in the grid is taken asIs estimated to fall on the firstThe frequency of the data points of the row is taken asFall on the firstData point frequency of rows asIs an estimate of (1), namely:
In the method, in the process of the invention, Representation ofAnd is also provided withIs used to determine the joint probability distribution of (1),Indicating that it falls on the firstThe number of data in the grid,Indicating that it falls on the firstThe number of data in the trellis;
Calculating input parameters And output parametersMutual information betweenThe method comprises the following steps:
In the method, in the process of the invention, Representation ofIs provided with a distribution of the edge probability of (c),Representation ofIs a boundary probability distribution of (1);
Calculating input parameters And output parametersMaximum mutual information coefficient betweenThe method comprises the following steps:
In the method, in the process of the invention, Indicating that the condition is satisfiedA maximum function at which F (n) is a function of n,Represents the smaller of m and n;
when the maximum mutual information coefficient is smaller than the coefficient threshold value, the corresponding input parameters are removed, so that a final input parameter combination, namely a secondary set transformation data set, is obtained;
step 4: in order to eliminate errors caused by inconsistent dimensions among different input parameters, normalization processing is carried out on data in the obtained secondary set of variable data to obtain a final set of variable data, wherein the method specifically comprises the following steps:
In the method, in the process of the invention, As the raw data is to be processed,For the data to be normalized,Representing parametersIs selected from the group consisting of a maximum value of (c),Representing parametersIs the minimum value of (a);
step 5: randomly dividing a final set-change data set according to a ratio of 3:1, wherein 75% is a training set, 25% is a testing set, and establishing a set-change prediction model based on a deep learning model, wherein the set-change prediction model comprises an input layer, a convolution layer, an activation function, a pooling layer, a convolution layer, a pooling layer, a full-connection layer, an activation function layer and an output layer which are sequentially arranged; wherein the activation function layer uses a ReLU as the activation function;
Inputting the training set into the set-change prediction model to train the set-change prediction model, and using the mean square error as a loss function in the training process, wherein the loss function is as follows:
In the method, in the process of the invention, In the case of a batch size of the product,Is the firstThe true values of the individual sets of data samples,Is the firstPredicted values of the individual sets of data samples;
Meanwhile, adding an adaptive adjustment learning rate in the training process, wherein the adaptive adjustment learning rate is as follows:
wherein, eta is the updated learning rate, eta 0 is the initial learning rate, For controlling the speed of learning rate decrease,The learning rate decreases more slowly as it approaches 1,The learning rate decreases faster as the learning rate approaches 0, and epoch is the training iteration number;
inputting the test set into a trained set-change prediction model for verification, and obtaining a final set-change prediction model after the verification is qualified;
Step 6: and obtaining the geological parameters and engineering parameters of the fracturing gas well, inputting the obtained geological parameters and engineering parameters into a final sleeve deformation prediction model to obtain the sleeve deformation, and predicting the sleeve deformation of the fracturing gas well through the sleeve deformation.
The foregoing is merely a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any modification and substitution based on the technical scheme and the inventive concept provided by the present invention should be covered in the scope of the present invention.
Claims (5)
1. The prediction method of the casing deformation in the fracturing process based on deep learning is characterized by comprising the following steps of:
Step 1: collecting geological parameters, engineering parameters and casing deformation of a fractured gas well, and constructing an original casing data set;
step 2: screening geological parameters and engineering parameters in the original set of variable data, setting a missing parameter threshold, calculating a parameter missing ratio of each geological parameter and engineering parameter, eliminating the geological parameters and engineering parameters with the parameter missing ratio being greater than the missing parameter threshold, and filling the geological parameters and engineering parameters with the parameter missing ratio being less than the missing parameter threshold and not being 0 by adopting a KNN algorithm to obtain a first-level set of variable data;
Step 3: the geological parameters and engineering parameters in the primary set of variable data are screened again, a coefficient threshold is set, the maximum mutual information coefficient of each geological parameter and engineering parameter is calculated, and the geological parameters and engineering parameters with the maximum mutual information coefficient smaller than the coefficient threshold are removed to obtain a secondary set of variable data;
Step 4: normalizing the secondary set variable data set to obtain a final set variable data set, and randomly dividing the final set variable data set according to the proportion of 3:1, wherein 75% is a training set and 25% is a testing set;
step 5: the method comprises the steps of establishing a set-variant prediction model based on a deep learning model, wherein the set-variant prediction model comprises an input layer, a convolution layer, an activation function, a pooling layer, a convolution layer, a pooling layer, a full connection layer, an activation function layer and an output layer which are sequentially arranged; the training set is input into the sleeve transformer prediction model to train the sleeve transformer prediction model, the test set is input into the trained sleeve transformer prediction model to be verified, and the final sleeve transformer prediction model is obtained after the verification is qualified;
Step 6: and obtaining the geological parameters and engineering parameters of the fracturing gas well, inputting the obtained geological parameters and engineering parameters into a final sleeve deformation prediction model to obtain the sleeve deformation, and predicting the sleeve deformation of the fracturing gas well through the sleeve deformation.
2. The method for predicting casing deformation in a deep learning based fracturing process of claim 1, wherein the geological parameters include fracture or fault inclination, fracture or fault length, distance of fracture or fault to casing deformation point, horizontal section burial depth, maximum horizontal ground stress and minimum horizontal ground stress; the engineering parameters include fracturing displacement, fracturing fluid volume, fracturing time, fracturing segment length, fracturing cluster number, reverse drainage rate and sand ratio.
3. The method for predicting casing deformation in a deep learning based fracturing process of claim 1, wherein the parameter deficiency ratio is a ratio of a parameter deficiency sample size to a parameter total sample size.
4. The method for predicting casing deformation in a deep learning based fracturing process according to claim 1, wherein in the training of the casing deformation prediction model in step 5, a mean square error is used as a loss function.
5. The method for predicting casing deformation in a deep learning-based fracturing process according to claim 1 or 4, wherein the learning rate is adaptively adjusted in the training process of the casing deformation prediction model in step 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411088244.XA CN118607082A (en) | 2024-08-09 | 2024-08-09 | Prediction method for casing deformation in fracturing process based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411088244.XA CN118607082A (en) | 2024-08-09 | 2024-08-09 | Prediction method for casing deformation in fracturing process based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118607082A true CN118607082A (en) | 2024-09-06 |
Family
ID=92561540
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411088244.XA Pending CN118607082A (en) | 2024-08-09 | 2024-08-09 | Prediction method for casing deformation in fracturing process based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118607082A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130259847A1 (en) * | 2012-04-02 | 2013-10-03 | Berg Pharma Llc | Interrogatory cell-based assays and uses thereof |
| CN111178611A (en) * | 2019-12-23 | 2020-05-19 | 广西电网有限责任公司 | Method for predicting daily electric quantity |
| US20210027770A1 (en) * | 2019-07-22 | 2021-01-28 | Capital One Services, Llc | Multi-turn dialogue response generation with persona modeling |
| CN112348168A (en) * | 2020-10-27 | 2021-02-09 | 国网四川省电力公司经济技术研究院 | Ultra-short-term load prediction method and system considering data loss and characteristic redundancy |
| CN114742262A (en) * | 2022-03-01 | 2022-07-12 | 山东鲁软数字科技有限公司 | Multi-dimensional parameter synchronous prediction method, prediction terminal and computer storage medium |
| CN115952620A (en) * | 2023-01-08 | 2023-04-11 | 西南石油大学 | Shale reservoir fracturing prevention construction method and key fracture control material |
| CN116011234A (en) * | 2023-01-13 | 2023-04-25 | 成都理工大学 | Pressure front sleeve change risk level judgment method integrating geomechanics and Bayes |
| KR102593981B1 (en) * | 2022-11-10 | 2023-10-25 | 주식회사 이노와이어리스 | method for processing missing value in network log data and analyzing root cause of communication defect thereby |
| CN117313919A (en) * | 2023-08-31 | 2023-12-29 | 中国石油天然气集团有限公司 | Horizontal well casing variable working condition prediction method, system, equipment and storage medium |
| CN117787074A (en) * | 2023-04-07 | 2024-03-29 | 福建工程学院 | Battery state of charge SOC prediction method combining convolutional network and sequence-to-sequence model |
-
2024
- 2024-08-09 CN CN202411088244.XA patent/CN118607082A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130259847A1 (en) * | 2012-04-02 | 2013-10-03 | Berg Pharma Llc | Interrogatory cell-based assays and uses thereof |
| US20210027770A1 (en) * | 2019-07-22 | 2021-01-28 | Capital One Services, Llc | Multi-turn dialogue response generation with persona modeling |
| CN111178611A (en) * | 2019-12-23 | 2020-05-19 | 广西电网有限责任公司 | Method for predicting daily electric quantity |
| CN112348168A (en) * | 2020-10-27 | 2021-02-09 | 国网四川省电力公司经济技术研究院 | Ultra-short-term load prediction method and system considering data loss and characteristic redundancy |
| CN114742262A (en) * | 2022-03-01 | 2022-07-12 | 山东鲁软数字科技有限公司 | Multi-dimensional parameter synchronous prediction method, prediction terminal and computer storage medium |
| KR102593981B1 (en) * | 2022-11-10 | 2023-10-25 | 주식회사 이노와이어리스 | method for processing missing value in network log data and analyzing root cause of communication defect thereby |
| CN115952620A (en) * | 2023-01-08 | 2023-04-11 | 西南石油大学 | Shale reservoir fracturing prevention construction method and key fracture control material |
| CN116011234A (en) * | 2023-01-13 | 2023-04-25 | 成都理工大学 | Pressure front sleeve change risk level judgment method integrating geomechanics and Bayes |
| CN117787074A (en) * | 2023-04-07 | 2024-03-29 | 福建工程学院 | Battery state of charge SOC prediction method combining convolutional network and sequence-to-sequence model |
| CN117313919A (en) * | 2023-08-31 | 2023-12-29 | 中国石油天然气集团有限公司 | Horizontal well casing variable working condition prediction method, system, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110363344B (en) | Probability integral parameter prediction method for optimizing BP neural network based on MIV-GP algorithm | |
| CN114595608B (en) | Fracturing construction parameter and working system parameter optimization method and system | |
| CN112364560B (en) | Intelligent prediction method for working hours of mine rock drilling equipment | |
| CN106677763B (en) | Dynamic integrated modeling-based oil well working fluid level prediction method | |
| CN110807544B (en) | Oil field residual oil saturation distribution prediction method based on machine learning | |
| CN112761628B (en) | Shale gas yield determination method and device based on long-term and short-term memory neural network | |
| CN113627639A (en) | Well testing productivity prediction method and system for carbonate fracture-cave reservoir | |
| CN113627068A (en) | Method and system for predicting well testing productivity of fracture-cavity type oil and gas reservoir | |
| CN114781951A (en) | Shale oil reservoir carbon dioxide huff-puff development well selection decision method and system | |
| CN110942182A (en) | Method for establishing typhoon prediction model based on support vector regression | |
| CN115185937A (en) | SA-GAN architecture-based time sequence anomaly detection method | |
| CN115982141A (en) | Characteristic optimization method for time series data prediction | |
| CN115860197A (en) | Data-driven coal bed gas yield prediction method and system | |
| CN117933497B (en) | TSA-ARIMA-CNN-based enterprise carbon emission prediction method | |
| CN118607082A (en) | Prediction method for casing deformation in fracturing process based on deep learning | |
| CN117078110A (en) | Reservoir engineering dessert evaluation method based on analytic hierarchy process | |
| CN116933631A (en) | Blasting block prediction method based on neural network optimization algorithm | |
| CN117253087A (en) | Rock type identification method and system based on convolutional neural network | |
| CN116842358A (en) | Soft measurement modeling method based on multi-scale convolution and self-adaptive feature fusion | |
| CN114862007A (en) | Short-period gas production rate prediction method and system for carbonate gas well | |
| CN115270623A (en) | Method for predicting embankment settlement | |
| CN116415625A (en) | Logging curve characteristic learning method and system based on deep neural network | |
| CN115759792A (en) | Horizontal well profile control and flooding effect analysis method based on whitening weighted gray model | |
| CN117522173B (en) | Natural gas hydrate depressurization exploitation productivity prediction method based on deep neural network | |
| CN115983088B (en) | Sea shale in-situ gas quantity prediction model established based on ensemble learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |