CN110163442A - A kind of gas well plug-ging prediction technique based on integrated study - Google Patents
A kind of gas well plug-ging prediction technique based on integrated study Download PDFInfo
- Publication number
- CN110163442A CN110163442A CN201910444971.8A CN201910444971A CN110163442A CN 110163442 A CN110163442 A CN 110163442A CN 201910444971 A CN201910444971 A CN 201910444971A CN 110163442 A CN110163442 A CN 110163442A
- Authority
- CN
- China
- Prior art keywords
- gas well
- ging
- data
- gas
- well plug
- 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 39
- 238000004519 manufacturing process Methods 0.000 claims abstract description 19
- 238000006116 polymerization reaction Methods 0.000 claims abstract description 17
- 238000013461 design Methods 0.000 claims abstract description 4
- 238000000605 extraction Methods 0.000 claims abstract description 4
- 238000004422 calculation algorithm Methods 0.000 claims description 28
- 238000012360 testing method Methods 0.000 claims description 14
- 238000004364 calculation method Methods 0.000 claims description 13
- 239000007788 liquid Substances 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 11
- XQCFHQBGMWUEMY-ZPUQHVIOSA-N Nitrovin Chemical compound C=1C=C([N+]([O-])=O)OC=1\C=C\C(=NNC(=N)N)\C=C\C1=CC=C([N+]([O-])=O)O1 XQCFHQBGMWUEMY-ZPUQHVIOSA-N 0.000 claims description 6
- 238000010801 machine learning Methods 0.000 claims description 4
- 238000004590 computer program Methods 0.000 claims description 3
- 238000002790 cross-validation Methods 0.000 claims description 3
- 238000007637 random forest analysis Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 claims description 3
- 230000004931 aggregating effect Effects 0.000 claims 1
- 238000012545 processing Methods 0.000 claims 1
- 241000521257 Hydrops Species 0.000 abstract description 14
- 206010030113 Oedema Diseases 0.000 abstract description 14
- 238000011161 development Methods 0.000 abstract description 12
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000011084 recovery Methods 0.000 abstract description 4
- 230000018109 developmental process Effects 0.000 description 10
- 238000011160 research Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 239000002245 particle Substances 0.000 description 6
- 230000008569 process Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- DGAQECJNVWCQMB-PUAWFVPOSA-M Ilexoside XXIX Chemical compound C[C@@H]1CC[C@@]2(CC[C@@]3(C(=CC[C@H]4[C@]3(CC[C@@H]5[C@@]4(CC[C@@H](C5(C)C)OS(=O)(=O)[O-])C)C)[C@@H]2[C@]1(C)O)C)C(=O)O[C@H]6[C@@H]([C@H]([C@@H]([C@H](O6)CO)O)O)O.[Na+] DGAQECJNVWCQMB-PUAWFVPOSA-M 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 239000007791 liquid phase Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000029052 metamorphosis Effects 0.000 description 1
- 239000012071 phase Substances 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 229910052708 sodium Inorganic materials 0.000 description 1
- 239000011734 sodium Substances 0.000 description 1
- 230000005514 two-phase flow Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/259—Fusion by voting
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Marketing (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Tourism & Hospitality (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Business, Economics & Management (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Primary Health Care (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Agronomy & Crop Science (AREA)
- Animal Husbandry (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The gas well plug-ging prediction technique based on integrated study that the invention discloses a kind of, the present invention specifically includes the following steps: S1, gas well initial data acquisition, the pretreatment of S2, Characteristic Extraction and data, the design of S3, base classifier, S4, ballot polymerization, the judgement of S5, gas well plug-ging state, the present invention relates to gas field development technical fields.The gas well plug-ging prediction technique based on integrated study, there is stronger specific aim and accuracy compared with the gas well plug-ging prediction model of traditional universality, it not only can more accurately judge the hydrops situation of gas well, there is better accuracy and convenience relative to traditional gas well plug-ging model, targetedly gas well plug-ging is obtained to the production status in different gas fields simultaneously and analyzes result, drainage gas production technology is instructed to construct preferably to reach discharge shaft bottom hydrops, restore gas well capacity, further excavate the later development potentiality of gas reservoir, achieve the purpose that the recovery ratio for improving gas reservoir.
Description
Technical field
The present invention relates to gas field development technical field, specially a kind of gas well plug-ging prediction technique based on integrated study.
Background technique
Current gas well plug-ging prediction technique is mainly based upon the critical flow physical theory model of individual particle drop for boundary
Face tension and drag coefficient are constant, and analyze in conjunction with empirical equation the hydrops situation of gas well, but are answered actual
With the calculated result of different models in the process, there are biggish deviations, and the general applicability of calculation method is not strong, to actual
On the one hand the reason of production operation and management bring different degrees of inconvenience, cause error is that liquid shape is had ignored in computation model
Caused by the resistance coefficient of state variation, another aspect is exactly that theoretical model is all using single drop as research object thus by drag force
Coefficient is to be set as fixed value, has ignored the metamorphosis of drop and is imitated by the droplet particles group for influencing each other and generating between drop
Should be so as to cause the calculating deviation of drag coefficient, traditional calculation method is mainly to a certain stage mistake in certain type gas fields
The research method that gets of analysis and conclusion toward data explain the shape of a certain stage gas well plug-ging in another type gas field
Condition, while the analysis of available data is not utilized rationally, gas field is directly influenced to the research of a certain gas field gas well plug-ging
How reasonable development makes more targeted, more reasonable, the more efficient weight for becoming each gas field and facing of gas well plug-ging prediction work
Problem is wanted, therefore we have the characteristics that enhancing recognizer generalization ability using using integrated study, proposed a kind of based on collection
It has solved the above problems at the gas well plug-ging prediction technique of study.
Summary of the invention
(1) the technical issues of solving
In view of the deficiencies of the prior art, the gas well plug-ging prediction technique based on integrated study that the present invention provides a kind of, solution
The calculated result of existing method of having determined different models in actual application process is there are biggish deviation, and calculation method
General applicability is not strong, brings different degrees of inconvenience, while traditional calculation method master to actual production operation and management
If the research method and conclusions that get to the analysis with the passing data of a certain stage in certain type gas fields explain other one
The situation of a certain stage gas well plug-ging in a type gas field, while the analysis of available data is not utilized rationally, thus to certain
The problem of research of one gas field gas well plug-ging directly influences gas field reasonable development.
(2) technical solution
In order to achieve the above object, the present invention is achieved by the following technical programs: a kind of gas well based on integrated study
Hydrops prediction technique, specifically includes the following steps:
The acquisition of S1, gas well initial data: gas well plug-ging simulation test data or connection gas field Production database are obtained, is mentioned
The creation data of individual well is taken, and data set is saved as to the format of csv file;
The pretreatment of S2, Characteristic Extraction and data: the data bi-directional scaling for first acquiring step S1 is allowed to fall into
One small specific sections removes the unit limitation of data, and is translated into nondimensional pure values, in order to not commensurate
Or the index of magnitude is able to carry out and compares and weight, while carrying out to missing values, invalid value and the repetition values that initial data is concentrated
Cleaning, correct or understand error value, removal data unit and scope limitation, to be translated into dimensionless, variance and
It is worth consistent pure values, for the feature of not commensurate or magnitude, is able to carry out and compares and weighted calculation, carried out with calculation formula
Data prediction, in pretreated regulation, the mean value of all data points is 0 for each attribute or each column, and variance is
1, random division is carried out to the collection containing label data handled well, is divided into training set and test set, and data set input vector is turned
It is changed to matrix;
The design of S3, base classifier: after the pretreatment for completing step S2 data, optimal gas well plug-ging is pre- in order to obtain
Model is surveyed, preferable three classifier algorithms of performance is selected, data is input in classifier algorithm in a manner of matrix, are passed through
Three machine learning algorithm models provide three sorter models for subsequent integrated study, and each base classifier algorithm passes through super
The way of search of parameter automatic search module obtains optimal classification device, after obtaining each base sorter model, needs to each
Base sorter model is verified;
S4, ballot polymerization: the verifying collection data matrix after step S2 pretreatment is input to step S3 trained three
In kind base sorter model, then by output as a result, including classification and weight as new input being input to ballot polymerization
In classifier, it is polymerize by way of ballot, the sorter model after polymerization obtains final result;
S5, gas well plug-ging state judge: entering data into after corresponding computer program output parameter as gas well
Predicted state.
Preferably, the creation data of monocrystalline includes gas production, oil jacket pressure, casing programme, payzone depth in the step S1
Degree, gas production, Liquid output and tubing size.
Preferably, the feature vector selected in the step S2 has: pay zone depth, pressure, gas production, Liquid output and oil pipe
Size.
Preferably, the calculation formula of data prediction is in the step S2WithY in formulaiIt represents " dimensionless, variance and the consistent pure values of mean value ", xiIt represents " individual data ",It represents " average value ", n represents " data number ", behalf " standard deviation ".
Preferably, training set accounts for 70% in the step S2, and test set accounts for 30%, and the dimension of matrix is (- 1,5), dimension
- 1 represents training set or test set data volume sum in degree.
Preferably, three classifier algorithms include that random forests algorithm, limit tree algorithm and guidance are poly- in the step S3
Hop algorithm.
Preferably, cross validation or leaving-one method are used in the step S3, to be verified to each base sorter model,
To determine the judging result of sorter model.
Preferably, the ballot polymerization in the step S4 is using the ballot mode for having weight, i.e., by three kinds of model predictions
It as a result is the average value of a certain class weight as standard, as soft ballot.
(3) beneficial effect
The gas well plug-ging prediction technique based on integrated study that the present invention provides a kind of.Have compared with prior art following
The utility model has the advantages that
(1), it is somebody's turn to do the gas well plug-ging prediction technique based on integrated study, with traditional critical flow based on individual particle drop
Physical theory model compares, and has abandoned traditional single Physical Modeling, is extracted by the algorithm of integrated study main
Creation data parameter as feature vector, optimized parameter is obtained in the way of grid search and forms base classifier, then
Submodel is polymerize by the way of soft ballot, finally the classifier mould after integrated study ballot classifier polymerization
Type exports final gas well plug-ging judging result, which can carry out the gas just for the gas field development situation of acquisition data
The gas well plug-ging in field is predicted, with stronger specific aim and accurately compared with the gas well plug-ging prediction model of traditional universality
Property, it not only can more accurately judge the hydrops situation of gas well, while can obtain to the production status in different gas fields
Targetedly gas well plug-ging analysis as a result, preferably instruct drainage gas production technology to construct with reach discharge shaft bottom hydrops, it is extensive
Multiple gas well capacity, further excavates the later development potentiality of gas reservoir, achievees the purpose that the recovery ratio for improving gas reservoir.
(2), it is somebody's turn to do the gas well plug-ging prediction technique based on integrated study, with traditional critical flow based on individual particle drop
Physical theory model compares, and gets rid of the limitation of traditional gas well plug-ging physical model, is not simply to research object liquid
Drop or liquid mould are analyzed, but use main creation data parameter as feature vector, are collected to gas well plug-ging situation
At study, there is better accuracy and convenience relative to traditional gas well plug-ging model.
(3), should gas well plug-ging prediction technique based on integrated study, by can be into for different gas field development situations
The different gas well plug-gings of row predict training set, have stronger specific aim compared with the gas well plug-ging prediction model of traditional universality
And adaptability.
Detailed description of the invention
Fig. 1 is flow chart of the method for the present invention figure;
Fig. 2 is the gas well plug-ging prediction result contrast table figure of present invention comparison case.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
The embodiment of the present invention provides a kind of technical solution referring to FIG. 1-2: a kind of gas well plug-ging based on integrated study is pre-
Survey method, specifically includes the following steps:
The acquisition of S1, gas well initial data: gas well plug-ging simulation test data or connection gas field Production database are obtained, is mentioned
The creation data of individual well is taken, and data set is saved as to the format of csv file;
The pretreatment of S2, Characteristic Extraction and data: since the precision and dimension of the test data of different gas wells may not
Unanimously, before being classified, data need to be pre-processed.The data bi-directional scaling that step S1 is acquired first, is allowed to fall
Enter a small specific sections, remove the unit limitation of data, and be translated into nondimensional pure values, in order to different lists
The index of position or magnitude is able to carry out and compares and weight, at the same missing values, invalid value and the repetition values that initial data is concentrated into
Row cleaning, corrects or understands error value, during gas liquid two-phase flow, gas phase and liquid phase are always carried out in different fluidised forms
Transformation, the key point of transformation are tolerance, liquid measure and pressure, and the purpose of gas well plug-ging prediction is that prediction is by the energy of gas
No drop to be drained into well head from shaft bottom, the empirical equation and actual conditions of bottom pressure source physical model have certain difference,
Well head pressure can obtain directly and accurately from apparatus, while the calculating parameter for influencing gas flow rate and flow is also
There are pressure influence, such as Reynolds number, drag coefficient.Therefore the data cases of real case of the invention are directed to, in removal data
Unit and scope limitation, dimensionless, variance and the consistent pure values of mean value are translated into, for not commensurate or magnitude
Feature, be able to carry out and compare and weighted calculation, carry out data prediction with calculation formula, in pretreated regulation, for
The mean value of all data points is 0 for each attribute or each column, variance 1, is carried out to the collection containing label data handled well random
It divides, is divided into training set and test set, and data set input vector is converted into matrix;
The design of S3, base classifier: after the pretreatment for completing step S2 data, optimal gas well plug-ging is pre- in order to obtain
Model is surveyed, preferable three classifier algorithms of performance is selected, data is input in classifier algorithm in a manner of matrix, are passed through
Three machine learning algorithm models provide three sorter models for subsequent integrated study, and each base classifier algorithm passes through super
The way of search of parameter automatic search module obtains optimal classification device, after obtaining each base sorter model, needs to each
Base sorter model is verified;
S4, ballot polymerization: the verifying collection data matrix after step S2 pretreatment is input to step S3 trained three
In kind base sorter model, then by output as a result, including classification and weight as new input being input to ballot polymerization
In classifier, it is polymerize by way of ballot, the sorter model after polymerization obtains final result, for gas well
Hydrops forecasting problem, each machine learning model performance is all very excellent, but passes through point after integrated study ballot polymerization
The accuracy rate of class device model model more single than other will be high.So Ensemble Learning Algorithms are asked for solving prediction gas well plug-ging
It inscribes and is more suitable for opposite other machines learning algorithm;
S5, gas well plug-ging state judge: entering data into after corresponding computer program output parameter as gas well
Predicted state, for simpler expression gas well state, setting " -1 " represent the non-hydrops of gas well, " 0 " represents the critical hydrops of gas well,
" 1 " represents gas well hydrops.
In the present invention, in step S1 the creation data of monocrystalline include gas production, oil jacket pressure, casing programme, pay zone depth,
Gas production, Liquid output and tubing size.
In the present invention, the feature vector selected in step S2 has: pay zone depth, pressure, gas production, Liquid output and oil pipe ruler
It is very little.
In the present invention, the calculation formula of data prediction is in step S2WithY in formulaiIt represents " dimensionless, variance and the consistent pure values of mean value ", xiIt represents " individual data ",It represents " average value ", n represents " data number ", behalf " standard deviation ".
In the present invention, training set accounts for 70% in step S2, and test set accounts for 30%, and the dimension of matrix is (- 1,5), dimension
In -1 represent training set or test set data volume sum.
In the present invention, three classifier algorithms include random forests algorithm, limit tree algorithm and guidance polymerization in step S3
Algorithm.
In the present invention, cross validation or leaving-one method are used in step S3, to be verified to each base sorter model, with
Determine the judging result of sorter model.
In the present invention, the ballot polymerization in step S4 is using the ballot mode for having weight, i.e., by three kinds of model prediction knots
Fruit is the average value of a certain class weight as standard, as soft ballot.
Compare case
For technical solution of the present invention and the advantage of preferably withdrawing deposit, examination is simulated with the famous special sodium gas well plug-ging of the industry
For testing, displaying is compared, as shown in table Fig. 2.
Such as table Fig. 2 it is found that prediction result of the invention is substantially better than traditional gas well plug-ging prediction technique, with actual observation
Hydrops situation degree of correspondence it is higher, so, this method not only simplifies the mechanism study of traditional complexity, and gas well plug-ging is pre-
The result of survey is higher, provides effective theory support and rationally guidance for the Efficient Development in gas field and whole decision and deployment,
Compared with traditional critical flow physical theory model based on individual particle drop, traditional single physical model has been abandoned
Method extracts main creation data parameter as feature vector, in the way of grid search by the algorithm of integrated study
It obtains optimized parameter and forms base classifier, then submodel is polymerize by the way of soft ballot, finally by integrated
Sorter model after study ballot classifier polymerization exports final gas well plug-ging judging result, and the judging result is just for adopting
The gas field development situation of collection data can carry out the gas well plug-ging prediction in the gas field, predict with the gas well plug-ging of traditional universality
Model, which is compared, has stronger specific aim and accuracy, not only can more accurately judge the hydrops situation of gas well, together
When the production status in different gas fields can be obtained targetedly gas well plug-ging analysis as a result, preferably instructing water drainage-gas recovery technology
Technology construction restores gas well capacity, further excavates the later development potentiality of gas reservoir, reach and mention to reach discharge shaft bottom hydrops
The purpose of the recovery ratio of high gas reservoir, while compared with traditional critical flow physical theory model based on individual particle drop,
The limitation for getting rid of traditional gas well plug-ging physical model, be not research object drop or liquid mould are analyzed simply, and
It is using main creation data parameter as feature vector, integrated study is carried out to gas well plug-ging situation, relative to traditional
Gas well plug-ging model has better accuracy and convenience, by that can carry out different gas for different gas field development situations
Well hydrops predicts training set, has stronger specific aim and adaptation compared with the gas well plug-ging prediction model of traditional universality
Property.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.
It although an embodiment of the present invention has been shown and described, for the ordinary skill in the art, can be with
A variety of variations, modification, replacement can be carried out to these embodiments without departing from the principles and spirit of the present invention by understanding
And modification, the scope of the present invention is defined by the appended.
Claims (8)
1. a kind of gas well plug-ging prediction technique based on integrated study, it is characterised in that: specifically includes the following steps:
The acquisition of S1, gas well initial data: gas well plug-ging simulation test data or connection gas field Production database are obtained, is extracted single
The creation data of well, and data set is saved as to the format of csv file;
The pretreatment of S2, Characteristic Extraction and data: the data bi-directional scaling for first acquiring step S1 is allowed to fall into one
Small specific sections remove the unit limitation of data, and are translated into nondimensional pure values, in order to not commensurate or amount
The index of grade, which is able to carry out, to be compared and weights, while missing values, invalid value and the repetition values concentrated to initial data are cleaned,
It corrects or understands error value, in the unit and scope limitation of removal data, it is consistent with mean value to be translated into dimensionless, variance
Pure values the feature of not commensurate or magnitude is able to carry out and is compared and weighted calculation, it is pre- to carry out data with calculation formula
Processing, in pretreated regulation, the mean value of all data points is 0 for each attribute or each column, variance 1, to place
The collection containing label data managed carries out random division, is divided into training set and test set, and data set input vector is converted to square
Battle array;
The design of S3, base classifier: after the pretreatment for completing step S2 data, optimal gas well plug-ging predicts mould in order to obtain
Type selects preferable three classifier algorithms of performance, data is input in classifier algorithm in a manner of matrix, pass through three
Machine learning algorithm model provides three sorter models for subsequent integrated study, and each base classifier algorithm passes through hyper parameter
The way of search of automatic search module obtains optimal classification device, after obtaining each base sorter model, needs to each base point
Class device model is verified;
S4, ballot polymerization: the verifying collection data matrix after step S2 pretreatment is input to the trained three kinds of bases of step S3
In sorter model, then by output as a result, including classification and weight as new input is input to ballot polymerization classification
In device, it is polymerize by way of ballot, the sorter model after polymerization obtains final result;
S5, gas well plug-ging state judge: entering data into after corresponding computer program output parameter as the prediction of gas well
State.
2. a kind of gas well plug-ging prediction technique based on integrated study according to claim 1, it is characterised in that: the step
The creation data of monocrystalline includes gas production, oil jacket pressure, casing programme, pay zone depth, gas production, Liquid output and oil pipe in rapid S1
Size.
3. a kind of gas well plug-ging prediction technique based on integrated study according to claim 1, it is characterised in that: the step
The feature vector selected in rapid S2 has: pay zone depth, pressure, gas production, Liquid output and tubing size.
4. a kind of gas well plug-ging prediction technique based on integrated study according to claim 1, it is characterised in that: the step
The calculation formula of data prediction is in rapid S2WithY in formulaiGeneration
Table " dimensionless, variance and the consistent pure values of mean value ", xiIt represents " individual data ",It represents " average value ", n represents " data number
Mesh ", behalf " standard deviation ".
5. a kind of gas well plug-ging prediction technique based on integrated study according to claim 1, it is characterised in that: the step
Training set accounts for 70% in rapid S2, and test set accounts for 30%, and the dimension of matrix is (- 1,5), and -1 represents training set or test in dimension
Collect data volume sum.
6. a kind of gas well plug-ging prediction technique based on integrated study according to claim 1, it is characterised in that: the step
Three classifier algorithms include random forests algorithm, limit tree algorithm and guidance aggregating algorithm in rapid S3.
7. a kind of gas well plug-ging prediction technique based on integrated study according to claim 1, it is characterised in that: the step
Cross validation or leaving-one method are used in rapid S3, to verify to each base sorter model, to determine sentencing for sorter model
Disconnected result.
8. a kind of gas well plug-ging prediction technique based on integrated study according to claim 1, it is characterised in that: the step
Ballot polymerization in rapid S4 is using the ballot mode for having weight, i.e., is the flat of a certain class weight by three kinds of model prediction results
Mean value is as standard, as soft ballot.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910444971.8A CN110163442A (en) | 2019-05-27 | 2019-05-27 | A kind of gas well plug-ging prediction technique based on integrated study |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910444971.8A CN110163442A (en) | 2019-05-27 | 2019-05-27 | A kind of gas well plug-ging prediction technique based on integrated study |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110163442A true CN110163442A (en) | 2019-08-23 |
Family
ID=67629018
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910444971.8A Pending CN110163442A (en) | 2019-05-27 | 2019-05-27 | A kind of gas well plug-ging prediction technique based on integrated study |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110163442A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111598724A (en) * | 2020-05-19 | 2020-08-28 | 四川革什扎水电开发有限责任公司 | Time-interval integration method for day-ahead prediction of warehousing flow of small and medium-sized reservoirs |
CN111784669A (en) * | 2020-06-30 | 2020-10-16 | 长沙理工大学 | Capsule endoscopy image multi-focus detection method |
CN112855127A (en) * | 2019-11-28 | 2021-05-28 | 北京国双科技有限公司 | Gas well accumulated liquid identification method and device |
Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6305216B1 (en) * | 1999-12-21 | 2001-10-23 | Production Testing Services | Method and apparatus for predicting the fluid characteristics in a well hole |
US20050228905A1 (en) * | 2004-03-17 | 2005-10-13 | Schlumberger Technology Corporation | Method and apparatus and program storage device adapted for automatic qualitative and quantitative risk assesssment based on technical wellbore design and earth properties |
US20080270328A1 (en) * | 2006-10-18 | 2008-10-30 | Chad Lafferty | Building and Using Intelligent Software Agents For Optimizing Oil And Gas Wells |
CN103590812A (en) * | 2013-10-21 | 2014-02-19 | 中国石油天然气股份有限公司 | Calculation method, calculation device and determination method for gas well liquid accumulation amount |
CN103670352A (en) * | 2012-09-18 | 2014-03-26 | 中国石油天然气股份有限公司 | Automatic control method for removing accumulated liquid in gas well |
CN104504611A (en) * | 2014-12-11 | 2015-04-08 | 中国石油天然气股份有限公司 | Method for determining whether liquid is accumulated in gas well and liquid accumulation degree of gas well |
CN104504604A (en) * | 2014-12-12 | 2015-04-08 | 中国地质大学(武汉) | Method for qualifying liquid loading of gas well |
US20160063402A1 (en) * | 2014-08-27 | 2016-03-03 | Digital H2O, Inc. | Oilfield water management |
US20160145994A1 (en) * | 2014-11-20 | 2016-05-26 | Petrochina Company Limited | Evaluation Method and Evaluation Device for Water Breakthrough Risk of Production Wells in Aquifer Drive Gas Reservoirs |
CN106246164A (en) * | 2016-08-29 | 2016-12-21 | 孟玲 | Coal bed gas well level monitoring system based on Fibre Optical Sensor |
CN106600442A (en) * | 2016-12-06 | 2017-04-26 | 中国海洋石油总公司 | Critical liquid carrying condition optimization method in deep water gas well testing |
CN107045671A (en) * | 2017-03-22 | 2017-08-15 | 重庆科技学院 | Water-producing gas well hydrops Risk Forecast Method |
CN108023876A (en) * | 2017-11-20 | 2018-05-11 | 西安电子科技大学 | Intrusion detection method and intruding detection system based on sustainability integrated study |
CN108021769A (en) * | 2017-12-25 | 2018-05-11 | 中国石油天然气股份有限公司 | Method for judging whether gas well normally carries liquid or not |
CN108351985A (en) * | 2015-06-30 | 2018-07-31 | 亚利桑那州立大学董事会 | Method and apparatus for large-scale machines study |
CN109026130A (en) * | 2018-08-17 | 2018-12-18 | 西安科技大学 | A kind of recognition methods of mine gas data exception |
WO2018236238A1 (en) * | 2017-06-20 | 2018-12-27 | Schlumberger Technology B.V. | Predicting wellbore flow performance |
CN109145948A (en) * | 2018-07-18 | 2019-01-04 | 宁波沙塔信息技术有限公司 | A kind of injection molding machine putty method for detecting abnormality based on integrated study |
KR101975436B1 (en) * | 2018-10-25 | 2019-05-07 | 동아대학교 산학협력단 | Apparatus and method for forecasting production for shale gas well in transient flow using machine learning method |
CN109741797A (en) * | 2018-12-10 | 2019-05-10 | 中国药科大学 | A method of small molecule compound water solubility grade is predicted using depth learning technology |
-
2019
- 2019-05-27 CN CN201910444971.8A patent/CN110163442A/en active Pending
Patent Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6305216B1 (en) * | 1999-12-21 | 2001-10-23 | Production Testing Services | Method and apparatus for predicting the fluid characteristics in a well hole |
US20050228905A1 (en) * | 2004-03-17 | 2005-10-13 | Schlumberger Technology Corporation | Method and apparatus and program storage device adapted for automatic qualitative and quantitative risk assesssment based on technical wellbore design and earth properties |
US20080270328A1 (en) * | 2006-10-18 | 2008-10-30 | Chad Lafferty | Building and Using Intelligent Software Agents For Optimizing Oil And Gas Wells |
CN103670352A (en) * | 2012-09-18 | 2014-03-26 | 中国石油天然气股份有限公司 | Automatic control method for removing accumulated liquid in gas well |
CN103590812A (en) * | 2013-10-21 | 2014-02-19 | 中国石油天然气股份有限公司 | Calculation method, calculation device and determination method for gas well liquid accumulation amount |
US20160063402A1 (en) * | 2014-08-27 | 2016-03-03 | Digital H2O, Inc. | Oilfield water management |
US20160145994A1 (en) * | 2014-11-20 | 2016-05-26 | Petrochina Company Limited | Evaluation Method and Evaluation Device for Water Breakthrough Risk of Production Wells in Aquifer Drive Gas Reservoirs |
CN104504611A (en) * | 2014-12-11 | 2015-04-08 | 中国石油天然气股份有限公司 | Method for determining whether liquid is accumulated in gas well and liquid accumulation degree of gas well |
CN104504604A (en) * | 2014-12-12 | 2015-04-08 | 中国地质大学(武汉) | Method for qualifying liquid loading of gas well |
CN108351985A (en) * | 2015-06-30 | 2018-07-31 | 亚利桑那州立大学董事会 | Method and apparatus for large-scale machines study |
CN106246164A (en) * | 2016-08-29 | 2016-12-21 | 孟玲 | Coal bed gas well level monitoring system based on Fibre Optical Sensor |
CN106600442A (en) * | 2016-12-06 | 2017-04-26 | 中国海洋石油总公司 | Critical liquid carrying condition optimization method in deep water gas well testing |
CN107045671A (en) * | 2017-03-22 | 2017-08-15 | 重庆科技学院 | Water-producing gas well hydrops Risk Forecast Method |
WO2018236238A1 (en) * | 2017-06-20 | 2018-12-27 | Schlumberger Technology B.V. | Predicting wellbore flow performance |
CN108023876A (en) * | 2017-11-20 | 2018-05-11 | 西安电子科技大学 | Intrusion detection method and intruding detection system based on sustainability integrated study |
CN108021769A (en) * | 2017-12-25 | 2018-05-11 | 中国石油天然气股份有限公司 | Method for judging whether gas well normally carries liquid or not |
CN109145948A (en) * | 2018-07-18 | 2019-01-04 | 宁波沙塔信息技术有限公司 | A kind of injection molding machine putty method for detecting abnormality based on integrated study |
CN109026130A (en) * | 2018-08-17 | 2018-12-18 | 西安科技大学 | A kind of recognition methods of mine gas data exception |
KR101975436B1 (en) * | 2018-10-25 | 2019-05-07 | 동아대학교 산학협력단 | Apparatus and method for forecasting production for shale gas well in transient flow using machine learning method |
CN109741797A (en) * | 2018-12-10 | 2019-05-10 | 中国药科大学 | A method of small molecule compound water solubility grade is predicted using depth learning technology |
Non-Patent Citations (8)
Title |
---|
DAVID A. WOOD: "A transparent Open-Box learning network provides insight to complex systems and a performance benchmark for more-opaque machine learning algorithms", 《ADVANCES IN GEO-ENERGY RESEARCH 》 * |
JAMES F. LEA: "《Gas Well Deliquification》", 12 February 2019 * |
康成瑞等: "天然气井井筒积液预测方法解析", 《新疆石油天然气》 * |
朱钰等: "《统计学》", 31 August 2012 * |
李品等: "气井井筒积液预测模型优选", 《当代化工》 * |
杜敬国等: "气井连续携液模型对比研究及新模型的现场验证", 《兰州石化职业技术学院学报》 * |
栾国华等: "应用人工神经网络方法预测气井积液", 《断块油气田》 * |
谢双喜等: "渤海气田气井井筒积液预测方法分析", 《长江大学学报(自科版)》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112855127A (en) * | 2019-11-28 | 2021-05-28 | 北京国双科技有限公司 | Gas well accumulated liquid identification method and device |
CN111598724A (en) * | 2020-05-19 | 2020-08-28 | 四川革什扎水电开发有限责任公司 | Time-interval integration method for day-ahead prediction of warehousing flow of small and medium-sized reservoirs |
CN111598724B (en) * | 2020-05-19 | 2022-07-22 | 四川革什扎水电开发有限责任公司 | Time-interval integration method for day-ahead prediction of warehousing flow of small and medium reservoirs |
CN111784669A (en) * | 2020-06-30 | 2020-10-16 | 长沙理工大学 | Capsule endoscopy image multi-focus detection method |
CN111784669B (en) * | 2020-06-30 | 2024-04-02 | 长沙理工大学 | Multi-range detection method for capsule endoscopic images |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Schwanghart et al. | TopoToolbox 2–MATLAB-based software for topographic analysis and modeling in Earth surface sciences | |
CN102362262B (en) | System and method for characterizing fractures in a subsurface reservoir | |
CN110163442A (en) | A kind of gas well plug-ging prediction technique based on integrated study | |
CN108985380B (en) | Point switch fault identification method based on cluster integration | |
Aldrich et al. | Analysis and visualization of discrete fracture networks using a flow topology graph | |
CN101379498A (en) | Methods, systems, and computer-readable media for fast updating of oil and gas field production models with physical and proxy simulators | |
CN105760673A (en) | Fluvial facies reservoir earthquake sensitive parameter template analysis method | |
CN109241627A (en) | The dynamic shoring method of probability hierarchical and the device of Automated Design supporting scheme | |
Alenezi et al. | A data-driven smart proxy model for a comprehensive reservoir simulation | |
CN105452598B (en) | The method for selecting and optimizing the oil field control for yield platform | |
Ali et al. | Clustering sparse sensor placement identification and deep learning based forecasting for wind turbine wakes | |
CN105893669A (en) | Global simulation performance predication method based on data digging | |
CN101718834A (en) | Method for analysing analog circuit fault propagation characteristic | |
CN118246714B (en) | Water conservancy and hydropower engineering construction energy consumption analysis method and system | |
Kolajoobi et al. | Investigating the capability of data-driven proxy models as solution for reservoir geological uncertainty quantification | |
CN115438823A (en) | Borehole wall instability mechanism analysis and prediction method and system | |
CN112085242B (en) | Low-resistance layer accurate prediction method and device based on big data deep learning method | |
CN102364490A (en) | Automatic synchronization recognition method based on hierarchical analyzing model | |
Ali et al. | Classification of the Reynolds stress anisotropy tensor in very large thermally stratified wind farms using colormap image segmentation | |
CN116911216B (en) | Reservoir oil well productivity factor assessment and prediction method | |
CN109034179A (en) | A kind of rock stratum classification method based on mahalanobis distance IDTW | |
CN108182178A (en) | Groundwater level analysis method and system based on event text data mining | |
Gu et al. | Calibration of traffic flow fundamental diagrams for network simulation applications: A two-stage clustering approach | |
CN110486009A (en) | A kind of automatic reverse method of parameter and system on infinity stratum | |
CN113970073B (en) | ResNet-based water supply network leakage accurate positioning method |
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 |