CN116911432A - Prediction method, storage medium and processor for horizontal well productivity - Google Patents
Prediction method, storage medium and processor for horizontal well productivity Download PDFInfo
- Publication number
- CN116911432A CN116911432A CN202310671981.1A CN202310671981A CN116911432A CN 116911432 A CN116911432 A CN 116911432A CN 202310671981 A CN202310671981 A CN 202310671981A CN 116911432 A CN116911432 A CN 116911432A
- Authority
- CN
- China
- Prior art keywords
- horizontal well
- data
- determining
- time period
- physical
- 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 87
- 238000003860 storage Methods 0.000 title claims abstract description 20
- 238000004519 manufacturing process Methods 0.000 claims abstract description 73
- 238000013528 artificial neural network Methods 0.000 claims abstract description 56
- 238000005457 optimization Methods 0.000 claims abstract description 22
- 238000007781 pre-processing Methods 0.000 claims abstract description 21
- 230000003247 decreasing effect Effects 0.000 claims description 13
- 230000004913 activation Effects 0.000 claims description 8
- 210000002569 neuron Anatomy 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 description 63
- 230000015654 memory Effects 0.000 description 20
- 238000010586 diagram Methods 0.000 description 18
- 239000003921 oil Substances 0.000 description 18
- 238000004590 computer program Methods 0.000 description 11
- 230000008569 process Effects 0.000 description 11
- 238000012549 training Methods 0.000 description 10
- 238000002790 cross-validation Methods 0.000 description 9
- 238000012360 testing method Methods 0.000 description 9
- 238000010801 machine learning Methods 0.000 description 8
- 238000010606 normalization Methods 0.000 description 7
- 239000004215 Carbon black (E152) Substances 0.000 description 6
- 229930195733 hydrocarbon Natural products 0.000 description 6
- 150000002430 hydrocarbons Chemical class 0.000 description 6
- 230000007774 longterm Effects 0.000 description 6
- 239000003208 petroleum Substances 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 239000011435 rock Substances 0.000 description 4
- 230000001186 cumulative effect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 239000012530 fluid Substances 0.000 description 3
- 238000003062 neural network model Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000010779 crude oil Substances 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005755 formation reaction Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000011065 in-situ storage Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000001537 neural effect Effects 0.000 description 2
- 239000003129 oil well Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000035699 permeability Effects 0.000 description 2
- 238000000053 physical method Methods 0.000 description 2
- 230000000638 stimulation Effects 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- BVKZGUZCCUSVTD-UHFFFAOYSA-L Carbonate Chemical compound [O-]C([O-])=O BVKZGUZCCUSVTD-UHFFFAOYSA-L 0.000 description 1
- 235000019738 Limestone Nutrition 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000011234 economic evaluation Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 230000007786 learning performance Effects 0.000 description 1
- 239000006028 limestone Substances 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 238000005293 physical law Methods 0.000 description 1
- 238000005086 pumping Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Mining & Mineral Resources (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Animal Husbandry (AREA)
- Agronomy & Crop Science (AREA)
- Quality & Reliability (AREA)
- Primary Health Care (AREA)
- Probability & Statistics with Applications (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Marine Sciences & Fisheries (AREA)
- Evolutionary Biology (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
Abstract
The embodiment of the application provides a prediction method, a storage medium and a processor for horizontal well productivity. Comprising the following steps: acquiring production data and physical constraint parameters of the horizontal well in a historical time period, wherein the physical constraint parameters comprise at least one of reservoir properties, logging curves, horizontal well information, fracturing parameters and well control changes; preprocessing production data and physical constraint parameters to obtain a preprocessed data set; determining a function equation of a prediction model by a physical driving neural network basic method; determining super parameters of the prediction model by a Bayesian optimization method to determine a model framework of the prediction model; constructing a prediction model according to the function equation and the model; dividing the preprocessing data set into a plurality of groups of data, and sequentially inputting each group of data into a prediction model; outputting the predicted capacity of the horizontal well in a predicted time period through a prediction model according to each group of data; and determining the actual predicted capacity of the horizontal well in the predicted time period according to the plurality of predicted capacities.
Description
Technical Field
The application relates to the technical field of oil and gas field development, in particular to a method for predicting capacity of a horizontal well, a capacity prediction system, a storage medium and a processor.
Background
Yield prediction is critical for economic exploration and development of non-conventional resources, especially in fracture design, economic evaluation and decision-making. However, predicting well dynamics remains challenging due to multiphase fluid flow and multi-scale complex fracture networks in tight oil formations, the physical laws behind which are not fully understood. With the rapid development of computer science and explosive accumulation of oilfield data, machine Learning (ML) has been widely used in the petroleum industry by virtue of its strong nonlinear approximation capability. The cumulative capacity of the first year is predicted by adopting random forests, self-adaptive enhancement, support vectors and a neural network algorithm and taking well information and fracturing yield increasing parameters as inputs. Formation thickness, depth, number of perforations, and log are used as inputs, and cumulative production for the first three months is used as an output. However, in the above ML model, the dependency between the input parameters and the target capacity generally depends on the data amount and quality. In addition, the prediction target is the accumulated capacity of fixed time, and the dynamic change of the oil well capacity along with time cannot be reflected.
To estimate dynamic time series capacity, recurrent neural networks and variants thereof, i.e., long and short term memory, gated recursive units, and bi-directional gated recursive units, are introduced into an ML-based method that can predict future capacity from previous historical capacity data. However, in practical applications of these pure ML methods for long-term productivity prediction, since the required data size increases with the complexity of the neural network, when the ML pattern is input into the input beyond the training data range, the output prediction result is inaccurate, and accurate prediction of the oil well productivity cannot be performed.
Disclosure of Invention
The embodiment of the application aims to provide a prediction method, a productivity prediction system, a storage medium and a processor for horizontal well productivity.
To achieve the above object, a first aspect of the present application provides a method for predicting capacity of a horizontal well, including:
acquiring production data and physical constraint parameters of the horizontal well in a historical time period, wherein the physical constraint parameters comprise at least one of reservoir properties, logging curves, horizontal well information, fracturing parameters and well control changes;
preprocessing production data and physical constraint parameters to obtain a preprocessed data set corresponding to the production data and the physical constraint parameters;
determining a function equation of a prediction model by a physical driving neural network basic method;
determining super parameters of the prediction model by a Bayesian optimization method to determine a model framework of the prediction model;
constructing a prediction model according to the function equation and the model;
dividing the preprocessing data set into a plurality of groups of data, and sequentially inputting each group of data into a prediction model;
outputting the predicted capacity of the horizontal well in a predicted time period through a prediction model according to each group of data;
and determining the actual predicted capacity of the horizontal well in the predicted time period according to the plurality of predicted capacities.
In one embodiment, determining the function equation of the predictive model by the physical driving neural network basic method includes: determining a physical loss term and a neural network loss term of a function equation; and determining a function equation of the prediction model according to the preset neural network loss term and the physical loss term.
In one embodiment, determining the physical loss term of the function equation includes determining according to equation (1):
wherein Lp (m, a) is a physical loss term,the predicted production energy is output by the Duong yield decreasing model in the historical time period, y is the actual production energy of the horizontal well in the historical time period, T is the historical time period, and a and m are constants.
In one embodiment, the method further comprises determining a Duong yield decremental model according to equation (2):
wherein q (T) is the real-time capacity of the horizontal well in the historical time period, N p For the accumulated capacity of the horizontal well in the historical time period, n is a crack time index, T is a preset time period, q i Is the initial capacity of the horizontal well over a historical period of time.
In one embodiment, determining the neural network loss term of the functional equation includes determining according to equation (3):
wherein L is data For the neural network to lose terms,and (3) outputting predicted production energy for the horizontal well in a historical time period through a Duong yield decreasing model, wherein y is the actual production energy of the horizontal well in the historical time period.
In one embodiment, determining the function equation of the predictive model from the predetermined neural network loss term and the physical loss term includes determining according to equation (4):
L(m,a,λ)＝(1-λ)L data +λL p (m,a) (4)
wherein L (m, a, lambda) is the total loss function of the prediction model, L data As a neural network loss term, lp (m, a) is a physical loss term, and λ is a weighting factor (0.ltoreq.λ.ltoreq.1).
In one embodiment, the super-parameters include at least one of a size of a history window, a number of hidden layers, a number of neurons per layer, and an activation function.
A second aspect of the application provides a processor configured to perform the above-described prediction method for horizontal well productivity.
A third aspect of the present application provides a productivity prediction system, comprising:
the system comprises a data acquisition module, a control module and a control module, wherein the data acquisition module is used for acquiring production data of a horizontal well in a historical time period and physical constraint parameters, and the physical constraint parameters comprise at least one of reservoir properties, logging curves, horizontal well information, fracturing parameters and well control changes;
the preprocessing module is used for preprocessing the production data and the physical constraint parameters to obtain a preprocessed data set corresponding to the production data and the physical constraint parameters;
the model construction module is used for determining a function equation of the prediction model through a physical driving neural network basic method; determining super parameters of the prediction model by a Bayesian optimization method to determine a model framework of the prediction model;
The prediction module divides the preprocessed data set into a plurality of groups of data, and sequentially inputs each group of data into the prediction model; outputting the predicted capacity of the horizontal well in a predicted time period through a prediction model according to each group of data; and determining the actual predicted capacity of the horizontal well in the predicted time period according to the plurality of predicted capacities.
A fourth aspect of the application provides a machine-readable storage medium having instructions stored thereon, which when executed by a processor, cause the processor to be configured to perform the above-described method of predicting horizontal well capacity.
Through the technical scheme, the accurate prediction of the productivity of the horizontal well is realized.
Additional features and advantages of embodiments of the application will be set forth in the detailed description which follows.
Drawings
The accompanying drawings are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain, without limitation, the embodiments of the application. In the drawings:
FIG. 1 schematically illustrates a flow diagram of a method for predicting horizontal well capacity in accordance with an embodiment of the present application;
FIG. 2 schematically illustrates a flow chart of a normalized pretreatment of raw data of a horizontal well for in-situ fracturing in accordance with an embodiment of the present application;
fig. 3 schematically shows a structural schematic of a fourth folding model according to an embodiment of the application;
FIG. 4 schematically illustrates a schematic diagram of the evolution of a loss function over six cross-validation periods in accordance with an embodiment of the present application;
FIG. 5 schematically illustrates a schematic diagram of predicted results for a different physical constraint parameter in a test set according to an embodiment of the present application;
FIG. 6 schematically illustrates a schematic diagram of the predictive performance of a PC-BiGRU model under different physical constraint parameters according to an embodiment of the application;
FIG. 7 schematically illustrates a block diagram of a capacity prediction system in accordance with an embodiment of the present application;
fig. 8 schematically shows an internal structural view of a computer device according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it should be understood that the detailed description described herein is merely for illustrating and explaining the embodiments of the present application, and is not intended to limit the embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Fig. 1 schematically shows a flow diagram of a method for predicting horizontal well capacity according to an embodiment of the application. As shown in fig. 1, in an embodiment of the present application, there is provided a method for predicting productivity of a horizontal well, including the steps of:
step 101, obtaining production data and physical constraint parameters of a horizontal well in a historical time period, wherein the physical constraint parameters comprise at least one of reservoir properties, logging curves, horizontal well information, fracturing parameters and well control changes.
Step 102, preprocessing the production data and the physical constraint parameters to obtain a preprocessed data set corresponding to the production data and the physical constraint parameters.
And step 103, determining a function equation of the prediction model through a physical driving neural network basic method.
And 104, determining super parameters of the prediction model through a Bayesian optimization method to determine a model framework of the prediction model.
And 105, constructing a prediction model according to the function equation and the model.
And 106, dividing the preprocessed data set into a plurality of groups of data, and sequentially inputting each group of data into the prediction model.
And 107, outputting the predicted capacity of the horizontal well in a predicted time period through a prediction model according to each group of data.
And step 108, determining the actual predicted capacity of the horizontal well in the predicted time period according to the plurality of predicted capacities.
Oil and gas are buried in hydrocarbon reservoirs ranging from tens to thousands of meters underground, and if it is to be produced, it is necessary to create a hydrocarbon path between the surface and the subsurface hydrocarbon reservoir, which is known as a well. A horizontal well is a special well in which the maximum well angle reaches or approaches 90 ° (typically not less than 86 °) and a length of horizontal well section is maintained in the layer of interest. In general, horizontal wells are used in thin hydrocarbon reservoirs or in fractured hydrocarbon reservoirs in order to increase the exposed area of the hydrocarbon reservoir. The horizontal well usually adopts a staged fracturing technology, which is a yield increasing transformation technology, when fracturing fluid enters a reservoir through perforation holes at a high speed, the perforation friction is generated and increases with the increase of pumping displacement, the bottom hole pressure is driven to rise, and when the bottom hole pressure exceeds the fracture pressure of a plurality of fracturing layers, namely, fracture joints are pressed on each layer, the fracturing pressure of each layer is required to be basically close, and the fracturing pressure can be regulated by the perforation friction. In the technical scheme, a horizontal well can be selected for exploitation of compact oil, the compact oil is clamped in or close to a compact reservoir stratum of a high-quality crude oil layer, petroleum aggregation formed without large-scale long-distance migration is a large-area continuous distribution petroleum resource which is symbiotic or close to a crude oil rock system, the lithology of the reservoir stratum mainly comprises compact sandstone, compact limestone and carbonate rock, the permeability of a overburden matrix is less than 0.1mD on average, and a single well has no natural industrial productivity. The dense oil is transported by a short distance, and the main occurrence space is in a dense layer close to the source rock.
Further, raw data of the fractured horizontal well of the petroleum exploitation site is collected, wherein the raw data comprises production data and physical constraint parameters of the horizontal well in a historical time period. The historical time period can be selected according to actual requirements, for example, the time period from the initial production day to the current day of the horizontal well can be selected. Physical constraints refer to the use of physical methods to limit the movement of an object, avoiding its movement beyond certain limits. In the present technical solution, the physical constraint parameter may refer to a parameter that limits the horizontal well and the petroleum to a certain extent by a physical method. For example, the physical constraint parameters may include at least one of reservoir properties, which may refer to permeability, porosity, and oil saturation of the tight oil reservoir, logging curves, horizontal well information, fracturing parameters, which may refer to the number of fracturing segments, total sand volume, total fluid volume, and slip volume, and well control changes, which may refer to the choke size, shut-in time, oil pressure, and production days. Wherein the well control variation is time series data, the time series production predictions can be constrained from the time dimension. Reservoir properties, logging curves, well information, and fracturing parameters reflect the spatial variation of different wells and constrain the production variation from spatial dimensions. After the processor acquires the collected raw data of the field fracturing horizontal well, the production data and the physical constraint parameters are preprocessed to obtain a preprocessed data set corresponding to the production data and the physical constraint parameters. Preprocessing may refer to performing a data normalization process on the production data and the physical constraint parameters, which may eliminate variability of different feature quantities, min-max normalization for all features between 0 and 1.
Further, a function equation of the prediction model is determined through a physical driving neural network basic method, super parameters of the prediction model are determined through a Bayesian optimization method, so that a model framework of the prediction model is determined, and the processor can construct the prediction model according to the function equation and the model. The function equation may be a loss function, where the loss function is a function that maps the random event or its related random variable to a non-negative real number to represent the "risk" or "loss" of the random event, and the prediction model may refer to a neural model. In particular, determining the functional equation of the predictive model by the physical driving neural network basic method may refer to defining the neural network loss function by physical parameter constraints and yield decreasing function physical term constraints. The super-parameters are tuning parameters in the machine learning algorithm, and the super-parameters are parameters of set values before the learning process is started, and the settings need to be considered. In general, the super parameters need to be optimized, and a group of optimal super parameters are selected for the learning machine so as to improve the learning performance and effect. In the technical scheme, the optimal super-parameters are obtained by selecting a Bayesian optimization method so as to determine the optimal model architecture of the prediction model.
Further, after constructing the prediction model by the physical driving neural network basic method and the bass optimization method, the processor divides the preprocessed data set into a plurality of groups of data, and sequentially inputs each group of data into the prediction model. And outputting the predicted capacity of the horizontal well in the predicted time period through a prediction model according to each group of data, and determining the actual predicted capacity of the horizontal well in the predicted time period according to the plurality of predicted capacities. Specifically, the data input into the prediction model can obtain a final prediction result by adopting a cross-validation mode, and the cross-validation can enable the estimated model to be more accurate and reliable. The cross-validation refers to dividing one part of training data input into a model into training sets, the other part of training data is a test set, and in each test process, changing different test sets each time until all test results are obtained, and taking an average value as a final result. In the technical scheme, the preprocessing data set can be divided into six groups of data, and the training data output to the prediction model is tested by adopting a six-time cross validation method. The five-fold dataset was used as the training set for training the model, and the remaining one-fold dataset was used as the test set to evaluate model performance. This process is repeated until all folds are tested once. After all the data are tested, the average value of the six groups of test results can be taken as a final prediction result to be output, and the processor can be used for actually predicting the productivity of the horizontal well in a prediction time period according to the result output by the prediction model.
Further, for example, production data and physical constraint parameters of the first day to the tenth day of the fracturing horizontal well of the petroleum exploitation site can be collected, the data is subjected to data normalization pretreatment to obtain a pretreatment data set, the pretreatment data set is output to a prediction model, and six times of cross validation is adopted to obtain the predicted capacity of the horizontal well of the eleventh day.
According to the technical scheme, accurate prediction of the productivity of the horizontal well is achieved.
FIG. 1 is a flow chart of a method for predicting production capacity of a horizontal well according to one embodiment. It should be understood that, although the steps in the flowchart of fig. 1 are shown in sequence as indicated by the arrows, the steps are not necessarily performed in sequence as indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in fig. 1 may include multiple sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, nor do the order in which the sub-steps or stages are performed necessarily performed in sequence, but may be performed alternately or alternately with at least a portion of other steps or sub-steps of other steps.
In one embodiment, determining the function equation of the predictive model by the physical driving neural network basic method includes: firstly, determining a physical loss term and a neural network loss term of a function equation, and secondly, determining the function equation of the prediction model according to the preset neural network loss term and the physical loss term. Specifically, the physical loss term of the function equation includes determining according to equation (1):
wherein Lp (m, a) is a physical loss term,the predicted production energy is output by the Duong yield decreasing model in the historical time period, y is the actual production energy of the horizontal well in the historical time period, T is the historical time period, and a and m are constants.
In the prediction of the production curve of an actual horizontal well, the differential equation for seepage is difficult to embed in the loss function, because the parameters such as pressure, saturation, etc. are usually unknown. The Duong yield decreasing curve is a wide approximation of unconventional oil and gas reservoir fracturing well production, and can be used for normalizing long-term production prediction results. The Duong yield decremental model is determined according to equation (2):
the method comprises the steps of determining a real-time capacity of a horizontal well in a historical time period, determining an accumulated capacity of the horizontal well in the historical time period, wherein n is a crack time index, T is a preset time period, and determining an initial capacity of the horizontal well in the historical time period.
Further, based on the fracture linear flow assumption:
where a and m are constants, the above formula is an empirical formula, and a=0.5 and m=1.23 can be generally taken.
Thus, the Duong yield decreasing formula can be obtained:
yield reduction ratio D (T) may be defined:
substituting formula (2-2) into (2-3) can result in:
and (3) making:
its derivative is:
substitution of the formulas (2-6) and (2-5) into the formula (2-4) can give:
further derivation of formulas (2-7) may result in:
the formula (2-3) can be modified as follows:
substituting the formula (2-8) into the formula (2-9) can obtain:
from the above calculation formula, a physical loss term, i.e., formula (1), can be defined.
In one embodiment, the Duong yield decremental formula is a broad approximation of the long term production of a fractured well, and the purpose of adding this formula to the loss function is to constrain the long term yield prediction trend. The initial production of a particular fractured well does not meet the yield decline and therefore does not directly minimize the physical loss term, but rather embeds it as a soft constraint in the total loss function. Specifically, the neural network loss term of the function equation includes determining according to equation (3):
wherein L is data For the neural network to lose terms,and (3) outputting predicted production energy for the horizontal well in a historical time period through a Duong yield decreasing model, wherein y is the actual production energy of the horizontal well in the historical time period.
Further, deriving the total loss function of the neural model based on the physical loss term and the neural network loss term includes determining according to equation (4):
L(m,a,λ)＝(1-λ)L data +λL p (m,a) (4)
wherein L (m, a, lambda) is the total loss function of the prediction model, L data As a neural network loss term, lp (m, a) is a physical loss term, and λ is a weighting factor (0.ltoreq.λ.ltoreq.1).
In one embodiment, the processor obtains optimal super parameters by selecting a Bayesian optimization method after determining the function equation of the prediction model by a physical driving neural network basic method to determine the optimal model architecture of the prediction model. The bayesian optimization method is a super-parameter optimization method commonly used in the field of machine learning at present, and can be regarded as the most advanced optimization framework at present. In this technical solution, the hyper-parameters of the prediction model may refer to the size of the history window, the number of hidden layers, the number of neurons per layer, and the activation function.
In a specific embodiment, a method for constructing a novel physical constraint deep learning model based on a bi-directional gating recursion unit and a deep hybrid neural network and predicting the capacity of a tight oil multistage fracturing horizontal well is provided. Specifically, by modifying the Python source program, in-situ fracturing horizontal well raw data is collected, including production data and physical constraint parameters, including reservoir properties, logging curves, well information, fracturing parameters, and well control variations, and a data normalization preprocessing process is performed to eliminate variability of different feature quantities.
As shown in fig. 2, a schematic diagram of a normalized pretreatment flow of raw data of a field fracturing horizontal well is provided. Specifically, first, data normalization is performed to eliminate variability of different feature amounts, and min-max normalization is used to bring all features between 0 and 1, and the following formula (5) is defined
Wherein Inp is T Andand represent the raw and normalized input features, inp, at time step T, respectively min And Inp max Representing the minimum and maximum values of the input sequence, respectively.
To meet the input requirements of the time series neural network model, the production data should be converted into input-output pairs through a sliding history window. The specific format conversion process is shown in fig. 2, assuming a production sequence with (n+3) data points. From day 1, the production data from the first n days is considered as input of the first sample, and the production information from day (n+1) is considered as output of the 1 st sample. Next, the sliding window is moved forward for one day. Production data from day 2 through (n+1) are used as input for the second sample, capacity data from day (n+2) are used as output for the first sample, and this process is continued until all data is covered. In this way, a production sequence with (n+3) data points is converted into three input-output samples.
Further, as shown in fig. 3, a structural schematic diagram of a fourth folding model is provided. Specifically, the time series of oil rate inputs are converted to three-dimensional tensors (5052, 8, 1) and input into a biglu layer having 81 nodes to obtain the time correlation between production sequences. Depending on the data dimension, physical constraints are passed to the full link layer and the convolutional layer, respectively, to extract useful representations. Where two-dimensional constraints (including well information, fracturing parameters, and operational variations) are transmitted through two dense layers and 72 nodes, while three-dimensional constraints (such as logging sequences) are transmitted to a convolutional layer. A dense layer is added after the time-series network layer by layer. Except that the activation function of the output layer is linear, the activation functions of other layers are optimized to be Sigmoid. In this way, the temporal and spatial physical parameters are integrated into a long-term predictive network architecture, which helps to output more accurate and physically consistent predictions.
Further, a Bayesian optimization method is adopted to optimize the super parameters of the PC-BiGRU combined neural network model. Due to the six cross-validations, there are six models in total waiting for optimization. The optimized hyper-parameters include the length of the history window, the number of hidden layers, and the number of neurons and activation functions in these layers. We approximated the search range of the layer by trial and error, with a history window of 14 and a neuron count of 64, based on the prediction accuracy and computational cost.
As shown in fig. 4, a schematic diagram of the evolution of the loss function over six cross-validation periods is provided. In particular, when the number of layers in all the modules is greater than 4, the calculation time increases significantly, and the accuracy decreases drastically. Therefore, the search range of the number of layers is set to 1 to 4. Bayesian optimization is then used to find the optimal hyper-parameters for the six models.
Further, the verification set is adopted to evaluate the accuracy of the pre-configured tight oil fracturing horizontal well productivity prediction network model obtained through training, and the quality of the pre-configured oil reservoir network model based on root mean square error is further checked. The root mean square error equation is defined as the following equation (6).
Wherein q is T Represents the capacity, q, of dense oil using the original collection T Represents the tight oil production predicted at the same time using the preconfigured reservoir network model, H represents the number of time steps.
Further, as shown in fig. 5, a schematic diagram of the prediction results of different physical constraint parameters in the test set is provided. And after the PC-BiGRU combined neural network model with the optimal super parameters is obtained, network model parameter optimization is carried out on the training set. The optimization parameters may be set as follows: adam was used as an optimizer, and the learning rate was set to 0.001. All weight matrices are initialized by Xavier initialization, and the bias matrix is initialized to zero. All models were run for 50 cycles with a batch size of 50. FIG. 4 depicts the evolution of the loss function with respect to period, indicating that all cross-validated PC-BiGRU models reached convergence after 50 periods without over-fitting problems. These models were coded with Python 3.7 under deep learning Keras libraries and were then run in the presence of a model Run on a 3.40GHz CPU computer. The influence of different physical constraints on the model performance will be discussed, which helps to determine the contribution of different physical constraints to the improvement of the productivity prediction accuracy. Well information and fracturing parameters are considered Engineering Constraints (ECs) because completion and fracturing are relevant to engineering construction. Log curves are classified as Geologic Constraints (GC) because they can indirectly reflect reservoir properties and rock mechanics. The operational changes are represented by OC and ALL representation takes into account ALL constraints. Different types of physical constraints are integrated into the PC-BiGRU model, respectively.
Further, as shown in fig. 6, a schematic diagram of the predictive performance of the PC-biglu model under different physical constraint parameters is provided. As can be seen from fig. 6, OC contributes most to model performance, followed by EC and GC. This may be due to the fact that daily operational changes contain rich dynamic information related to oil rate fluctuations. In addition, completion and fracturing parameters may reflect stimulation scale, meaning stimulation amplitude, while geologic features characterize reservoir heterogeneity and inter-well variability. It can be seen that in the multi-step lead capacity forecast task, GC and EC are responsible for constraining the overall trend of long-term capacity, OC is responsible for focusing on subtle fluctuations in time-series capacity.
According to the technical scheme, the production capacity prediction of the tight oil multistage fracturing horizontal well is based on the physical constraint bidirectional circulating neural network. The method comprises the steps of collecting original data of a field fracturing horizontal well, performing data normalization preprocessing, performing cross validation on a preprocessed data set, weakening the influence of random partitioning of the data set, and objectively evaluating the prediction effect of the neural network. Secondly, an optimal super-parameter is obtained by adopting a Bayesian optimization method, an optimal model framework is determined, a neural network loss function is defined through physical parameter constraint and yield decreasing function physical item constraint, deep learning model training is carried out, a deep learning prediction result is evaluated on a test set, and the technical problem of low precision caused by driving fracturing horizontal well productivity prediction by means of pure data is solved.
In one embodiment, as shown in FIG. 7, there is provided a capacity prediction system 700 comprising:
a data acquisition module 710 for acquiring production data of the horizontal well over a historical period of time and physical constraint parameters including at least one of reservoir properties, logging curves, horizontal well information, fracturing parameters, and well control changes;
a preprocessing module 720, configured to preprocess the production data and the physical constraint parameters to obtain a preprocessed data set corresponding to the production data and the physical constraint parameters;
A model construction module 730 for determining a function equation of the prediction model by a physical driving neural network basic method; determining super parameters of the prediction model by a Bayesian optimization method to determine a model framework of the prediction model;
the prediction module 740 divides the preprocessed data set into a plurality of groups of data, and sequentially inputs each group of data into the prediction model; outputting the predicted capacity of the horizontal well in a predicted time period through a prediction model according to each group of data; and determining the actual predicted capacity of the horizontal well in the predicted time period according to the plurality of predicted capacities.
The embodiment of the application provides a storage medium, wherein a program is stored on the storage medium, and the program is executed by a processor to realize the prediction method for the horizontal well productivity.
The embodiment of the application provides a processor for running a program, wherein the program runs to execute the method for predicting the productivity of a horizontal well.
In one embodiment, a computer device is provided, which may be a server, and the internal structure of which may be as shown in fig. 8. The computer device includes a processor a01, a network interface a02, a memory (not shown) and a database (not shown) connected by a system bus. Wherein the processor a01 of the computer device is adapted to provide computing and control capabilities. The memory of the computer device includes internal memory a03 and nonvolatile storage medium a04. The nonvolatile storage medium a04 stores an operating system B01, a computer program B02, and a database (not shown in the figure). The internal memory a03 provides an environment for the operation of the operating system B01 and the computer program B02 in the nonvolatile storage medium a04. The database of the computer device is used for storing prediction data for the capacity of the horizontal well. The network interface a02 of the computer device is used for communication with an external terminal through a network connection. The computer program B02, when executed by the processor a01, implements a predictive method for horizontal well productivity.
It will be appreciated by those skilled in the art that the structure shown in FIG. 8 is merely a block diagram of some of the structures associated with the present inventive arrangements and is not limiting of the computer device to which the present inventive arrangements may be applied, and that a particular computer device may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.
The embodiment of the application provides equipment, which comprises a processor, a memory and a program stored in the memory and capable of running on the processor, wherein the processor realizes the following steps when executing the program: acquiring production data and physical constraint parameters of the horizontal well in a historical time period, wherein the physical constraint parameters comprise at least one of reservoir properties, logging curves, horizontal well information, fracturing parameters and well control changes; preprocessing production data and physical constraint parameters to obtain a preprocessed data set corresponding to the production data and the physical constraint parameters; determining a function equation of a prediction model by a physical driving neural network basic method; determining super parameters of the prediction model by a Bayesian optimization method to determine a model framework of the prediction model; constructing a prediction model according to the function equation and the model; dividing the preprocessing data set into a plurality of groups of data, and sequentially inputting each group of data into a prediction model; outputting the predicted capacity of the horizontal well in a predicted time period through a prediction model according to each group of data; and determining the actual predicted capacity of the horizontal well in the predicted time period according to the plurality of predicted capacities.
In one embodiment, determining the function equation of the predictive model by the physical driving neural network basic method includes: determining a physical loss term and a neural network loss term of a function equation; and determining a function equation of the prediction model according to the preset neural network loss term and the physical loss term.
In one embodiment, determining the physical loss term of the function equation includes determining according to equation (1):
wherein Lp (m, a) is a physical loss term,the predicted production energy is output by the Duong yield decreasing model in the historical time period, y is the actual production energy of the horizontal well in the historical time period, T is the historical time period, and a and m are constants.
In one embodiment, the method further comprises determining a Duong yield decremental model according to equation (2):
wherein q (T) is the real-time capacity of the horizontal well in the historical time period, N p For the cumulative capacity of the horizontal well over the historical time period, n is the fracture time index, T isPreset time period, q i Is the initial capacity of the horizontal well over a historical period of time.
In one embodiment, determining the neural network loss term of the functional equation includes determining according to equation (3):
wherein L is data For the neural network to lose terms,and (3) outputting predicted production energy for the horizontal well in a historical time period through a Duong yield decreasing model, wherein y is the actual production energy of the horizontal well in the historical time period.
In one embodiment, determining the function equation of the predictive model from the predetermined neural network loss term and the physical loss term includes determining according to equation (4):
L(m,a,λ)＝(1-λ)L data +λL p (m,a) (4)
wherein L (m, a, lambda) is the total loss function of the prediction model, L data As a neural network loss term, lp (m, a) is a physical loss term, and λ is a weighting factor (0.ltoreq.λ.ltoreq.1).
In one embodiment, the super-parameters include at least one of a size of a history window, a number of hidden layers, a number of neurons per layer, and an activation function.
The application also provides a computer program product adapted to perform, when executed on a data processing device, a program initialized with the method steps of: acquiring production data and physical constraint parameters of the horizontal well in a historical time period, wherein the physical constraint parameters comprise at least one of reservoir properties, logging curves, horizontal well information, fracturing parameters and well control changes; preprocessing production data and physical constraint parameters to obtain a preprocessed data set corresponding to the production data and the physical constraint parameters; determining a function equation of a prediction model by a physical driving neural network basic method; determining super parameters of the prediction model by a Bayesian optimization method to determine a model framework of the prediction model; constructing a prediction model according to the function equation and the model; dividing the preprocessing data set into a plurality of groups of data, and sequentially inputting each group of data into a prediction model; outputting the predicted capacity of the horizontal well in a predicted time period through a prediction model according to each group of data; and determining the actual predicted capacity of the horizontal well in the predicted time period according to the plurality of predicted capacities.
In one embodiment, determining the function equation of the predictive model by the physical driving neural network basic method includes: determining a physical loss term and a neural network loss term of a function equation; and determining a function equation of the prediction model according to the preset neural network loss term and the physical loss term.
In one embodiment, determining the physical loss term of the function equation includes determining according to equation (1):
/>
wherein Lp (m, a) is a physical loss term,the predicted production energy is output by the Duong yield decreasing model in the historical time period, y is the actual production energy of the horizontal well in the historical time period, T is the historical time period, and a and m are constants.
In one embodiment, the method further comprises determining a Duong yield decremental model according to equation (2):
wherein q (T) is the real-time capacity of the horizontal well in the historical time period, N p For the accumulated capacity of the horizontal well in the historical time period, n is a crack time index, T is a preset time period, q i Is the initial capacity of the horizontal well over a historical period of time.
In one embodiment, determining the neural network loss term of the functional equation includes determining according to equation (3):
wherein L is data For the neural network to lose terms,and (3) outputting predicted production energy for the horizontal well in a historical time period through a Duong yield decreasing model, wherein y is the actual production energy of the horizontal well in the historical time period.
In one embodiment, determining the function equation of the predictive model from the predetermined neural network loss term and the physical loss term includes determining according to equation (4):
L(m,a,λ)＝(1-λ)L data +λL p (m,a) (4)
wherein L (m, a, lambda) is the total loss function of the prediction model, L data As a neural network loss term, lp (m, a) is a physical loss term, and λ is a weighting factor (0.ltoreq.λ.ltoreq.1).
In one embodiment, the super-parameters include at least one of a size of a history window, a number of hidden layers, a number of neurons per layer, and an activation function.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.
Claims (10)
1. A prediction method for horizontal well productivity, the prediction method comprising:
acquiring production data and physical constraint parameters of the horizontal well in a historical time period, wherein the physical constraint parameters comprise at least one of reservoir properties, logging curves, horizontal well information, fracturing parameters and well control changes;
preprocessing the production data and the physical constraint parameters to obtain a preprocessed data set corresponding to the production data and the physical constraint parameters;
determining a function equation of a prediction model by a physical driving neural network basic method;
determining super parameters of the prediction model through a Bayesian optimization method to determine a model framework of the prediction model;
constructing the prediction model according to the function equation and the model;
dividing the preprocessing data set into a plurality of groups of data, and sequentially inputting each group of data into the prediction model;
Outputting the predicted capacity of the horizontal well in a predicted time period through the prediction model according to each group of data;
and determining the actual predicted capacity of the horizontal well in the predicted time period according to the predicted capacities.
2. The method for predicting capacity of a horizontal well as set forth in claim 1, wherein said determining a function equation of the prediction model by a physical driving neural network basic method comprises:
determining a physical loss term and a neural network loss term of the function equation;
and determining a function equation of the prediction model according to a preset neural network loss term and the physical loss term.
3. The method of predicting horizontal well productivity of claim 2, wherein determining the physical loss term of the functional equation comprises determining according to equation (1):
wherein Lp (m, a) is the physical loss term,predicted production output by the Duong yield decrementing model for the horizontal well over the historical period of time,y is the actual capacity of the horizontal well in the historical time period, T is the historical time period, and a and m are constants.
4. A method for predicting capacity of a horizontal well as set forth in claim 3, further comprising determining the Duong yield decremental model according to equation (2):
Wherein q (T) is the real-time capacity of the horizontal well in the historical time period, N p For the accumulated capacity of the horizontal well in the historical time period, n is a crack time index, T is the preset time period, q i And (3) initial capacity of the horizontal well in the historical time period is obtained.
5. The method of claim 2, wherein determining the neural network loss term for the functional equation comprises determining according to equation (3):
wherein L is data For the neural network loss term(s),and outputting predicted production energy for the horizontal well in the historical time period through a Duong production decreasing model, wherein y is the actual production energy of the horizontal well in the historical time period.
6. The prediction method for horizontal well productivity according to claim 2, wherein the function equation for determining the prediction model from a predetermined neural network loss term and the physical loss term comprises determining according to formula (4):
L(m,a,λ)＝(1-λ)L data +λL p (m,a)(4)
wherein L (m, a, lambda) is the total loss function of the predictive model, L data And Lp (m, a) is the physical loss term, and lambda is a weighting factor (0 is less than or equal to lambda is less than or equal to 1) for the neural network loss term.
7. The method of claim 1, wherein the super parameters include at least one of a size of a history window, a number of hidden layers, a number of neurons per layer, and an activation function.
8. A processor configured to perform the prediction method for horizontal well capacity according to any one of claims 1 to 7.
9. A capacity prediction system, comprising:
the system comprises a data acquisition module, a control module and a control module, wherein the data acquisition module is used for acquiring production data of a horizontal well in a historical time period and physical constraint parameters, and the physical constraint parameters comprise at least one of reservoir properties, logging curves, horizontal well information, fracturing parameters and well control changes;
the preprocessing module is used for preprocessing the production data and the physical constraint parameters to obtain a preprocessed data set corresponding to the production data and the physical constraint parameters;
the model construction module is used for determining a function equation of the prediction model through a physical driving neural network basic method; determining super parameters of the prediction model through a Bayesian optimization method to determine a model framework of the prediction model;
the prediction module divides the preprocessing data set into a plurality of groups of data, and sequentially inputs each group of data into the prediction model; outputting the predicted capacity of the horizontal well in a predicted time period through the prediction model according to each group of data; and determining the actual predicted capacity of the horizontal well in the predicted time period according to the predicted capacities.
10. A machine-readable storage medium having instructions stored thereon, which when executed by a processor, cause the processor to be configured to perform the prediction method for horizontal well capacity according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310671981.1A CN116911432A (en) | 2023-06-07 | 2023-06-07 | Prediction method, storage medium and processor for horizontal well productivity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310671981.1A CN116911432A (en) | 2023-06-07 | 2023-06-07 | Prediction method, storage medium and processor for horizontal well productivity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116911432A true CN116911432A (en) | 2023-10-20 |
Family
ID=88350070
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310671981.1A Pending CN116911432A (en) | 2023-06-07 | 2023-06-07 | Prediction method, storage medium and processor for horizontal well productivity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116911432A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117522173A (en) * | 2024-01-04 | 2024-02-06 | 山东科技大学 | Natural gas hydrate depressurization exploitation productivity prediction method based on deep neural network |
-
2023
- 2023-06-07 CN CN202310671981.1A patent/CN116911432A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117522173A (en) * | 2024-01-04 | 2024-02-06 | 山东科技大学 | Natural gas hydrate depressurization exploitation productivity prediction method based on deep neural network |
| CN117522173B (en) * | 2024-01-04 | 2024-04-26 | 山东科技大学 | Natural gas hydrate depressurization exploitation productivity prediction method based on deep neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hongliang et al. | Production prediction at ultra-high water cut stage via Recurrent Neural Network | |
| Wei et al. | Predicting injection profiles using ANFIS | |
| CN109709607B (en) | Method and device for predicting thickness of thin sandstone reservoir | |
| Zhang et al. | Application of fuzzy comprehensive evaluation to evaluate the effect of water flooding development | |
| Mohd Razak et al. | Transfer learning with recurrent neural networks for long-term production forecasting in unconventional reservoirs | |
| Mishra et al. | Evaluation and development of a predictive model for geophysical well log data analysis and reservoir characterization: machine learning applications to lithology prediction | |
| CN116911432A (en) | Prediction method, storage medium and processor for horizontal well productivity | |
| Sudakov et al. | Artificial neural network surrogate modeling of oil reservoir: A case study | |
| Tariq et al. | An intelligent data-driven model for Dean–Stark water saturation prediction in carbonate rocks | |
| Sankaran et al. | Data Analytics in Reservoir Engineering | |
| Hou et al. | Data-driven optimization of brittleness index for hydraulic fracturing | |
| Tan et al. | Fracturing productivity prediction model and optimization of the operation parameters of shale gas well based on machine learning | |
| WO2022011015A1 (en) | Optimization of discrete fracture network (dfn) using streamlines and machine learning | |
| Garavand et al. | In situ stress assessment based on plastic behavior of borehole breakouts and machine learning | |
| Yousefzadeh et al. | History matching and robust optimization using proxies | |
| Mathew et al. | Application of Machine Learning to Interpret Steady State Drainage Relative Permeability Experiments | |
| George | Predicting Oil Production Flow Rate Using Artificial Neural Networks-The Volve Field Case | |
| CN112749835B (en) | Reservoir productivity prediction method, device and equipment | |
| Asante et al. | Forecasting Oil Recovery Using Long Short Term Memory Neural Machine Learning Technique | |
| Wang et al. | Capturing intrinsic features from field data for predicting the production of natural gas | |
| Temizel et al. | Effective use of data-driven methods in brown fields | |
| Qin et al. | Predicting heavy oil production by hybrid data-driven intelligent models | |
| Yang et al. | Hybrid Deep Learning Approach for Hydraulic Fracture Morphology Diagnosis in Coalbed Methane Wells | |
| Verma et al. | Quantifying sand fraction from seismic attributes using modular artificial neural network | |
| Du et al. | A similarity measure recognized by morphological characteristics analysis of well logging curves: application to the knowledge domain of sandstone reservoir |
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 |