Method for predicting gas production rate of staged fractured horizontal well and related equipment

Abstract

The invention discloses a staged fractured horizontal well gas production rate prediction method and related equipment, and the method comprises the steps: inputting each main control factor corresponding to each parameter of a yield decline model into a trained artificial neural network model corresponding to each parameter, obtaining parameters of a yield decline model output by each trained artificial neural network model; and on the basis of the parameters of the yield decline model, the gas production predicted quantity of the staged fractured horizontal well is obtained through the yield decline model. Through the method and the device, the technical problem that the gas production rate of the staged fractured horizontal well cannot be predicted due to the fact that the parameters of the yield decline model cannot be determined in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of natural gas development, and in particular to a segmented fracturing horizontal well gas production prediction method and related equipment. BACKGROUND

[0002] At present, segmented fracturing horizontal wells are widely used in unconventional fields such as shale oil and gas, tight oil and gas, etc. The segmented fracturing horizontal well gas production decline law is complex, and the parameters of the production decline model can be obtained only based on the full life cycle dynamic data, and then the segmented fracturing horizontal well gas production can be predicted based on the parameters of the production decline model.

[0003] However, due to the long natural production cycle of unconventional oil and gas reservoirs, the dynamic data recorded is relatively short, which cannot reflect the production characteristics of the gas well throughout its life cycle, and thus the parameters of the production decline model cannot be determined.

[0004] Therefore, there is an urgent need for a technical solution that can determine the parameters of the production decline model and predict the segmented fracturing horizontal well gas production in the absence of dynamic data throughout the life cycle of the segmented fracturing horizontal well. SUMMARY

[0005] The present application provides a segmented fracturing horizontal well gas production prediction method and related equipment, which can solve the technical problem that the parameters of the production decline model cannot be determined in the prior art, and the segmented fracturing horizontal well gas production cannot be predicted.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a segmented fracturing horizontal well gas production prediction method, which comprises:

[0008] Each main control factor corresponding to each parameter of the production decline model is input into the trained artificial neural network model corresponding to each parameter to obtain the parameters of the production decline model output by each trained artificial neural network model.

[0009] Based on the parameters of the production decline model, the segmented fracturing horizontal well gas production prediction is obtained through the production decline model.

[0010] In a second aspect, the present application provides a segmented fracturing horizontal well gas production prediction device, which comprises:

[0011] The artificial neural network model module is configured to input each master factor corresponding to each parameter of the production decline model into a trained artificial neural network model corresponding to the parameter, and obtain the parameter of the production decline model output by each trained artificial neural network model.

[0012] The production decline model module is configured to obtain the gas production prediction of the staged fracturing horizontal well by the production decline model based on the parameter of the production decline model.

[0013] In a third aspect, an electronic device is provided, and the electronic device includes a memory and a processor. The processor is configured to read and execute a computer program stored in the memory, so as to implement the method for visual authorization of starting a train.

[0014] In a fourth aspect, a computer storage medium is provided, and the computer storage medium stores computer executable instructions. The computer executable instructions are executed to implement the method for visual authorization of starting a train.

[0015] The technical scheme provided by the embodiments of the present application has the following beneficial effects:

[0016] The parameter of the production decline model output by each trained artificial neural network model is obtained through the trained artificial neural network model corresponding to each parameter. The production decline model is parameterized based on the obtained parameter of the production decline model, and the gas production prediction of the staged fracturing horizontal well is obtained through the parameterized production decline model. The technical problem that the gas production of the staged fracturing horizontal well cannot be predicted due to the inability to determine the parameter of the production decline model in the related art is solved. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical scheme in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0018] Figure 1 The flowchart of the embodiment of the method for predicting the gas production of the staged fracturing horizontal well of the present application;

[0019] Figure 2 The flowchart of another embodiment of the method for predicting the gas production of the staged fracturing horizontal well of the present application;

[0020] Figure 3a The schematic diagram of the influence degree of each influencing factor on the initial production of the production decline model of the present application;

[0021] Figure 3b Fig. 1 is a diagram illustrating the influence of each influencing factor on the initial decline rate of the production decline model of the present application;

[0022] Figure 3c Fig. 2 is a diagram illustrating the influence of each influencing factor on the decline index of the production decline model of the present application;

[0023] Figure 4 Fig. 3 is a diagram illustrating the basic model of the segmented fracturing horizontal well of the present application;

[0024] Figure 5 Fig. 4 is a diagram illustrating the functional modules of an embodiment of the segmented fracturing horizontal well gas production prediction device of the present application;

[0025] Figure 6 Fig. 5 is a diagram illustrating the structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0027] To make the objects, technical solutions, and advantages of the present application clearer, the embodiments of the present application will be described in detail below with reference to the drawings.

[0028] In a first aspect, the embodiments of the present application provide a segmented fracturing horizontal well gas production prediction method.

[0029] In an embodiment, with reference to Figure 1 , Figure 1 Fig. 1 is a flowchart of an embodiment of the segmented fracturing horizontal well gas production prediction method of the present application. As shown in Figure 1 the segmented fracturing horizontal well gas production prediction method includes:

[0030] Step S10, inputting each main control factor corresponding to each parameter of the production decline model into each trained artificial neural network model corresponding to the parameter, to obtain the parameter of the production decline model output by each trained artificial neural network model;

[0031] In the present embodiment, the parameters of the production decline model include the initial production q i , the production decline rate D i , and the decline index b. The corresponding main control factors of the initial production q i , the corresponding main control factors of the production decline rate D i , and the corresponding main control factors of the decline index b are obtained.

[0032] The initial output q i The corresponding initial output q of each main control factor input i The corresponding trained artificial neural network model obtains the initial output q i The initial output q of the production decline model output by the corresponding trained artificial neural network model i .

[0033] The output decline rate D i The corresponding output decline rate D of each main control factor input i The corresponding trained artificial neural network model obtains the production decline rate D i The corresponding training completed artificial neural network model outputs the production decline model's production decline rate D i .

[0034] The respective main control factors corresponding to the decreasing index b are input into the trained artificial neural network model corresponding to the decreasing index b, and the decreasing index b of the output decreasing model outputted by the trained artificial neural network model corresponding to the decreasing index b is obtained.

[0035] Step S20: Based on the parameters of the production decline model, the predicted gas production of the staged fractured horizontal well is obtained through the production decline model.

[0036] In this embodiment, the production decline model is the Arps production decline model, and the parameters of the production decline model (initial production q i , output decline rate D i and the decline index b). Parameters based on the production decline model (initial production q i , output decline rate D i And the decline index b) adjust the parameters of the Arps production decline model, and use the adjusted Arps production decline model to predict the gas production of the staged fractured horizontal well to obtain the predicted gas production of the staged fractured horizontal well. Among them, the Arps decline model is the most widely used empirical model, using the initial production q i , output decline rate D i The model is determined by three parameters: α and the decreasing exponent b.

[0037] The Arps production decline model formula is:

[0038]

[0039] The corresponding cumulative gas production formula is:

[0040]

[0041] Where Q is the cumulative gas production (m 3 ), q is the gas production (m 3 / d),q i is the initial output (m 3 / d), D i is the production decline rate (d -1 ), b is the decreasing exponent (dimensionless).

[0042] In this embodiment, the parameters of the production decline model output by each trained artificial neural network model are obtained using the trained artificial neural network model corresponding to each parameter. The production decline model is then adjusted based on the obtained parameters. The adjusted production decline model can then be used to obtain a predicted gas production for a staged fractured horizontal well. This resolves the technical issue in related arts where the inability to determine the parameters of the production decline model makes it impossible to predict the gas production of a staged fractured horizontal well.

[0043] Optionally, in one embodiment, referring to Figure 2 , Figure 2 FIG. 1 is a flow chart of another embodiment of the method for predicting gas production of a staged fractured horizontal well according to the present invention. Figure 2 As shown, the method for predicting gas production of staged fractured horizontal wells also includes:

[0044] Step S001, obtaining the value range corresponding to each influencing factor, wherein the influencing factor is the influencing factor of the gas production of the staged fracturing horizontal well;

[0045] In this embodiment, the influencing factors, i.e., the factors affecting the gas production of staged fracturing horizontal wells, include the following geological engineering parameters: reservoir thickness h, porosity Permeability K m , gas saturation S gi , formation pressure P i , crack length X f , dimensionless conductivity of fracture C fD , crack height penetration rate h f , the angle between the fracture and the horizontal wellbore θ, the number of fractures n f , horizontal well section length L and bottom hole pressure P wf Obtain the value range corresponding to each impact factor, that is, the upper and lower limits corresponding to each impact factor.

[0046] Step S002, obtaining the gas production of the staged fractured horizontal wells under N groups of simulation schemes based on the upper and lower limits of the values ​​of the influencing factors;

[0047] In some specific embodiments, step S002 specifically includes:

[0048] Based on the upper limit value of each influence factor and the lower limit value of each influence factor, N groups of influence factor combinations are obtained through a two-level fractional factorial design.

[0049] N groups of segmented fracturing horizontal well gas production corresponding to the N groups of influence factor combinations are obtained through the CMG-IMEX software.

[0050] In this embodiment, the values of each influence factor are shown in Table 1:

[0051] Table 1 Values of each influence factor

[0052] Influencing factor Unit Lower value limit Base value Upper value limit h m 5 15 20 φ ％ 6 8 15 K m ]]> mD 0.0001 0.0005 0.002 [SA gi ]] ％ 65 80 100 P i ]]> MPa 20 50 60 x f ]]> m 150 200 400 [C fD ]]> Dimensionless 10 100 300 h f ]]> Dimensionless 0.1 0.8 1.0 θ ° 45 75 90 n f ]]> Bar 20 30 60 L m 1000 1500 2000 P wf ]]> MPa 5 10 20

[0053] Based on the upper limit value of each influence factor and the lower limit value of each influence factor, N groups of influence factor combinations are obtained through a two-level fractional factorial design, wherein the factorial design is a design test for studying the possible effects of multiple factors on responses. The fractional factorial design is a design in which the experimenter only performs a selected subset or partial test run of the full factorial design.

[0054] The CMG-IMEX software is a full-featured three-phase, four-component implicit-explicit black oil simulator for simulating the primary recovery and secondary recovery processes of conventional oil and gas reservoirs. Each group of influence factor combinations is a simulation scheme, and the segmented fracturing horizontal well gas production under N simulation schemes can be obtained through the CMG-IMEX software.

[0055] In step S003, the N groups of gas production are fitted through a production decline model to determine the parameters of the production decline model corresponding to each group of gas production.

[0056] In this embodiment, taking N as 128 as an example, the segmented fracturing horizontal well gas production under 128 simulation schemes is fitted through the Arps production decline model to obtain the parameters (initial production q i , production decline rate D i and decline exponent b) of the production decline model corresponding to the 128 groups of gas production. Part of the simulation schemes (i.e., the combination of values of each influence factor) and the parameters of the Arps production decline model are shown in Table 2:

[0057] Table 2 Parameters of part of the simulation schemes and the Arps production decline model

[0058]

[0059] Wherein, “-1” is used to represent that the lower limit value of the basic model parameter of the segmented fracturing horizontal well is taken, and “1” is used to represent that the upper limit value of the basic model parameter of the segmented fracturing horizontal well is taken.

[0060] Step S004, based on the influence degree of each influence factor on each parameter, obtaining the corresponding master control factor of each parameter;

[0061] In some specific embodiments, step S004 specifically comprises:

[0062] Selecting any parameter as a target parameter, and substituting each influence factor into a preset formula to calculate the influence degree of each influence factor on the target parameter, the preset formula being as follows:

[0063]

[0064] wherein, Effect X is used to represent the influence degree of the influence factor X on the target parameter OF, N is used to represent the number of times of fitting the gas production through the production decline model, OF min is used to represent the value of the target parameter OF corresponding to the lower limit value of the influence factor X, OF max is used to represent the value of the target parameter OF corresponding to the upper limit value of the influence factor X;

[0065] Selecting the first m influence factors with the largest influence degree on the target parameter as the master control factors corresponding to the target parameter;

[0066] Selecting any parameter from the parameters that have not been selected as a target parameter, and returning to execute the step of substituting each influence factor into the preset formula to calculate the influence degree of each influence factor on the target parameter, until the master control factors corresponding to each parameter are obtained.

[0067] In this embodiment, any parameter is selected from the parameters (initial production q i , production decline rate D i and decline index b) of the production decline model as a target parameter, and the target parameter is the initial production q i , and each influence factor includes: reservoir thickness h, porosity permeability K m , gas saturation S gi , formation pressure P i , fracture length x f , fracture dimensionless conductivity C fD , fracture height penetration rate h f , fracture and horizontal wellbore angle θ, fracture number n f , horizontal well section length L and bottom hole pressure P wf For example, substituting each influence factor into the preset formula , the influence degree of each influence factor on the target parameter can be calculated.

[0068] For example, Figure 3a、 Figure 3b and Figure 3c As shown in Table 3, each impact factor has different influence on each parameter of the production decline model. The first m impact factors with the greatest influence on the target parameter are selected as the initial production q i The corresponding master control factor. It should be noted that when the first m impact factors with the greatest influence on the target parameter are selected, the number of the first m impact factors with the greatest influence on the target parameter can be the same or different.

[0069] Any parameter is selected from the parameters (production decline rate D i and decline index b) that have not been selected as the target parameter, and the step of substituting each impact factor into the preset formula to calculate the influence of each impact factor on the target parameter is returned to execute, and the cycle is repeated, so that the production decline rate D i The corresponding master control factor and the corresponding master control factor of the decline index b.

[0070] The master control factor corresponding to each of the parameters is shown in Table 3:

[0071] Table 3 Master control factor corresponding to each parameter of the production decline model

[0072]

[0073] Wherein, "√" is used to represent the selected master control factor, and "×" is used to represent the unselected master control factor.

[0074] In step S005, the artificial neural network model is trained based on the master control factor corresponding to each of the parameters and each of the parameters, and a trained artificial neural network model corresponding to each of the parameters is obtained.

[0075] In some specific embodiments, step S005 specifically includes:

[0076] For any one of the parameters of the production decline model, based on the upper limit value and the lower limit value of the master control factor corresponding to the parameter, the gas production of the horizontal well with staged fracturing under W groups of simulation schemes is obtained.

[0077] The W groups of the gas production are fitted by the production decline model respectively, and the parameter of the production decline model corresponding to each group of the gas production is determined.

[0078] The master control factors corresponding to W' parameters are selected from W parameters as training samples, and the W' parameters are used as labels to train the artificial neural network model, and a trained artificial neural network model corresponding to the parameter is obtained.

[0079] Similarly, the trained artificial neural network model corresponding to each of the parameters is obtained.

[0080] In this embodiment, for any one of the parameters of the production decline model, any one parameter is the initial production q i For example, based on the initial production q i The upper and lower limit values of the corresponding main control factor (h, K m , P i , x f , C fD , h f , n f , P wf ) are used for two-level partial factor design, and W groups of factor combinations are obtained, that is, W groups of simulation schemes. The segmented fracturing horizontal well gas production under the W groups of simulation schemes is obtained by CMG-IMEX software.

[0081] The segmented fracturing horizontal well gas production under the W groups of simulation schemes is fitted by the production decline model, that is, the segmented fracturing horizontal well gas production is fitted W times by the production decline model, and the initial production q i , the production decline rate D i and the decline index b corresponding to each group of gas production are obtained, that is, W initial productions q i , production decline rates D i and decline indexes b are obtained.

[0082] From the W initial productions q i , the corresponding main control factors of W' initial productions q i are selected as training samples, and the W' initial productions q i are used as labels to train the artificial neural network model, and the trained artificial neural network model corresponding to the initial production q i is obtained.

[0083] Similarly, the trained artificial neural network model corresponding to the production decline rate D i and the trained artificial neural network model corresponding to the decline index b are obtained.

[0084] Step S006, the trained artificial neural network model is detected for credibility;

[0085] In some specific embodiments, step S006 specifically includes:

[0086] The parameters of the production decline model under the W groups of simulation schemes are obtained by the trained artificial neural network model corresponding to each of the parameters;

[0087] Based on the parameters of the production decline model of the W group, the cumulative gas production of the segmented fractured horizontal well in different preset years is predicted by the production decline model, and serves as a predicted value.

[0088] Each of the influence factors corresponding to the simulation schemes of the W group is respectively substituted into the basic model of the segmented fractured horizontal well, and the cumulative gas production of the segmented fractured horizontal well in different preset years under the simulation schemes of the W group is respectively predicted, and serves as a true value.

[0089] The error value between the predicted value and the true value is calculated.

[0090] If the error value is less than a threshold value, the credibility detection of the trained artificial neural network model passes.

[0091] In this embodiment, the parameters of the production decline model under the simulation schemes of the W group are obtained by the trained artificial neural network model corresponding to each parameter of the production decline model, that is, the initial production q i The initial production q i of the production decline model under the simulation schemes of the W group is obtained by the corresponding trained artificial neural network model. i The production decline rate D i of the production decline model under the simulation schemes of the W group is obtained by the corresponding trained artificial neural network model. i The decline index b of the production decline model under the simulation schemes of the W group is obtained by the corresponding trained artificial neural network model.

[0092] Based on the initial production q i , the production decline rate D i and the decline index b of the W group, the parameters of the production decline model are adjusted, and the cumulative gas production of the segmented fractured horizontal well in different preset years is predicted by the production decline model after the adjustment, and serves as a predicted value. Specifically, taking different preset years including 1 year, 5 years, 10 years, 15 years and 20 years as an example, after adjusting the production decline model each time, the cumulative gas production of the segmented fractured horizontal well in 1 year, 5 years, 10 years, 15 years and 20 years is respectively predicted, and serves as a predicted value. It is easy to think that there are 5*W predicted values.

[0093] Based on the above table 2, each group of initial production q i , production decline rate D i and decline index b corresponds to a group of influence factors, and each of the W groups of influence factors under the simulation schemes of the W group is substituted into the basic model of the segmented fractured horizontal well, and the cumulative gas production of the segmented fractured horizontal well in 1 year, 5 years, 10 years, 15 years and 20 years under the simulation schemes of the W group is respectively predicted, and serves as a true value. It is easy to think that there are 5*W true values. Among them, refer to Figure 4 , Figure 4This is a schematic diagram of the basic model of the staged fracturing horizontal well of the present invention. Figure 4 As shown, the gray slices are fractures, the cylinders are horizontal wellbores, and the fractures are arranged in parallel and penetrated by the horizontal wellbores. The basic model of staged fracturing horizontal wells is established. The reservoir is assumed to be homogeneous and of equal thickness, with a size of 2000m×400m×15m (length y e × Width x e × height h), the fractures are symmetrical about both sides of the horizontal wellbore and are evenly distributed in parallel along the horizontal wellbore. The gas well maintains a constant bottomhole pressure for 20 years. The relevant basic values ​​are shown in Table 1.

[0094] Calculate the error values ​​between 5*W predicted values ​​and 5*W true values. In this embodiment, the error values ​​include mean absolute percentage error (MAPE), mean absolute error (MAE), and root mean square error (RMSE). For example, the three error calculation results are shown in Table 4:

[0095] Table 4 Three error calculation results

[0096]

[0097]

[0098] The mean absolute percentage error (MAPE) is calculated as follows:

[0099]

[0100] The mean absolute error (MAE) is calculated as follows:

[0101]

[0102] The root mean square error RMSE calculation formula is as follows:

[0103]

[0104] Among them, n is used to represent the number of predicted values ​​or true values, i Used to represent the i-th real value, Forecast i It is used to represent the first predicted value. It is easy to imagine that in this embodiment, n=5*W.

[0105] If the mean absolute percentage error (MAPE), mean absolute error (MAE), and / or root mean square error (RMSE) are less than the corresponding threshold, the trained artificial neural network model passes the credibility test.

[0106] Step S007: If the credibility test is passed, the trained artificial neural network model is used as the trained artificial neural network model.

[0107] In the embodiment, if the credibility detection of the trained artificial neural network model passes, the trained artificial neural network model is taken as a trained artificial neural network model for subsequent actual use, and each parameter of the production decline model is obtained based on the trained artificial neural network model.

[0108] In another embodiment, if the credibility detection of the trained artificial neural network model does not pass, the parameters of the artificial neural network model are adjusted, and the artificial neural network model is retrained.

[0109] In a second aspect, the embodiment of the present application further provides a segmented fractured horizontal well gas production prediction device.

[0110] In an embodiment, the segmented fractured horizontal well gas production prediction device comprises: Figure 5 Figure 5 FIG. 1 is a functional module schematic diagram of an embodiment of the segmented fractured horizontal well gas production prediction device of the present application. As shown in FIG. 1, the segmented fractured horizontal well gas production prediction device comprises: Figure 5

[0111] The artificial neural network model module 10 is configured to input each master control factor corresponding to each parameter of the production decline model into the trained artificial neural network model corresponding to the parameter, to obtain the parameter of the production decline model output by each trained artificial neural network model;

[0112] The production decline model module 20 is configured to obtain the segmented fractured horizontal well gas production prediction value through the production decline model based on the parameter of the production decline model.

[0113] Optionally, in an embodiment, the segmented fractured horizontal well gas production prediction device further comprises an artificial neural network model training module configured to:

[0114] obtain the value range corresponding to each influence factor, wherein the influence factor is an influencing factor of the segmented fractured horizontal well gas production;

[0115] obtain the segmented fractured horizontal well gas production under N groups of simulation schemes based on the upper limit and the lower limit of the value of each influence factor;

[0116] fit each of the N groups of the gas production through the production decline model, to determine each parameter of the production decline model corresponding to each group of the gas production;

[0117] determine the master control factor corresponding to each parameter based on the influence degree of each influence factor on each parameter;

[0118] train the artificial neural network model based on the master control factor corresponding to each parameter and each parameter, to obtain a trained artificial neural network model corresponding to each parameter.​​

[0119] performing credibility detection on the trained artificial neural network model;

[0120] if the credibility detection passes, taking the trained artificial neural network model as a trained artificial neural network model.

[0121] Optionally, in an embodiment, the artificial neural network model training module is specifically configured to:

[0122] based on each influence factor taking the upper limit value and each influence factor taking the lower limit value, obtaining N groups of influence factor combinations through two-level fractional factorial design;

[0123] obtaining N groups of segmented fracturing horizontal well gas production rates corresponding to the N groups of influence factor combinations through the CMG-IMEX software.

[0124] Optionally, in an embodiment, the artificial neural network model training module is specifically configured to:

[0125] selecting any parameter as a target parameter, and substituting each influence factor into a preset formula to calculate the influence degree of each influence factor on the target parameter, the preset formula being as follows:

[0126]

[0127] wherein, Effect X is used to represent the influence degree of the influence factor X on the target parameter OF, N is used to represent the number of times of gas production rate fitting through the production decline model, OF min is used to represent the value of the target parameter OF corresponding to the influence factor X taking the lower limit value, OF max is used to represent the value of the target parameter OF corresponding to the influence factor X taking the upper limit value;

[0128] selecting the first m influence factors with the largest influence degree on the target parameter as the main control factors corresponding to the target parameter;

[0129] selecting any parameter from the parameters that have not been selected as a target parameter, and returning to perform the step of substituting each influence factor into the preset formula to calculate the influence degree of each influence factor on the target parameter, until the main control factors corresponding to each parameter are obtained.

[0130] Optionally, in an embodiment, the artificial neural network model training module is specifically configured to:

[0131] for any parameter in each parameter of the production decline model, obtaining the segmented fracturing horizontal well gas production rate under W groups of simulation schemes based on the upper limit value and the lower limit value of the main control factor corresponding to the parameter.

[0132] fitting the gas production of each of the W groups by a production decline model, to determine the parameters of the production decline model corresponding to the gas production of each group;

[0133] selecting W' main control factors corresponding to W' parameters as training samples, and training the artificial neural network model with W' parameters as labels, to obtain the trained artificial neural network model corresponding to the parameters;

[0134] By analogy, the trained artificial neural network model corresponding to each of the parameters is obtained.

[0135] Optionally, in an embodiment, the artificial neural network model training module is specifically configured to:

[0136] obtaining the parameters of the production decline model under the W groups of simulation schemes by the trained artificial neural network model corresponding to each of the parameters;

[0137] predicting the cumulative gas production of the segmented fractured horizontal well at different preset years based on the parameters of the W groups of production decline models by the production decline model, and taking the prediction value;

[0138] respectively, the cumulative gas production of the segmented fractured horizontal well at different preset years under the W groups of simulation schemes is predicted by respectively substituting the corresponding influence factors under the W groups of simulation schemes into the basic model of the segmented fractured horizontal well, and taking the true value;

[0139] calculating the error value between the prediction value and the true value;

[0140] If the error value is less than a threshold value, the trained artificial neural network model passes the credibility detection.

[0141] The functions of each module in the segmented fractured horizontal well gas production prediction device correspond to the steps in the segmented fractured horizontal well gas production prediction method, and the functions and implementation processes will not be described here.

[0142] In a third aspect, the embodiments of the present application also provide an electronic device, which has a structure as shown in Figure 6 The processor is configured to read and execute a computer program stored in the memory to implement the method for visual authorization of train dispatching.

[0143] In a fourth aspect, the embodiments of the present application also provide a computer storage medium, which stores computer executable instructions, and the computer executable instructions implement the method for visual authorization of train dispatching when executed.

[0144] Finally, it should be noted that in some of the processes described in the embodiments herein, there can be additional or fewer processes, and the processes described can be combined or performed in an order other than the described order. Additionally, the processes described can be performed in real time or off-line.

[0145] The above descriptions are only the preferred embodiments of the application, not intended to limit the application. Although the application has been described by referring to the aforesaid embodiments, for those skilled in the art, the technical solutions recorded in the foregoing embodiments can be modified or some technical features can be replaced equivalently, and any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. A method for predicting gas production from a horizontally fractured well with multiple segments, characterized in that, The method comprises: inputting each main control factor corresponding to each parameter of the production decline model into a trained artificial neural network model corresponding to the parameter to obtain a parameter of the production decline model output by each trained artificial neural network model; obtaining a segmented fracturing horizontal well gas production prediction value through the production decline model based on the parameter of the production decline model.

2. The method of predicting gas production rate of a sectionally fractured horizontal well according to claim 1, characterized in that, The method further comprises: obtaining a value range corresponding to each influencing factor, wherein the influencing factor is an influencing factor of segmented fracturing horizontal well gas production; obtaining segmented fracturing horizontal well gas production under N groups of simulation schemes based on upper and lower limits of each influencing factor; fitting N groups of the gas production through a production decline model respectively to determine each parameter of the production decline model corresponding to each group of the gas production; obtaining a main control factor corresponding to each parameter based on the influence degree of each influencing factor on each parameter; training an artificial neural network model based on the main control factor corresponding to each parameter and each parameter to obtain a trained artificial neural network model corresponding to each parameter; detecting the trained artificial neural network model for credibility; if the credibility detection passes, using the trained artificial neural network model as a trained artificial neural network model.

3. The method of predicting gas production rate of a sectionally fractured horizontal well according to claim 2, characterized in that, The step of obtaining segmented fracturing horizontal well gas production under N groups of simulation schemes based on upper and lower limits of each influencing factor comprises: obtaining N groups of influencing factor combinations through two-level partial factor design based on each influencing factor with an upper limit value and each influencing factor with a lower limit value; obtaining N groups of segmented fracturing horizontal well gas production corresponding to the N groups of influencing factor combinations through CMG-IMEX software.

4. The method of predicting gas production rate of a sectionally fractured horizontal well according to claim 2, wherein, The step of obtaining a main control factor corresponding to each parameter based on the influence degree of each influencing factor on each parameter comprises: selecting any parameter as a target parameter, and substituting each influencing factor into a preset formula to calculate the influence degree of each influencing factor on the target parameter, wherein the preset formula is as follows: wherein Effect X is used to represent the degree of influence of the influence factor X on the target parameter OF, N is used to represent the number of times of gas production fitting by the production decline model, OF min is used to represent the value of the target parameter OF corresponding to the lower limit value of the influence factor X, OF max is used to represent the value of the target parameter OF corresponding to the upper limit value of the influence factor X; selecting the first m influencing factors with the greatest influence degree on the target parameter as the main control factors corresponding to the target parameter; selecting any parameter as a target parameter from the parameters that have not been selected, and returning to the step of substituting each influencing factor into the preset formula to calculate the influence degree of each influencing factor on the target parameter until the main control factors corresponding to each parameter are obtained.

5. The method of predicting gas production rate of a sectionally fractured horizontal well according to claim 2, wherein, The step of training an artificial neural network model based on the main control factor corresponding to each parameter and each parameter to obtain a trained artificial neural network model corresponding to each parameter comprises: for any parameter of the parameters of the production decline model, obtaining segmented fracturing horizontal well gas production under W groups of simulation schemes based on upper and lower limit values of the main control factor corresponding to the parameter; fitting W groups of the gas production through a production decline model respectively to determine the parameter of the production decline model corresponding to each group of the gas production; W parameters are selected from the W parameters ’ The master factor corresponding to the W parameters is taken as a training sample, and the W ’ The parameters are taken as labels to train an artificial neural network model, and a trained artificial neural network model corresponding to the parameters is obtained. by analogy, obtaining a trained artificial neural network model corresponding to each parameter.

6. The method of predicting gas production from a sectionally fractured horizontal well according to claim 5, wherein, The step of performing credibility detection on the trained artificial neural network model comprises: obtaining each parameter of the production decline model under W groups of simulation schemes through each trained artificial neural network model corresponding to the parameter; predicting the cumulative gas production of the segmented fractured horizontal well under different preset years through the production decline model based on the parameters of the W groups of production decline models, and taking the prediction value; respectively inputting each corresponding influence factor under W groups of simulation schemes into the basic model of the segmented fractured horizontal well, respectively predicting the cumulative gas production of the segmented fractured horizontal well under different preset years under W groups of simulation schemes, and taking the true value; calculating the error value between the prediction value and the true value; if the error value is less than a threshold value, the credibility detection of the trained artificial neural network model is passed.

7. A device for predicting gas production from a horizontally fractured well with multiple segments, characterized in that, The device comprises: An artificial neural network model module configured to input each main control factor corresponding to each parameter of the production decline model into each trained artificial neural network model corresponding to the parameter, and obtain the parameter of the production decline model output by each trained artificial neural network model. A production decline model module configured to obtain the segmented fractured horizontal well gas production prediction value through the production decline model based on the parameters of the production decline model.

8. The device for predicting gas production from a sectionally fractured horizontal well according to claim 7, characterized in that, The device further comprises an artificial neural network model training module configured to: obtain the value range corresponding to each influence factor, wherein the influence factor is an influencing factor of the segmented fractured horizontal well gas production; obtain the segmented fractured horizontal well gas production under N groups of simulation schemes based on the upper and lower limits of the value of each influence factor; fit N groups of the gas production through the production decline model to determine each parameter of the production decline model corresponding to each group of gas production; determine the main control factor corresponding to each parameter based on the influence degree of each influence factor on each parameter; train the artificial neural network model based on the main control factor corresponding to each parameter and each parameter to obtain a trained artificial neural network model corresponding to each parameter; perform credibility detection on the trained artificial neural network model; if the credibility detection is passed, the trained artificial neural network model is taken as the trained artificial neural network model.

9. An electronic device, comprising: comprises: a memory and a processor; The processor is configured to read and execute the computer program stored in the memory to implement the segmented fractured horizontal well gas production prediction method in any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer executable instructions, and the computer executable instructions implement the segmented fractured horizontal well gas production prediction method in any one of claims 1-6 when executed.

Images