Method for training casing inner diameter calculation model and method for determining casing inner diameter

Abstract

The embodiment of the application provides a casing inner diameter calculation model training method and a method for determining the casing inner diameter, and belongs to the technical field of computers. The casing inner diameter calculation model training method comprises the following steps: acquiring microseismic data and a shear modulus, determining a distance range of fault slip based on the microseismic data and the shear modulus, acquiring a geological parameter variation range of geological engineering parameters and a wellbore parameter variation range of wellbore parameters, determining engineering parameters based on the distance range, the geological parameter variation range and the wellbore parameter variation range, performing numerical simulation according to the engineering parameters to obtain a casing inner diameter result corresponding to the engineering parameters, training a BP neural network model based on the engineering parameters and the casing inner diameter result to obtain a casing inner diameter calculation model. Through the casing inner diameter calculation model, the calculation error is effectively reduced, and the convenience of casing inner diameter calculation is improved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and more specifically to a training method for a casing inner diameter calculation model and a method for determining the casing inner diameter. Background Technology

[0002] Horizontal wells combined with hydraulic fracturing have become core technologies for unconventional oil and gas development. However, casing deformation is a frequent problem during hydraulic fracturing. Research indicates that the main form of casing deformation is shear deformation, induced by hydraulic fracturing-induced fault slip, which then shears the casing. The reduced inner diameter of the sheared casing prevents the insertion of bridge plugs, hindering fracturing operations and resulting in "missing sections," significantly reducing oil and gas production. To ensure successful fracturing, smaller diameter bridge plugs are typically used for wells with shear deformation that prevent bridge plug insertion. However, due to the unpredictable amount of casing deformation after shearing, replacement bridge plugs often fail to penetrate, necessitating secondary or even tertiary replacements. Furthermore, using multi-arm caliper tools to measure the diameter at the shear deformation site significantly increases well completion costs.

[0003] To address this, the casing inner diameter after shear deformation following fault slip can be calculated. However, there is currently no analytical solution for this calculation, so numerical simulation is the only option. Current research largely relies on commercially available numerical simulation software to study the degree of casing deformation and its influencing factors under different conditions. This approach suffers from drawbacks such as numerous engineering parameters, complex parameter adjustments, slow calculation speed, and the need for post-processing, making it difficult for non-technical personnel to calculate the casing inner diameter. To ensure that casing inner diameter calculations are truly applicable to field drilling and completion operations, there is an urgent need to establish an efficient and simple method for calculating casing inner diameters. This method should enable field technicians to perform casing inner diameter calculations after simple training, and to optimize bridge plug selection based on the results. Summary of the Invention

[0004] To address the aforementioned shortcomings in the prior art, the purpose of this invention is to provide a training method for a casing inner diameter calculation model and a method for determining the casing inner diameter.

[0005] To achieve the above objectives, the first aspect of the present invention provides a training method for a casing inner diameter calculation model, comprising:

[0006] Acquire microseismic data and shear modulus;

[0007] The distance range of fault slip is determined based on microseismic data and shear modulus;

[0008] Obtain the range of variation of geological parameters and wellbore parameters;

[0009] Engineering parameters are determined based on the range of distance, the range of geological parameter variations, and the range of wellbore parameter variations.

[0010] Numerical simulation is performed based on engineering parameters to obtain the casing inner diameter corresponding to the engineering parameters;

[0011] The BP neural network model is trained based on engineering parameters and casing inner diameter results to obtain the casing inner diameter calculation model.

[0012] In this embodiment of the invention, the microseismic data includes stress drop data and moment magnitude. Determining the fault slip distance range based on the acquired microseismic data and shear modulus includes:

[0013] The acquired microseismic data and shear modulus are input into the distance calculation formula to determine the distance range of fault slip. The distance calculation formula includes:

[0014]

[0015] In the formula, D represents the distance, Δσ represents the stress drop data, G represents the shear modulus, and r represents the fault radius. M w Indicates moment magnitude.

[0016] In this embodiment of the invention, the engineering parameters include geological engineering parameters and wellbore parameters. Numerical simulation is performed based on the engineering parameters to obtain the casing inner diameter result corresponding to the engineering parameters, including:

[0017] Input the wellbore parameters into the preset numerical model;

[0018] Input all geological and engineering parameters within the range of geological parameter variations into the numerical model;

[0019] The results of determining the inner diameter of the first casing corresponding to each geological engineering parameter based on the wellbore parameters;

[0020] Update the wellbore parameters based on the range of wellbore parameter changes until all wellbore parameters within the range of wellbore parameter changes have been traversed.

[0021] The casing inner diameter result is determined based on the first casing inner diameter result corresponding to all wellbore parameters.

[0022] In this embodiment of the invention, the engineering parameters also include distance parameters. The casing inner diameter result is determined based on the first casing inner diameter result corresponding to all wellbore parameters, including:

[0023] Input the distance parameters into the numerical model;

[0024] The inner diameter of the second casing is determined based on the inner diameter of the first casing corresponding to the distance parameter and all wellbore parameters.

[0025] Update the distance parameters based on the distance range until all distance parameters in the distance range have been traversed;

[0026] Use the second casing inner diameter result corresponding to all distance parameters as the casing inner diameter result.

[0027] In this embodiment of the invention, the BP neural network model is trained based on engineering parameters and casing inner diameter results, including:

[0028] The engineering parameters are input into the BP neural network model to obtain the output parameters of the BP neural network model;

[0029] Determine whether the error between the output parameters and the sleeve inner diameter result is within the preset error range;

[0030] If the error is not within the preset error range, adjust the connection weights in the BP neural network model based on the error until the error is within the preset error range or the number of adjustments reaches the preset number of learning iterations.

[0031] In this embodiment of the invention, the BP neural network model includes an input layer, a hidden layer, and an output layer. Engineering parameters are input into the BP neural network model to obtain the output parameters of the BP neural network model, including:

[0032] Determine the first threshold of the input layer neurons; and

[0033] Determine the second threshold of the output layer neurons;

[0034] Determine the first connection weights between the input layer and the hidden layer; and

[0035] Determine the second connection weight coefficients between the hidden layer and the output layer;

[0036] The output parameters of the hidden layer are determined based on engineering parameters, the first connection coefficient, the first threshold, and the hidden layer activation function. The input layer comprises multiple input layer neurons, and the hidden layer comprises multiple hidden layer neurons. The formula for calculating the output parameters of the hidden layer neurons is as follows:

[0037]

[0038] Where p represents the engineering parameters, and i represents the i-th neuron in the hidden layer. Let represent the output parameter of the i-th hidden layer neuron with engineering parameters p, g(.) represent the hidden layer activation function: the sigmoid function, j represent the j-th input layer neuron, m represent the number of input layer neurons, and Wij This represents the first connection weight coefficient between the i-th hidden layer neuron and the j-th input layer neuron. θ represents the engineering parameters of the j-th input layer neuron. i This represents the first threshold of the i-th hidden layer neuron, and n represents the number of hidden layer neurons;

[0039] The output parameters of the output layer are determined based on the output parameters of the hidden layer, the second connection weight coefficient, the second threshold, and the output layer activation function. These parameters serve as the output parameters of the BP neural network model. The output layer consists of output layer neurons, and the formula for calculating the output parameters of these neurons is as follows:

[0040]

[0041] Where p represents the engineering parameters, and k represents the k-th neuron in the output layer. This represents the output parameter of the k-th output layer neuron when the engineering parameter is p. The output layer activation function is a linear activation function, where i represents the i-th hidden layer neuron, n represents the number of hidden layer neurons, and W... ki θ represents the second connection weight coefficient between the k-th output layer neuron and the i-th hidden layer neuron. k This represents the second threshold of the k-th output layer neuron, o i p This represents the output parameter of the i-th hidden layer neuron when the engineering parameter is p.

[0042] In this embodiment of the invention, the engineering parameters include multiple parameters, and determining whether the error between the output parameters and the sleeve inner diameter result is within a preset error range includes:

[0043] Determine the quantity of all engineering parameters;

[0044] Based on the number of engineering parameters, all output parameters, and all casing inner diameter results, determine whether the error between the output parameters and the casing inner diameter results is within the preset error range;

[0045] The formulas for calculating error include:

[0046]

[0047] Among them, e p This represents the error between the output parameter and the inner diameter of the sleeve, where L represents the number of engineering parameters and k represents the k-th output layer neuron. This represents the output parameters of the k-th output layer neuron. This indicates the casing inner diameter when the engineering parameter is p.

[0048] A second aspect of the present invention provides a method for determining the inner diameter of a sleeve, comprising:

[0049] Obtain actual project parameters;

[0050] The actual engineering parameters are input into the casing inner diameter calculation model to obtain the casing inner diameter corresponding to the actual engineering parameters; wherein the casing inner diameter calculation model is obtained by the training method of the casing inner diameter calculation model as described in the above embodiment.

[0051] A third aspect of the present invention provides a training method configured to perform a casing inner diameter calculation model as described in the above embodiments, or a method for determining the casing inner diameter.

[0052] A fourth aspect of the present invention provides a machine-readable storage medium storing instructions for causing a machine to perform a training method for a casing inner diameter calculation model as described in the above embodiments or a method for determining the casing inner diameter.

[0053] The above technical solution acquires microseismic data and shear modulus, determines the fault slip distance range based on the microseismic data and shear modulus, obtains the variation ranges of geological parameters and wellbore parameters, and determines engineering parameters based on the distance range, geological parameter variation range, and wellbore parameter variation range. Numerical simulation is performed based on the engineering parameters to obtain the casing inner diameter result corresponding to the engineering parameters. A BP neural network model is trained based on the engineering parameters and casing inner diameter result to obtain the casing inner diameter calculation model. A training dataset for the casing inner diameter calculation model is established using numerical simulation results under different parameter combinations. The BP neural network algorithm is used to extract features from the dataset to establish the casing inner diameter calculation model. This effectively reduces calculation errors and saves calculation time, greatly improving calculation efficiency. It overcomes the shortcomings of traditional calculation software in calculating casing stress, such as numerous engineering parameters, complex parameter adjustment, slow calculation speed, and the need for post-processing, thus improving the convenience of casing inner diameter calculation.

[0054] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0055] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0056] Figure 1 This is a flowchart illustrating a training method for a casing inner diameter calculation model according to an embodiment of the present invention.

[0057] Figure 2 This is a schematic diagram of a fault slip scenario according to an embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram illustrating the range of geological parameter variations according to an embodiment of the present invention;

[0059] Figure 4 This is a schematic diagram illustrating the range of wellbore parameter variations according to an embodiment of the present invention;

[0060] Figure 5 This is a schematic diagram of a BP neural network architecture according to an embodiment of the present invention;

[0061] Figure 6 This is a flowchart illustrating a method for determining the inner diameter of a casing according to an embodiment of the present invention. Detailed Implementation

[0062] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0063] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0064] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0065] Figure 1 This is a flowchart illustrating a training method for a casing inner diameter calculation model according to an embodiment of the present invention. Figure 1 As shown in the embodiment of the present invention, a training method for a casing inner diameter calculation model is provided. Taking the application of this method to a processor as an example, the method may include the following steps:

[0066] Step S100: Obtain microseismic data and shear modulus;

[0067] In this embodiment, it should be noted that the microseismic data may include data obtained based on microseismic measurements; the shear modulus is a material constant, which is the ratio of shear stress to strain, one of the mechanical property indicators of a material, and is the ratio of shear stress to shear strain under shear stress within the elastic deformation ratio limit. It characterizes the material's ability to resist shear strain. The shear modulus can be obtained through testing or by converting the elastic modulus formula.

[0068] Step S200: Determine the distance range of fault slip based on microseismic data and shear modulus;

[0069] refer to Figure 2 In a fault slip scenario, the casing encased in a cement ring is interspersed between the fixed fault segment and the slipping fault segment. The distance the slipping fault segment slides along the slip direction is the fault slip distance.

[0070] In this embodiment, it should be noted that the distance of fault slip determines the change in the inner diameter of the casing after shearing. When determining the inner diameter of the casing, the fault slip distance needs to be taken into consideration.

[0071] Specifically, microseismic data includes stress drop data and moment magnitude. Based on the acquired microseismic data and shear modulus, the distance range of fault slip is determined, including:

[0072] The acquired microseismic data and shear modulus are input into the distance calculation formula to determine the distance range of fault slip. The distance calculation formula includes:

[0073]

[0074] In the formula, D represents the distance, Δσ represents the stress drop data, G represents the shear modulus, and r represents the fault radius. M w Indicates moment magnitude.

[0075] It should be noted that stress drop includes the stress difference at a point on the fault before and after rupture during an earthquake; moment magnitude is determined by the magnitude of the seismic moment, M0 representing the seismic moment, a physical quantity describing the mechanical strength of an earthquake. Specifically, after acquiring stress drop data, moment magnitude, and shear modulus, the processor can determine the fault slip distance according to the distance calculation formula. When any value of stress drop data, moment magnitude, or shear modulus changes, the fault slip distance will change accordingly. Therefore, based on all the stress drop data, moment magnitude, and corresponding shear modulus values, the slip distances of multiple faults can be determined, thus obtaining the range of fault slip distances.

[0076] Step S300: Obtain the variation range of geological parameters and wellbore parameters;

[0077] In this embodiment, it should be noted that there are many factors affecting casing shear deformation. Based on the parameter variation range and parameter properties, they can be broadly categorized into geological engineering parameters and wellbore parameters. The variation range of geological parameters includes the variation range of geological engineering parameters; the variation range of wellbore parameters includes the variation range of wellbore parameters. Geological engineering parameters may include: stress drop, moment magnitude, formation shear modulus, casing internal pressure, cement sheath elastic modulus, cement sheath Poisson's ratio, formation elastic modulus, and formation Poisson's ratio; wellbore parameters may include: wellbore size, casing outer diameter, casing wall thickness, and casing steel grade. (Reference) Figure 3 In one embodiment, the range of geological engineering parameter variations and the step size of parameter variations are determined based on the analysis of drilling and completion operation data. (Reference) Figure 4 In one embodiment, the casing specifications are based on commonly used wellbore dimensions, casing outer diameter, casing wall thickness, and casing steel grade in engineering sites. The casing specification types include casing outer diameter, casing wall thickness, and casing steel grade. According to the wellbore dimensions and drilling standards, the casing specification type used for each type of wellbore size is fixed, with a one-to-one correspondence between different wellbore dimensions and casing specifications. Conventional fracturing wells have wellbore dimensions of 152.4mm, 215.9mm, 244.5mm, and 311.2mm, resulting in four different types of casing specifications. Figure 4 A, B, C, and D in the diagram mainly represent the number of casing specifications and types under different wellbore sizes in conventional fracturing wells.

[0078] Step S400: Determine engineering parameters based on distance range, geological parameter variation range, and wellbore parameter variation range;

[0079] In this embodiment, it should be noted that the engineering parameters include a combination of distance parameters within the distance range, geological engineering parameters within the geological variation range, and wellbore parameters within the wellbore parameter variation range. There are multiple engineering parameters, including all combinations of distance parameters, geological engineering parameters, and wellbore parameters.

[0080] Step S500: Perform numerical simulation based on engineering parameters to obtain the casing inner diameter result corresponding to the engineering parameters;

[0081] In this embodiment, it should be noted that the casing inner diameter result includes the casing inner diameter obtained after numerical simulation. The casing inner diameter varies under different distance parameters, geological engineering parameters, and wellbore parameters. When the engineering parameters include all combinations of distance parameters, geological engineering parameters, and wellbore parameters, the casing inner diameter result corresponding to each engineering parameter can be obtained separately after numerical simulation based on the engineering parameters.

[0082] Step S600: Train the BP neural network model based on engineering parameters and casing inner diameter results to obtain the casing inner diameter calculation model.

[0083] In this embodiment, it should be noted that the BP (Back Propagation) neural network is a multi-layer feedforward neural network trained using an error backpropagation algorithm. After determining the casing inner diameter result corresponding to the engineering parameters, each set of engineering parameters and casing inner diameter results is considered as a dataset. All datasets are used as training data for the BP neural network model to obtain the trained BP neural network model as the casing inner diameter calculation model. It is understood that, in one embodiment, all datasets can be divided into training, testing, and validation sets. For example, the datasets can be divided into training, testing, and validation sets in a ratio of 0.7:0.15:0.15 to complete the training, testing, and validation of the BP neural network model.

[0084] The training method for the aforementioned casing inner diameter calculation model involves acquiring microseismic data and shear modulus, determining the fault slip distance range based on these data, obtaining the variation ranges of geological parameters and wellbore parameters, determining engineering parameters based on these ranges, performing numerical simulations to obtain the corresponding casing inner diameter results, and training a BP neural network model based on these engineering parameters and casing inner diameter results to obtain the casing inner diameter calculation model. A training dataset for the casing inner diameter calculation model is established using numerical simulation results under different parameter combinations. The BP neural network algorithm is then used to extract features from the dataset to establish the casing inner diameter calculation model. This method effectively reduces calculation errors and saves calculation time, significantly improving calculation efficiency. It overcomes the shortcomings of traditional calculation software for casing stress calculations, such as numerous engineering parameters, complex parameter adjustments, slow calculation speed, and the need for post-processing, thus enhancing the convenience of casing inner diameter calculation.

[0085] In one embodiment, the engineering parameters include geological engineering parameters and wellbore parameters. Numerical simulation is performed based on these engineering parameters to obtain the casing inner diameter corresponding to the engineering parameters, including:

[0086] Input the wellbore parameters into the preset numerical model;

[0087] Input all geological and engineering parameters within the range of geological parameter variations into the numerical model;

[0088] The results of determining the inner diameter of the first casing corresponding to each geological engineering parameter based on the wellbore parameters;

[0089] Update the wellbore parameters based on the range of wellbore parameter changes until all wellbore parameters within the range of wellbore parameter changes have been traversed.

[0090] The casing inner diameter result is determined based on the first casing inner diameter result corresponding to all wellbore parameters.

[0091] In this embodiment, it should be noted that a numerical model is a model that uses mathematical symbols and language rules, such as variables, equations and inequalities, and mathematical operations, to describe the characteristics and internal relationships of things. In this embodiment, a numerical model is used to determine the casing inner diameter result corresponding to the engineering parameters. Different engineering parameters correspond to different casing inner diameter results when numerically simulated. Each set of engineering parameters and casing inner diameter results is considered as a dataset. All datasets are used as training data for a BP neural network model. By determining a dataset with a larger data volume, the accuracy of the BP neural network model during training is improved.

[0092] Specifically, the processor inputs wellbore parameters into a preset numerical model, and also inputs all geological engineering parameters within the range of geological parameter variations into the numerical model, obtaining the casing inner diameter result corresponding to the combination of the wellbore parameters and each geological engineering parameter, which is used as the first casing inner diameter result. (Reference) Figure 3 Based on the variation range and step size of geological parameters, during numerical simulation, changing the geological engineering parameters can generate 10*6*6*5*5*5*5*5 = 1,125,000 sets of data, forming a dataset. Further, the wellbore parameters are updated based on their variation range until all wellbore parameters within that range have been traversed. The first casing inner diameter result corresponding to all wellbore parameters is then used as the casing inner diameter result. (Reference) Figure 4 Based on the dataset formed by changing geological engineering parameters, and combined with the update of wellbore parameters, 1125000*(A+B+C+D) sets of data can be determined, thus forming a relatively large dataset. This is understandable. Figure 4 The number of sleeve specifications is limited, but in actual applications, the number of sleeve specifications can be increased according to specific needs.

[0093] In this embodiment, numerical simulation is performed by traversing all wellbore parameters and geological engineering parameters to obtain the casing inner diameter result, which provides a large amount of training data for subsequent training of the BP neural network model and improves the accuracy of the casing inner diameter calculation model.

[0094] In one embodiment, the engineering parameters also include distance parameters, and the casing inner diameter result is determined based on the first casing inner diameter result corresponding to all wellbore parameters, including:

[0095] Input the distance parameters into the numerical model;

[0096] The inner diameter of the second casing is determined based on the inner diameter of the first casing corresponding to the distance parameter and all wellbore parameters.

[0097] Update the distance parameters based on the distance range until all distance parameters in the distance range have been traversed;

[0098] Use the second casing inner diameter result corresponding to all distance parameters as the casing inner diameter result.

[0099] It should be noted that the first casing inner diameter results generated under different combinations of wellbore parameters and geological engineering parameters form a large dataset. In this embodiment, the dataset can be further expanded by incorporating changes in distance parameters. Specifically, the processor inputs distance parameters into the numerical model to determine the second casing inner diameter results based on the first casing inner diameter results corresponding to the distance parameters and all wellbore parameters. The distance parameters are updated based on the distance range until all distance parameters within the range have been traversed, and the second casing inner diameter results corresponding to all distance parameters are used as the casing inner diameter results.

[0100] In this embodiment, numerical simulation is performed by traversing all distance parameters to obtain the casing inner diameter result. The training data for training the BP neural network model is further increased, thereby improving the accuracy of the casing inner diameter calculation model.

[0101] In one embodiment, training the BP neural network model based on engineering parameters and casing inner diameter results includes:

[0102] The engineering parameters are input into the BP neural network model to obtain the output parameters of the BP neural network model;

[0103] Determine whether the error between the output parameters and the sleeve inner diameter result is within the preset error range;

[0104] If the error is not within the preset error range, adjust the connection weights in the BP neural network model based on the error until the error is within the preset error range or the number of adjustments reaches the preset number of learning iterations.

[0105] It should be noted that the BP neural network is a multi-layer feedforward network that uses the steepest gradient descent method to backpropagate error information to achieve the learning objective. The learning process of the BP neural network includes two stages: forward propagation of input information and backward adjustment of error information. In the forward propagation stage, the input information is passed from the input layer through the hidden layers to the output layer. If the expected output is not obtained at the output layer, the output error will be backpropagated layer by layer until it reaches the input layer. Along the way, the connection weights and thresholds between neurons in each layer are modified to gradually minimize the error. The forward propagation and backward adjustment processes are performed alternately until the output error of the BP neural network is reduced to a preset error range or a preset number of learning iterations is reached. In this embodiment, the BP neural network model is based on the BP neural network, the engineering parameters are the input information, the output parameters are the output obtained by the BP neural network model based on the engineering parameters, and the sleeve inner diameter result corresponding to the engineering parameters is the expected output of the BP neural network model. The preset error range is set in advance to determine whether the connection weight coefficients in the BP neural network model need to be adjusted. Specifically, the processor inputs engineering parameters into the BP neural network model, obtains the output parameters of the BP neural network model, determines the error between the output parameters and the inner diameter of the sleeve, and judges whether the error is within the preset error range. If the error is not within the preset error range, the connection weight coefficients in the BP neural network model are adjusted based on the error until the error is within the preset error range or the number of adjustments reaches the preset number of learning times.

[0106] In this embodiment, the connection weight coefficients in the BP neural network model are adjusted based on the error between the output parameters and the casing inner diameter result, thereby reducing the calculation error of the BP neural network model and improving the calculation accuracy.

[0107] In one embodiment, the BP neural network model includes an input layer, hidden layers, and an output layer. Engineering parameters are input into the BP neural network model to obtain its output parameters, including:

[0108] Determine the first threshold of the input layer neurons; and

[0109] Determine the second threshold of the output layer neurons;

[0110] Determine the first connection weights between the input layer and the hidden layer; and

[0111] Determine the second connection weight coefficients between the hidden layer and the output layer;

[0112] refer to Figure 5 The BP neural network architecture includes an input layer, hidden layers, and an output layer. The input layer contains m neurons, the output layer contains 1 neuron, and the hidden layers contain n neurons, where x1, x2, ..., xn are the input, output, and output neurons respectively.m Let y be the input of the BP neural network, t be the target output of the output layer neuron, and e be the error of the output layer neuron. Initially, each layer of neurons has initial connection weights and thresholds. The threshold of the input layer neuron is the first threshold; the threshold of the output layer neuron is the second threshold; the connection weights between the input layer and the hidden layer are the first connection weights; and the connection weights between the hidden layer and the output layer are the second connection weights.

[0113] The output parameters of the hidden layer are determined based on engineering parameters, the first connection coefficient, the first threshold, and the hidden layer activation function. The input layer comprises multiple input layer neurons, and the hidden layer comprises multiple hidden layer neurons. The formula for calculating the output parameters of the hidden layer neurons is as follows:

[0114]

[0115] Where p represents the engineering parameters, and i represents the i-th neuron in the hidden layer. Let represent the output parameter of the i-th hidden layer neuron with engineering parameters p, g(.) represent the hidden layer activation function: the sigmoid function, j represent the j-th input layer neuron, m represent the number of input layer neurons, and W ij This represents the first connection weight coefficient between the i-th hidden layer neuron and the j-th input layer neuron. θ represents the engineering parameters of the j-th input layer neuron. i This represents the first threshold of the i-th hidden layer neuron, and n represents the number of hidden layer neurons;

[0116] The output parameters of the output layer are determined based on the output parameters of the hidden layer, the second connection weight coefficient, the second threshold, and the output layer activation function. These parameters serve as the output parameters of the BP neural network model. The output layer consists of output layer neurons, and the formula for calculating the output parameters of these neurons is as follows:

[0117]

[0118] Where p represents the engineering parameters, and k represents the k-th neuron in the output layer. This represents the output parameter of the k-th output layer neuron when the engineering parameter is p. The output layer activation function is a linear activation function, where i represents the i-th hidden layer neuron, n represents the number of hidden layer neurons, and W... ki θ represents the second connection weight coefficient between the k-th output layer neuron and the i-th hidden layer neuron. k This represents the second threshold of the k-th output layer neuron. This represents the output parameter of the i-th hidden layer neuron when the engineering parameter is p.

[0119] In one embodiment, the engineering parameters include multiple parameters, and determining whether the error between the output parameters and the casing inner diameter result is within a preset error range includes:

[0120] Determine the quantity of all engineering parameters;

[0121] Based on the number of engineering parameters, all output parameters, and all casing inner diameter results, determine whether the error between the output parameters and the casing inner diameter results is within the preset error range;

[0122] It should be noted that the engineering parameters include all combinations of distance parameters, geological engineering parameters, and wellbore parameters. There are multiple combinations. By determining the number of all engineering parameters, and based on the number of engineering parameters, all output parameters, and all casing inner diameter results, it can be determined whether the error between all output parameters and casing inner diameter results is within the preset error range.

[0123] The formulas for calculating error include:

[0124]

[0125] Among them, e p This represents the error between the output parameter and the inner diameter of the sleeve, where L represents the number of engineering parameters and k represents the k-th output layer neuron. This represents the output parameters of the k-th output layer neuron. This indicates the casing inner diameter when the engineering parameter is p.

[0126] In this embodiment, it should be noted that the connection weight coefficients include a first connection weight coefficient and a second connection weight coefficient. The connection weight coefficients in the BP neural network model adjusted based on the error include:

[0127] The first adjustment amount is determined based on the preset learning rate, error, and the first adjustment formula.

[0128] The first connection weight coefficient is adjusted based on the first adjustment amount;

[0129] It should be noted that the preset learning rate includes the learning rate of the BP neural network, which is generally taken as a value between 0.01 and 0.1. The first connection weight coefficient includes the connection weight coefficient between the output layer and the hidden layer, and the first adjustment amount includes the adjustment amount for the first connection weight coefficient. After determining the first adjustment amount, the first adjustment amount is added to the original first connection weight coefficient to obtain the new first connection weight coefficient.

[0130] The first adjustment formula includes:

[0131]

[0132] Where, Δw ki This represents the first adjustment amount of the first connection weight coefficient between the k-th output layer neuron and the i-th hidden layer neuron, where η represents the preset learning rate. This represents the output parameters of the k-th output layer neuron. This indicates the casing inner diameter when the engineering parameter is p. This represents the derivative of the activation function of the output layer. This represents the output parameter of the i-th hidden layer neuron when the engineering parameter is p;

[0133] The second adjustment amount is determined based on the preset learning rate, error, and the second adjustment formula.

[0134] Adjust the second connection weight coefficient based on the second adjustment amount;

[0135] The second connection weight coefficient includes the connection weight coefficient between the hidden layer and the input layer, and the second adjustment amount includes the adjustment amount for adjusting the second connection weight coefficient. After determining the second adjustment amount, the second adjustment amount is added to the original second connection weight coefficient to obtain the new second connection weight coefficient.

[0136] The second adjustment formula includes:

[0137]

[0138] Where, Δw ij η represents the second adjustment amount of the second connection weight coefficient between the i-th hidden layer neuron and the j-th input layer neuron, L represents the preset learning rate, L represents the number of engineering parameters, and k represents the k-th output layer neuron. This represents the output parameters of the output layer neurons. This indicates the casing inner diameter when the engineering parameter is p. w represents the derivative of the activation function of the output layer. ki Let g'(.) represent the first connection weight coefficient between the k-th output layer neuron and the i-th hidden layer neuron, and g'(.) represent the derivative of the hidden layer activation function. This represents the engineering parameters of the j-th neuron in the input layer.

[0139] Figure 6 This is a schematic flowchart of a method for determining the inner diameter of a casing according to an embodiment of the present invention. Figure 6 As shown in the embodiment of the present invention, a method for determining the inner diameter of a sleeve is provided. Taking the application of this method to a processor as an example, the method may include the following steps:

[0140] Step S700: Obtain the actual parameters of the project;

[0141] Step S800: Input the actual engineering parameters into the casing inner diameter calculation model to obtain the casing inner diameter corresponding to the actual engineering parameters; wherein the casing inner diameter calculation model is obtained by the training method of the casing inner diameter calculation model as described in the above embodiment.

[0142] In this embodiment, it should be noted that the actual engineering parameters include those known in practical applications, including geological engineering parameters and wellbore parameters. The casing inner diameter calculation model is a trained calculation model that can calculate the corresponding casing inner diameter based on the input actual engineering parameters. Specifically, after the processor obtains the actual engineering parameters, it inputs the actual engineering parameters into the casing inner diameter calculation model to obtain the casing inner diameter corresponding to the actual engineering parameters.

[0143] The method described above for determining the inner diameter of the casing effectively reduces calculation errors and saves calculation time by using a trained casing inner diameter calculation model. This greatly improves calculation efficiency and overcomes the shortcomings of traditional calculation software in calculating casing stress, such as numerous engineering parameters, complex parameter adjustments, slow calculation speed, and the need for post-processing. This enhances the convenience of casing inner diameter calculation.

[0144] This invention provides a processor for running a program, wherein the program executes a training method for the casing inner diameter calculation model or a method for determining the casing inner diameter.

[0145] This invention provides a storage medium storing a program that, when executed by a processor, implements a training method for the casing inner diameter calculation model or a method for determining the casing inner diameter.

[0146] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can 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, etc.) containing computer-usable program code.

[0147] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can 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, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0148] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0149] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0150] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0151] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0152] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media 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 technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0153] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0154] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A training method for a casing inner diameter calculation model, characterized in that, include: Acquire microseismic data and shear modulus, wherein the microseismic data includes stress drop data and moment magnitude; The distance range of fault slip is determined based on the microseismic data and the shear modulus; Obtain the range of variation of geological parameters and wellbore parameters; Engineering parameters are determined based on the distance range, the geological parameter variation range, and the wellbore parameter variation range. Numerical simulation is performed based on the engineering parameters to obtain the casing inner diameter corresponding to the engineering parameters; The BP neural network model is trained based on the engineering parameters and the casing inner diameter result to obtain the casing inner diameter calculation model; The determination of the fault slip distance range based on the microseismic data and the shear modulus includes: The acquired microseismic data and shear modulus are input into the distance calculation formula to determine the distance range of fault slip. The distance calculation formula includes: In the formula, D Indicates distance, Δσ Represented as stress drop data, G Indicates shear modulus, r Represents the fault radius, where, , , Indicates moment magnitude.

2. The method according to claim 1, characterized in that, The engineering parameters include geological engineering parameters and wellbore parameters. The step of performing numerical simulation based on the engineering parameters to obtain the casing inner diameter corresponding to the engineering parameters includes: Input the wellbore parameters into the preset numerical model; Input all geological engineering parameters within the range of geological parameter variations into the numerical model; The inner diameter of the first casing corresponding to each of the geological engineering parameters is determined based on the wellbore parameters. The wellbore parameters are updated based on the range of changes in the wellbore parameters until all wellbore parameters within the range of changes in the wellbore parameters have been traversed. The casing inner diameter result is determined based on the first casing inner diameter result corresponding to all wellbore parameters.

3. The method according to claim 2, characterized in that, The engineering parameters also include distance parameters. The step of determining the casing inner diameter result based on the first casing inner diameter result corresponding to all wellbore parameters includes: Input the distance parameters into the numerical model; The inner diameter of the second casing is determined based on the first casing inner diameter result corresponding to the distance parameter and all wellbore parameters. The distance parameters are updated based on the distance range until all distance parameters within the distance range have been traversed. Use the second casing inner diameter result corresponding to all distance parameters as the casing inner diameter result.

4. The method according to claim 1, characterized in that, The training of the BP neural network model based on the engineering parameters and the casing inner diameter includes: The engineering parameters are input into the BP neural network model to obtain the output parameters of the BP neural network model; Determine whether the error between the output parameter and the sleeve inner diameter result is within a preset error range; If the error is not within the preset error range, the connection weight coefficients in the BP neural network model are adjusted based on the error until the error is within the preset error range or the number of adjustments reaches the preset number of learning iterations.

5. The method according to claim 4, characterized in that, The BP neural network model includes an input layer, a hidden layer, and an output layer. The step of inputting the engineering parameters into the BP neural network model to obtain the output parameters of the BP neural network model includes: Determine the first threshold of the input layer neurons; and Determine the second threshold of the output layer neurons; Determine the first connection weight coefficients between the input layer and the hidden layer; and Determine the second connection weight coefficient between the hidden layer and the output layer; The output parameters of the hidden layer are determined based on the engineering parameters, the first connection weight coefficient, the first threshold, and the hidden layer activation function. The input layer includes multiple input layer neurons, and the hidden layer includes multiple hidden layer neurons. The formula for calculating the output parameters of the hidden layer neurons is as follows: in, p Indicates engineering parameters, i Indicates the hidden layer number 1 i One neuron, Indicates the hidden layer number 1 i The hidden layer neurons have engineering parameters of p The corresponding output parameters at that time The hidden layer activation function is represented by the sigmoid function. j Indicates the first j One input layer neuron, m This indicates the number of neurons in the input layer. Wij Indicates the first i The hidden layer neurons and the first j The first connection weight coefficients between each input layer neuron Indicates the first j Engineering parameters of each input layer neuron. θi Indicates the first i The first threshold of each hidden layer neuron n This indicates the number of neurons in the hidden layer; The output parameters of the output layer are determined based on the output parameters of the hidden layer, the second connection weight coefficient, the second threshold, and the output layer activation function, and are used as the output parameters of the BP neural network model. The output layer includes output layer neurons, and the formula for calculating the output parameters of the output layer neurons is as follows: in, p Indicates engineering parameters, k Indicates the output layer number k One output layer neuron, Indicates the first k Each output layer neuron has engineering parameters of p The corresponding output parameters at that time This indicates the output layer activation function: a linear activation function. i Indicates the first i One hidden layer neuron, n This indicates the number of neurons in the hidden layer. Wki Indicates the first k The output layer neuron and the first i The second connection weight coefficients between neurons in the hidden layer θk Indicates the first k The second threshold of each output layer neuron Indicates the first i The hidden layer neurons have engineering parameters of p The corresponding output parameters.

6. The method according to claim 5, characterized in that, The engineering parameters include multiple parameters, and determining whether the error between the output parameter and the sleeve inner diameter result is within a preset error range includes: Determine the quantity of all engineering parameters; Based on the number of the engineering parameters, all the output parameters, and all the casing inner diameter results, determine whether the error between the output parameters and the casing inner diameter results is within a preset error range; The formula for calculating the error includes: in, ep This indicates the error between the output parameter and the sleeve inner diameter result. L Indicates the number of engineering parameters. k Indicates the first k One output layer neuron, Indicates the first k The output parameters of each output layer neuron Indicates the engineering parameters as p The corresponding inner diameter of the casing.

7. A method for determining the inner diameter of a casing, characterized in that, include: Obtain actual project parameters; The actual engineering parameters are input into the casing inner diameter calculation model to obtain the casing inner diameter corresponding to the actual engineering parameters; wherein, the casing inner diameter calculation model is obtained by the training method of the casing inner diameter calculation model as described in any one of claims 1 to 6.

8. A processor, characterized in that, The method is configured to perform a training method for a casing inner diameter calculation model according to any one of claims 1 to 6, or a method for determining the casing inner diameter according to claim 7.

9. A machine-readable storage medium, characterized in that, The machine-readable storage medium stores instructions for causing the machine to perform a training method for the casing inner diameter calculation model as described in any one of claims 1 to 6, or a method for determining the casing inner diameter as described in claim 7.

Images