Longitudinal wave velocity logging curve correction method and device and electronic equipment

Abstract

The invention relates to the technical field of oil-gas exploration and geophysical logging, and discloses a longitudinal wave velocity logging curve correction method and device and electronic equipment, and the method comprises the steps: constructing input sample data and output sample data; performing iterative training on the depth feedforward neural network model through a preset model training strategy according to the input sample data and the output sample data to obtain a longitudinal wave velocity logging curve correction model; and inputting all the expanded-diameter section logging curves meeting preset conditions in the expanded-diameter section logging curve set into a longitudinal wave velocity logging curve correction model for longitudinal wave velocity logging curve correction to obtain a corrected expanded-diameter section longitudinal wave velocity logging curve. According to the invention, the rationality of longitudinal wave velocity logging curve correction can be improved.

Description

Technical Field

[0001] The invention relates to the technical field of oil and gas exploration and geophysical logging, and in particular to a method, device and electronic equipment for correcting a longitudinal wave velocity logging curve. Background Art

[0002] Well logging curves are an important basis for well seismic analysis and reservoir prediction. However, in mudstone formations, diameter expansion and collapse often occur due to changes in drilling speed and other reasons, resulting in abnormal diameter expansion in acoustic wave, density and other curves, affecting synthetic record calibration, rock physics analysis and subsequent reservoir prediction. Therefore, it is necessary to correct the abnormal values ​​of the curves affected by diameter expansion to make them more consistent with geological laws.

[0003] At present, the commonly used methods for P-wave velocity curve correction include parameter fitting method, empirical formula method, rock physics modeling, etc.

[0004] The parameter fitting method is also called the standard layer statistical method. Generally, the non-expanded section is selected as the stable section, and the target curve of the stable section is intersected and fitted with other curves to obtain the fitting relationship between the parameters. This relationship is then applied to the expanded section to correct the target curve of the expanded section. The disadvantage of this method is that the intersection fitting can often only fit simple functional relationships between parameters, while in reality, there may be more complex nonlinear relationships between the parameter curves.

[0005] The most commonly used empirical formula method for calculating P-wave velocity is the Faust formula method, which uses the empirical relationship between P-wave impedance and resistivity in mudstone sections to correct the P-wave velocity curve. The shortcoming of this method is that it only considers the influence of resistivity on P-wave velocity, resulting in low prediction accuracy.

[0006] Therefore, there is an urgent need for a correction method that can reasonably correct the P-wave velocity logging curve of the expansion section to solve the above technical problems. Summary of the invention

[0007] In view of the above problems, the present invention provides a method, device and electronic equipment for correcting the P-wave velocity logging curve, which can improve the rationality of the P-wave velocity logging curve correction in the expansion section, thereby laying a good foundation for subsequent logging response characteristics and rock physics analysis, synthetic record calibration, reservoir prediction and other work.

[0008] The present invention provides a method for correcting a longitudinal wave velocity logging curve, the method comprising:

[0009] Construct input sample data and output sample data;

[0010] Iteratively train the deep feedforward neural network model according to the input sample data and the output sample data through a preset model training strategy to obtain a P-wave velocity logging curve correction model;

[0011] All the expanded diameter section logging curves that meet the preset conditions in the expanded diameter section logging curve set are input into the P-wave velocity logging curve correction model to perform P-wave velocity logging curve correction, and obtain the corrected expanded diameter section P-wave velocity logging curve.

[0012] Furthermore, constructing input sample data and output sample data includes:

[0013] According to the preset construction strategy, the logging curve sets of the stable section and the expanded diameter section in the target layer section of the target well are respectively constructed;

[0014] All stable segment logging curves that meet preset conditions in the stable segment logging curve set are used as input sample data, and the stable segment P-wave velocity logging curves in the stable segment logging curve set are used as output sample data.

[0015] Furthermore, the preset build strategies include:

[0016] The well section in the target layer section of the target well with a well diameter greater than a preset well diameter threshold is regarded as a stable section, and the well section in the target layer section of the target well with a well diameter not greater than the preset well diameter threshold is regarded as an expanded diameter section;

[0017] Extracting the stable segment logging curve corresponding to the stable segment from each logging curve of the target layer segment of the target well respectively, so as to construct a stable segment logging curve set including all the stable segment logging curves;

[0018] The expanded diameter section logging curve corresponding to the expanded diameter section is respectively intercepted from each logging curve of the target layer section of the target well to construct an expanded diameter section logging curve set including all expanded diameter section logging curves.

[0019] Further, determining a stable section logging curve that meets preset conditions includes:

[0020] Deleting the stable segment P-wave velocity logging curve from the stable segment logging curve set to obtain a target curve set;

[0021] Perform correlation analysis on each stable segment logging curve in the target curve set and the stable segment P-wave velocity logging curve to obtain correlation analysis results;

[0022] The stable section logging curve that meets the preset conditions is determined based on the correlation analysis results.

[0023] Furthermore, the preset model training strategy includes a least squares method.

[0024] The present invention also provides a P-wave velocity logging curve correction device, the device comprising:

[0025] A construction module, used for constructing input sample data and output sample data;

[0026] A training module, used to iteratively train the deep feedforward neural network model according to the input sample data and the output sample data through a preset model training strategy to obtain a P-wave velocity logging curve correction model;

[0027] The correction module is used to input the expanded diameter section logging curve that meets the preset conditions in the expanded diameter section logging curve set into the P-wave velocity logging curve correction model to perform P-wave velocity logging curve correction and obtain the corrected expanded diameter section P-wave velocity logging curve.

[0028] Furthermore, the building blocks include:

[0029] A set construction unit is used to respectively construct a set of logging curves of a stable section and an expanded diameter section in a target layer section of a target well according to a preset construction strategy;

[0030] The sample data determination unit is used to use all stable segment logging curves that meet preset conditions in the stable segment logging curve set as input sample data and use the stable segment P-wave velocity logging curve in the stable segment logging curve set as output sample data.

[0031] Furthermore, the building blocks include:

[0032] The curve determination unit is used to delete the stable segment P-wave velocity logging curve from the stable segment logging curve set to obtain a target curve set; and to perform correlation analysis on each stable segment logging curve in the target curve set with the stable segment P-wave velocity logging curve to obtain a correlation analysis result; and to determine the stable segment logging curve that meets the preset conditions according to the correlation analysis result.

[0033] The present invention also provides a computer-readable storage medium, wherein the computer program stored in the computer-readable storage medium implements the steps of the above method when executed by one or more processors.

[0034] The present invention also provides an electronic device, comprising a memory and one or more processors, wherein a computer program is stored in the memory, and when the computer program is executed by the one or more processors, the steps of the above method are executed.

[0035] The P-wave velocity logging curve correction method, device and electronic device provided by the present invention iteratively train the deep feedforward neural network model according to the input sample data and the output sample data through the preset model training strategy to obtain the P-wave velocity logging curve correction model. When correcting the P-wave velocity logging curve of the expansion section, all the expansion section logging curves that meet the preset conditions in the logging curve set of the expansion section are used as model input to obtain the corrected P-wave velocity logging curve of the expansion section. The present invention can fully explore the nonlinear relationship between the input sample data and the output sample data by iteratively training the deep feedforward neural network model, and then apply this relationship to the expansion section, and use the trained model to correct the P-wave velocity logging curve. Compared with the conventional parameter fitting method and the empirical formula method, the present invention takes into account the influence of multiple factors, has higher learning accuracy, and the obtained curve correction result is more reliable, which improves the rationality of the curve correction, thereby laying a good foundation for subsequent logging response characteristics and rock physics analysis, synthetic record calibration, reservoir prediction and other work. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] It should also be noted that, for the convenience of description, only the parts related to the present invention are shown in the drawings. The drawings constituting a part of the specification of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments and descriptions in the present invention are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0038] Figure 1 A flowchart of the steps of a method for correcting a longitudinal wave velocity logging curve provided in the first embodiment of the present invention;

[0039] Figure 2 A flowchart of a preset construction strategy step provided in Embodiment 1 of the present invention;

[0040] Figure 3 A schematic diagram of a deep feedforward neural network model architecture provided in Embodiment 1 of the present invention;

[0041] Figure 4 This is a schematic structural diagram of a P-wave velocity logging curve correction device provided in the second embodiment of the present invention;

[0042] Figure 5 This is a schematic diagram of the structure of the building blocks provided in the second embodiment of the present invention;

[0043] Figure 6 This is another schematic diagram of the structure of building blocks provided in the second embodiment of the present invention;

[0044] Figure 7 The target well target layer logging curve in the third embodiment of the present invention;

[0045] Figure 8 This is a schematic diagram of the training accuracy of the deep feedforward network model in the third embodiment of the present invention;

[0046] Fig. 9 This is a schematic diagram of the correction result of the longitudinal wave velocity curve in the third embodiment of the present invention;

[0047] Fig.10 This is a schematic diagram of the comparison of synthetic records before and after correction in Embodiment 3 of the present invention;

[0048] Fig.11 This is a schematic diagram of the structure of an electronic device provided in Embodiment 6 of the present invention;

[0049] Reference numerals:

[0050] Figures 4 to 6 In: 401-construction module, 402-training module, 403-correction module, 4011-set construction unit, 4012-sample data determination unit, 4013-curve determination unit;

[0051] Fig.11 In it: 1100 - electronic device, 1101 - processor, 1102 - communication bus, 1103 - user interface, 1104 - communication interface, 1105 - memory. DETAILED DESCRIPTION

[0052] The present invention will be further described below in conjunction with the embodiments shown in the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments of the present invention, and it should be understood that the present invention is not limited to the exemplary embodiments described here.

[0053] It should be noted that the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention unless specifically stated otherwise.

[0054] It should also be understood that any component, data or structure mentioned in the embodiments of the present invention can generally be understood as one or more, unless explicitly limited or otherwise indicated in the context.

[0055] In addition, the term "and / or" in the present disclosure is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present disclosure generally indicates that the associated objects before and after are in an "or" relationship.

[0056] It should also be understood that the description of the various embodiments of the present invention focuses on the differences between the various embodiments, and the same or similar aspects thereof can be referenced to each other, and for the sake of brevity, they will not be described one by one.

[0057] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0058] The embodiments of the present invention may not discuss in detail the technologies, methods, and devices known to ordinary technicians in the relevant fields, but where appropriate, the technologies, methods, and devices should be considered as part of the specification.

[0059] Embodiments of the present invention can be applied to electronic devices such as terminal devices, computer systems, servers, etc., which can operate with many other general or special computing system environments or configurations. Examples of well-known terminal devices, computing systems, environments and / or configurations suitable for use with electronic devices such as terminal devices, computer systems, servers, etc. include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, small computer systems, large computer systems, and distributed cloud computing technology environments including any of the above systems, etc.

[0060] Electronic devices such as terminal devices, computer systems, servers, etc. can be described in the general context of computer system executable instructions (such as program modules) executed by computer systems. Generally, program modules can include routines, programs, object programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. Computer systems / servers can be implemented in a distributed cloud computing environment, where tasks are performed by remote processing devices linked through a communication network. In a distributed cloud computing environment, program modules can be located on local or remote computing system storage media including storage devices.

[0061] As can be seen from the background technology, well logging curves are an important basis for well seismic analysis and reservoir prediction. However, in mudstone formations, diameter expansion and collapse often occur due to changes in drilling speed and other reasons, resulting in abnormal diameter expansion in acoustic wave, density and other curves, affecting synthetic record calibration, rock physics analysis and subsequent reservoir prediction work. Therefore, it is necessary to correct the abnormal values ​​of the curves affected by diameter expansion so that they can better conform to geological laws.

[0062] At present, the commonly used methods for P-wave velocity curve correction include parameter fitting method, empirical formula method, rock physics modeling, etc.

[0063] The parameter fitting method is also called the standard layer statistical method. Generally, the non-expanded section is selected as the stable section, and the target curve of the stable section is intersected and fitted with other curves to obtain the fitting relationship between the parameters. This relationship is then applied to the expanded section to correct the target curve of the expanded section. The disadvantage of this method is that the intersection fitting can often only fit simple functional relationships between parameters, while in reality, there may be more complex nonlinear relationships between the parameter curves.

[0064] The most commonly used empirical formula method for calculating P-wave velocity is the Faust formula method, which uses the empirical relationship between P-wave impedance and resistivity in mudstone sections to correct the P-wave velocity curve. The shortcoming of this method is that it only considers the influence of resistivity on P-wave velocity, resulting in low prediction accuracy.

[0065] Therefore, the present application provides a correction method that can reasonably correct the longitudinal wave velocity logging curve of the expansion section to solve the above technical problems.

[0066] Embodiment 1

[0067] In the first embodiment of the present invention, Figure 1 As shown, a method for correcting a longitudinal wave velocity logging curve is provided, and the method specifically comprises the following steps:

[0068] Step S101: construct input sample data and output sample data.

[0069] Step S102: iteratively training the deep feedforward neural network model according to the input sample data and the output sample data through a preset model training strategy to obtain a P-wave velocity logging curve correction model.

[0070] Step S103: inputting all the expanded diameter section logging curves that meet the preset conditions in the expanded diameter section logging curve set into the P-wave velocity logging curve correction model to perform P-wave velocity logging curve correction to obtain the corrected expanded diameter section P-wave velocity logging curve.

[0071] Among them, the deep feedforward neural network, also known as the multilayer perceptron (MLP), is a typical deep learning model in which neurons are arranged in layers. Each neuron is only connected to the neurons in the previous layer, receives the output of the previous layer, and outputs it to the next layer, with no feedback between layers. The deep feedforward neural network can learn the representation and features of data through multi-layer nonlinear transformations, thereby realizing various machine learning tasks.

[0072] Optionally, in step S101, constructing input sample data and output sample data specifically includes the following steps:

[0073] Step S1011: constructing a set of well logging curves of a stable section and an expanded diameter section in a target layer section of a target well according to a preset construction strategy;

[0074] Step S1012: taking all stable segment logging curves that meet preset conditions in the stable segment logging curve set as input sample data, and taking the stable segment P-wave velocity logging curves in the stable segment logging curve set as output sample data.

[0075] Specifically, the selection of the target layer section of the target well is determined by the technicians according to actual needs, and the present invention does not impose any special restrictions on this.

[0076] For example, a typical well in the study area may be selected as a target well.

[0077] Optionally, in step S1011, Figure 2 As shown, the preset construction strategy specifically includes the following steps:

[0078] Step S10111: The well section in the target layer section of the target well with a well diameter greater than a preset well diameter threshold is regarded as a stable section, and the well section in the target layer section of the target well with a well diameter not greater than the preset well diameter threshold is regarded as an expanded diameter section.

[0079] It should be noted that the preset well diameter threshold value can be determined by technicians according to actual conditions, and the present invention does not specifically limit the preset well diameter threshold value.

[0080] Step S10112: extract the stable segment logging curve corresponding to the stable segment from each logging curve of the target layer segment of the target well, so as to construct a stable segment logging curve set including all stable segment logging curves.

[0081] Step S10113: extracting the expanded diameter section logging curve corresponding to the expanded diameter section from each logging curve of the target layer segment of the target well, so as to construct an expanded diameter section logging curve set including all expanded diameter section logging curves.

[0082] Among them, the target layer section of the target well includes multiple logging curves, such as compressional wave velocity (logging symbol is VP), density (logging symbol is RHO), well diameter (logging symbol is CAL), deep lateral resistivity (logging symbol is LLD), shallow lateral resistivity (logging symbol is LLS), drilling depth (logging symbol is MD), natural potential (logging symbol is SP) and other logging curves. Each logging curve reflects the change of its corresponding parameter with depth.

[0083] After the target layer segment of the target well is divided into a stable segment and an expanded diameter segment according to step S10111, the stable segment corresponds to a depth range, and the expanded diameter segment corresponds to a depth range. Therefore, in step S10112 and step S10113, the logging curve of the corresponding depth can be intercepted from each logging curve of the target layer segment of the target well according to the depth range of the stable segment and the depth range of the expanded diameter segment, respectively, so as to obtain a logging curve set of the stable segment including multiple stable segment logging curves and a logging curve set of the expanded diameter segment including multiple expanded diameter segment logging curves.

[0084] Optionally, in step S102, the preset model training strategy includes a least squares method.

[0085] The least squares method (LS), also known as the least square method, is a mathematical tool widely used in many disciplines such as error estimation, uncertainty, system identification, prediction, and forecasting. It finds the best function matching data by minimizing the sum of squares of errors. The least squares method can be used to easily obtain unknown data and minimize the sum of squares of errors between the obtained data and the actual data.

[0086] For example, in the actual training process, it is assumed that the target P-wave velocity curve (stable P-wave velocity logging curve) to be corrected is the variable y, and all stable segment logging curves that meet the preset conditions (curves with high correlation with the stable segment P-wave velocity logging curve and less affected by the expansion) are of two types, x1 and x2. Iterative training of the deep feedforward neural network model by the least squares method includes:

[0087] Take x1 and x2 as input sample data and y as output sample data, and establish a deep feedforward neural network model with 2 hidden layers and 3 neurons in each hidden layer (the schematic diagram of the deep feedforward neural network model is shown in Figure 3 As shown), conduct learning training.

[0088] The activation function of the hidden layer neurons is the nonlinear Sigmoid logic function, that is,

[0089] The w and b of the deep feedforward neural network model are the weight coefficient and bias term of the network model respectively. The neurons in each layer of the network are connected and transmitted to the neurons in the next layer through their respective weight coefficients w and bias terms b. These two parameters need to be initialized before training.

[0090] x1 and x2 are input into the deep feedforward neural network model, and the weighted calculation is first performed as follows (1) to obtain z1, z2, and z3:

[0091]

[0092] Pass z1, z2, and z3 to the first hidden layer neurons for nonlinear Sigmoid logic function conversion to obtain f1(z1) and f2(z2).

[0093] The result after nonlinear transformation is passed to the neurons of the next layer through weighted summation calculation, and so on for nonlinear operations between parameters.

[0094] During the training process, the conjugate gradient method is used to optimize the objective function (Formula 2) according to the principle that the variance of each layer is as equal as possible. The calculation is terminated when the output result and the target curve y reach the least square error. At this time, the obtained w and b are the optimal solutions, and the model corresponding to the optimal solution is the P-wave velocity logging curve correction model obtained through training.

[0095] Among them, the conjugate gradient method optimization objective function includes:

[0096]

[0097] In formula (2), is the L2 norm of w, λ is the weight coefficient of the norm, P is the number of sample points of the logging curve in the input sample data, L is the total number of layers of the deep neural network model, l is the l-th layer network; m is the number of neurons in the l-th layer; i is the i-th neuron in the l-th layer, j is the j-th neuron in the l-1-th layer, and f() is the nonlinear Sigmoid logic function.

[0098] In addition, during the model training process, the number of hidden layers or the number of different neurons of the deep feedforward neural network model can be adjusted according to the error variation law between the output result and the target curve y.

[0099] Furthermore, in one implementation, determining a stable segment logging curve that meets a preset condition includes the following specific steps:

[0100] Step S10121: Delete the stable segment P-wave velocity logging curve from the stable segment logging curve set to obtain a target curve set.

[0101] Step S10122: performing correlation analysis on each stable segment logging curve in the target curve set and the stable segment P-wave velocity logging curve to obtain correlation analysis results.

[0102] Step S10123: Determine the stable segment logging curve that meets the preset conditions according to the correlation analysis results.

[0103] Specifically, in step S10122, performing correlation analysis includes:

[0104] The correlation between each stable section logging curve in the target curve set and the stable section P-wave velocity logging curve is analyzed, and the influence of the diameter expansion on each stable section logging curve in the target curve set is analyzed.

[0105] In step S10123, determining the stable segment logging curve that meets the preset conditions according to the correlation analysis results includes:

[0106] Several types of logging curves with high correlation with P-wave velocity and small influence of diameter expansion are determined as stable segment logging curves meeting preset conditions, and then all stable segment logging curves meeting preset conditions are determined from the stable segment logging curve set.

[0107] It is understandable that in step S103, all the expansion section logging curves meeting the preset conditions are of the same type as all the stable section logging curves meeting the preset conditions, that is, all the expansion section logging curves meeting the preset conditions need to be determined based on all the stable section logging curves meeting the preset conditions.

[0108] For example, when all stable section logging curves that meet the preset conditions include deep lateral resistivity (LLD) logging curves, shallow lateral resistivity (LLS) logging curves, drilling depth logging (MD) logging curves, natural potential (SP) logging curves, and natural gamma (GR) logging curves, then all expanded section logging curves that meet the preset conditions also include deep lateral resistivity (LLD) logging curves, shallow lateral resistivity (LLS) logging curves, drilling depth logging (MD) logging curves, natural potential (SP) logging curves, and natural gamma (GR) logging curves.

[0109] The P-wave velocity logging curve correction method provided in this embodiment is to iteratively train the deep feedforward neural network model according to the input sample data and the output sample data through the preset model training strategy to obtain the P-wave velocity logging curve correction model. When correcting the P-wave velocity logging curve of the expansion section, all the expansion section logging curves that meet the preset conditions in the expansion section logging curve set are used as model input to obtain the corrected P-wave velocity logging curve of the expansion section. By iteratively training the deep feedforward neural network model, the nonlinear relationship between the input sample data and the output sample data can be fully explored, and then this relationship is applied to the expansion section, and the P-wave velocity logging curve is corrected using the trained model. Compared with the conventional parameter fitting method and the empirical formula method, the present invention takes into account the influence of multiple factors, has higher learning accuracy, and the obtained curve correction result is more reliable, which improves the rationality of the curve correction, thereby laying a good foundation for subsequent logging response characteristics and rock physics analysis, synthetic record calibration, reservoir prediction and other work.

[0110] Embodiment 2

[0111] In the second embodiment of the present invention, Figure 4 As shown, a longitudinal wave velocity logging curve correction device is provided, and the device specifically comprises:

[0112] Construction module 401, used to construct input sample data and output sample data;

[0113] The training module 402 is used to iteratively train the deep feedforward neural network model according to the input sample data and the output sample data through a preset model training strategy to obtain a P-wave velocity logging curve correction model;

[0114] The correction module 403 is used to input all the expanded diameter section logging curves that meet the preset conditions in the expanded diameter section logging curve set into the P-wave velocity logging curve correction model to perform P-wave velocity logging curve correction and obtain the corrected expanded diameter section P-wave velocity logging curve.

[0115] Alternatively, if Figure 5 As shown, building block 401 includes:

[0116] A set construction unit 4011 is used to respectively construct a set of logging curves of a stable section and an expanded diameter section in a target layer section of a target well according to a preset construction strategy;

[0117] The sample data determination unit 4012 is used to use all stable segment logging curves that meet preset conditions in the stable segment logging curve set as input sample data, and use the stable segment P-wave velocity logging curves in the stable segment logging curve set as output sample data.

[0118] Optionally, the preset model training strategy includes a least squares method.

[0119] Alternatively, if Figure 6 As shown, the building block 401 also includes:

[0120] The curve determination unit 4013 is used to delete the stable segment P-wave velocity logging curve from the stable segment logging curve set to obtain a target curve set; and perform correlation analysis on each stable segment logging curve in the target curve set with the stable segment P-wave velocity logging curve to obtain a correlation analysis result; and to determine the stable segment logging curve that meets the preset conditions according to the correlation analysis result.

[0121] The P-wave velocity logging curve correction device provided in this embodiment iteratively trains the deep feedforward neural network model according to the input sample data and the output sample data through the preset model training strategy to obtain the P-wave velocity logging curve correction model. When correcting the P-wave velocity logging curve of the expansion section, all the expansion section logging curves that meet the preset conditions in the logging curve set of the expansion section are used as model input to obtain the corrected P-wave velocity logging curve of the expansion section. By iteratively training the deep feedforward neural network model, the nonlinear relationship between the input sample data and the output sample data can be fully explored, and then this relationship is applied to the expansion section, and the P-wave velocity logging curve is corrected using the trained model. Compared with the conventional parameter fitting method and the empirical formula method, the present invention takes into account the influence of multiple factors, has higher learning accuracy, and the obtained curve correction result is more reliable, which improves the rationality of the curve correction, thereby laying a good foundation for subsequent logging response characteristics and rock physics analysis, synthetic record calibration, reservoir prediction and other work.

[0122] Embodiment 3

[0123] In the third embodiment of the present invention, a specific example of correcting a P-wave velocity logging curve using the P-wave velocity logging curve correction method in the first embodiment is provided.

[0124] In this embodiment, a typical well S1 in a block in the western part of the Ordos Basin is selected as the target well. Figure 7 As shown, there is an expansion anomaly in the P-wave velocity of the target layer, the Shanxi Formation of the Permian System. The correction method of an embodiment of the present invention is used to correct the P-wave velocity logging curve of the expansion section.

[0125] Through analysis, it is determined that the target layer with a well diameter less than 23 cm is a stable section, and the well diameter greater than 23 cm is an expansion section that needs to be corrected. The stable section is selected to construct the pseudo well W1 (the logging curve set of the pseudo well W1 is the logging curve set of the stable section).

[0126] The correlation between the P-wave velocity logging curve of Well W1 and other parameter logging curves was analyzed. It was found that there was a good correlation between P-wave velocity and five parameters, namely, deep lateral resistivity (LLD), shallow lateral resistivity (LLS), drilling sounding (MD), spontaneous potential (SP) and natural gamma (GR), and these five parameters were basically not affected by diameter expansion. Therefore, the deep lateral resistivity (LLD), shallow lateral resistivity (LLS), drilling sounding (MD), spontaneous potential (SP) and natural gamma (GR) logging curves in the pseudo-well W1 were used as input data of the deep feedforward neural network model, and the P-wave velocity logging curve in the pseudo-well W1 was used as output data to construct a sample set.

[0127] Through the model parameter test, a deep feedforward neural network model with 3 hidden layers, 50 neurons in each layer, and 500 iterations was finally established. The model training accuracy reached 0.926 (such as Figure 8 ), which shows the validity and rationality of the model.

[0128] The five parameter logging curves of deep lateral resistivity (LLD), shallow lateral resistivity (LLS), drilling depth (MD), spontaneous potential (SP), and natural gamma (GR) of the target layer expansion section of S1 well were input into the constructed model, and the corrected P-wave velocity logging curve was obtained as follows: Fig. 9 As shown, the corrected P-wave velocity logging curve is used to regenerate a new synthetic record and calibrate it with the well side track of the seismic data. Fig.10 As shown in the figure, through the calibration comparison of the synthetic records before and after correction, it can be seen that the wave group relationship of the synthetic records after correction is more reasonable, which lays a good foundation for subsequent rock physical analysis and reservoir prediction.

[0129] Embodiment 4

[0130] In the fourth embodiment of the present invention, a computer program product is also provided. The computer program product includes a computer program or instructions. When the computer program or instructions are executed by a processor, all or part of the steps of the method for correcting the longitudinal wave velocity logging curve described in the above embodiment are implemented.

[0131] Among them, the P-wave velocity logging curve correction method includes:

[0132] Construct input sample data and output sample data;

[0133] Iteratively train the deep feedforward neural network model according to the input sample data and the output sample data through a preset model training strategy to obtain a P-wave velocity logging curve correction model;

[0134] All the expanded diameter section logging curves that meet the preset conditions in the expanded diameter section logging curve set are input into the P-wave velocity logging curve correction model to perform P-wave velocity logging curve correction, and obtain the corrected expanded diameter section P-wave velocity logging curve.

[0135] Optionally, construct input sample data and output sample data, including:

[0136] According to the preset construction strategy, the logging curve sets of the stable section and the expanded diameter section in the target layer section of the target well are respectively constructed;

[0137] All stable segment logging curves that meet preset conditions in the stable segment logging curve set are used as input sample data, and the stable segment P-wave velocity logging curves in the stable segment logging curve set are used as output sample data.

[0138] Optionally, the preset build strategies include:

[0139] The well section in the target layer section of the target well with a well diameter greater than a preset well diameter threshold is regarded as a stable section, and the well section in the target layer section of the target well with a well diameter not greater than the preset well diameter threshold is regarded as an expanded diameter section;

[0140] Extracting the stable segment logging curve corresponding to the stable segment from each logging curve of the target layer segment of the target well respectively, so as to construct a logging curve set of the stable segment including all the stable segment logging curves;

[0141] The expanded diameter section logging curve corresponding to the expanded diameter section is respectively intercepted from each logging curve of the target layer section of the target well to construct an expanded diameter section logging curve set including all expanded diameter section logging curves.

[0142] Optionally, determining a stable segment logging curve that meets a preset condition includes:

[0143] Deleting the stable segment P-wave velocity logging curve from the stable segment logging curve set to obtain a target curve set;

[0144] Perform correlation analysis on each stable segment logging curve in the target curve set and the stable segment P-wave velocity logging curve to obtain correlation analysis results;

[0145] The stable section logging curve that meets the preset conditions is determined based on the correlation analysis results.

[0146] Optionally, the preset model training strategy includes a least squares method.

[0147] Further, the computer program product may include one or more computer executable components configured to execute the embodiments when the program is run; the computer program product may also include a computer program tangibly contained on a readable medium, the computer program including program code for executing any method in the embodiments of the present invention. In such an embodiment, the computer program may be downloaded and installed from a network through a communication part, and / or installed from a removable medium.

[0148] Embodiment 5

[0149] In the fifth embodiment of the present invention, a computer-readable storage medium is further provided. The computer program stored in the computer-readable storage medium, when executed by one or more processors, implements all or part of the steps of the method for correcting the longitudinal wave velocity logging curve described in the above embodiment.

[0150] Among them, the P-wave velocity logging curve correction method includes:

[0151] Construct input sample data and output sample data;

[0152] Iteratively train the deep feedforward neural network model according to the input sample data and the output sample data through a preset model training strategy to obtain a P-wave velocity logging curve correction model;

[0153] All the expanded diameter section logging curves that meet the preset conditions in the expanded diameter section logging curve set are input into the P-wave velocity logging curve correction model to perform P-wave velocity logging curve correction, and obtain the corrected expanded diameter section P-wave velocity logging curve.

[0154] Optionally, construct input sample data and output sample data, including:

[0155] According to the preset construction strategy, the logging curve sets of the stable section and the expanded diameter section in the target layer section of the target well are respectively constructed;

[0156] All stable segment logging curves that meet preset conditions in the stable segment logging curve set are used as input sample data, and the stable segment P-wave velocity logging curves in the stable segment logging curve set are used as output sample data.

[0157] Optionally, the preset build strategies include:

[0158] The well section in the target layer section of the target well with a well diameter greater than a preset well diameter threshold is regarded as a stable section, and the well section in the target layer section of the target well with a well diameter not greater than the preset well diameter threshold is regarded as an expanded diameter section;

[0159] Extracting the stable segment logging curve corresponding to the stable segment from each logging curve of the target layer segment of the target well respectively, so as to construct a logging curve set of the stable segment including all the stable segment logging curves;

[0160] The expanded diameter section logging curve corresponding to the expanded diameter section is respectively intercepted from each logging curve of the target layer section of the target well to construct an expanded diameter section logging curve set including all expanded diameter section logging curves.

[0161] Optionally, determining a stable segment logging curve that meets a preset condition includes:

[0162] Deleting the stable segment P-wave velocity logging curve from the stable segment logging curve set to obtain a target curve set;

[0163] Perform correlation analysis on each stable segment logging curve in the target curve set and the stable segment P-wave velocity logging curve to obtain correlation analysis results;

[0164] The stable section logging curve that meets the preset conditions is determined based on the correlation analysis results.

[0165] Optionally, the preset model training strategy includes a least squares method.

[0166] Each functional unit in each embodiment of the present invention may be integrated into a processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. If the functions are implemented in the form of software functional units and sold or used as independent products, they may be stored in a computer-readable storage medium.

[0167] Computer readable storage media can use any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any combination of the above. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (Random Access Memory, referred to as RAM), a read-only memory (Read-Only Memory, referred to as ROM), an erasable programmable read-only memory (Electrically Erasable Programmable read only memory, referred to as EPROM), an optical fiber, a portable compact disc read-only memory (Compact Disc Read-Only Memory, referred to as CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0168] Embodiment 6

[0169] In the sixth embodiment of the present invention, an electronic device 1100 is also provided, which may be a mobile phone, a computer, or a tablet computer. Fig.11 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention is shown in FIG. Fig.11 As shown, the electronic device 1100 includes: at least one processor 1101, at least one communication bus 1102, a user interface 1103, at least one external communication interface 1104, and a memory 1105. The communication bus 1102 is configured to realize the connection and communication between these components. The user interface 1103 may include a display screen, and the external communication interface 1104 may include a standard wired interface and a wireless interface. The memory 1105 stores a computer program, and the memory 1105 and one or more processors 1101 are connected to each other in communication. When the computer program is executed by one or more processors, the processor 1101 is configured to execute the computer program stored in the memory to implement all or part of the steps of the longitudinal wave velocity logging curve correction method in the above embodiment.

[0170] Among them, the P-wave velocity logging curve correction method includes:

[0171] Construct input sample data and output sample data;

[0172] Iteratively train the deep feedforward neural network model according to the input sample data and the output sample data through a preset model training strategy to obtain a P-wave velocity logging curve correction model;

[0173] All the expanded diameter section logging curves that meet the preset conditions in the expanded diameter section logging curve set are input into the P-wave velocity logging curve correction model to perform P-wave velocity logging curve correction, and obtain the corrected expanded diameter section P-wave velocity logging curve.

[0174] Optionally, construct input sample data and output sample data, including:

[0175] According to the preset construction strategy, the logging curve sets of the stable section and the expanded diameter section in the target layer section of the target well are respectively constructed;

[0176] All stable segment logging curves that meet preset conditions in the stable segment logging curve set are used as input sample data, and the stable segment P-wave velocity logging curves in the stable segment logging curve set are used as output sample data.

[0177] Optionally, the preset build strategies include:

[0178] The well section in the target layer section of the target well with a well diameter greater than a preset well diameter threshold is regarded as a stable section, and the well section in the target layer section of the target well with a well diameter not greater than the preset well diameter threshold is regarded as an expanded diameter section;

[0179] Extracting the stable segment logging curve corresponding to the stable segment from each logging curve of the target layer segment of the target well respectively, so as to construct a logging curve set of the stable segment including all the stable segment logging curves;

[0180] The expanded diameter section logging curve corresponding to the expanded diameter section is respectively intercepted from each logging curve of the target layer section of the target well to construct an expanded diameter section logging curve set including all expanded diameter section logging curves.

[0181] Optionally, determining a stable segment logging curve that meets a preset condition includes:

[0182] Deleting the stable segment P-wave velocity logging curve from the stable segment logging curve set to obtain a target curve set;

[0183] Perform correlation analysis on each stable segment logging curve in the target curve set and the stable segment P-wave velocity logging curve to obtain correlation analysis results;

[0184] The stable section logging curve that meets the preset conditions is determined based on the correlation analysis results.

[0185] Optionally, the preset model training strategy includes a least squares method.

[0186] The processor can be an application specific integrated circuit (ASIC), a digital signal processor (DSP), a programmable logic device (PLD), a field programmable gate array (FPGA), a controller, a microcontroller, a microprocessor or other electronic components to execute all or part of the steps of the longitudinal wave velocity logging curve correction method described in the above embodiment, which will not be repeated in this embodiment.

[0187] The memory can be implemented by any type of volatile or non-volatile storage device or a combination of them, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0188] The P-wave velocity logging curve correction method, device and electronic device provided by the present invention iteratively train the deep feedforward neural network model according to the input sample data and the output sample data through the preset model training strategy to obtain the P-wave velocity logging curve correction model. When correcting the P-wave velocity logging curve of the expansion section, all the expansion section logging curves that meet the preset conditions in the logging curve set of the expansion section are used as model input to obtain the corrected P-wave velocity logging curve of the expansion section. By iteratively training the deep feedforward neural network model, the nonlinear relationship between the input sample data and the output sample data can be fully explored, and then this relationship is applied to the expansion section, and the P-wave velocity logging curve is corrected using the trained model. Compared with the conventional parameter fitting method and the empirical formula method, the present invention takes into account the influence of multiple factors, has higher learning accuracy, and the obtained curve correction result is more reliable, which improves the rationality of the curve correction, thereby laying a good foundation for subsequent logging response characteristics and rock physics analysis, synthetic record calibration, reservoir prediction and other work.

[0189] The basic principles of the present invention are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and facilitation of understanding, and are not limiting. The above details do not limit the present invention to being implemented by adopting the above specific details.

[0190] The block diagrams of the devices, apparatuses, equipment, and systems involved in the present invention are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems may be connected, arranged, and configured in any manner. Words such as "including," "comprising," "having," and the like are open-ended words, meaning "including but not limited to," and may be used interchangeably therewith.

[0191] It should also be noted that in the apparatus, device and method disclosed herein, each component or each step can be decomposed and / or reassembled, and such decomposition and / or reassembly should be regarded as an equivalent solution of the present invention.

[0192] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present invention. Therefore, the present invention is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0193] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present invention to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A method for correcting a longitudinal wave velocity logging curve, characterized in that: The method comprises: Construct input sample data and output sample data; Iteratively training a deep feedforward neural network model according to the input sample data and the output sample data through a preset model training strategy to obtain a P-wave velocity logging curve correction model; All the expanded diameter section logging curves that meet the preset conditions in the expanded diameter section logging curve set are input into the P-wave velocity logging curve correction model to perform P-wave velocity logging curve correction to obtain the corrected expanded diameter section P-wave velocity logging curve.

2. The method for correcting the longitudinal wave velocity logging curve according to claim 1, characterized in that: The constructing of input sample data and output sample data includes: According to the preset construction strategy, the logging curve sets of the stable section and the expanded diameter section in the target layer section of the target well are respectively constructed; All stable segment logging curves that meet the preset conditions in the stable segment logging curve set are used as input sample data, and the stable segment P-wave velocity logging curves in the stable segment logging curve set are used as output sample data.

3. The method for correcting the longitudinal wave velocity logging curve according to claim 2, characterized in that: The preset construction strategy includes: The well section in the target layer section of the target well with a well diameter greater than a preset well diameter threshold is regarded as a stable section, and the well section in the target layer section of the target well with a well diameter not greater than the preset well diameter threshold is regarded as an expanded diameter section; Extracting the stable segment logging curve corresponding to the stable segment from each logging curve of the target layer segment of the target well respectively, so as to construct a logging curve set of the stable segment including all the stable segment logging curves; The expanded diameter section logging curve corresponding to the expanded diameter section is respectively intercepted from each logging curve of the target layer section of the target well to construct a logging curve set of the expanded diameter section including all the expanded diameter section logging curves.

4. The method for correcting the longitudinal wave velocity logging curve according to claim 2, characterized in that: Determining a stable segment logging curve that meets the preset conditions includes: Deleting the stable segment P-wave velocity logging curve from the stable segment logging curve set to obtain a target curve set; Performing correlation analysis on each stable segment logging curve in the target curve set and the stable segment P-wave velocity logging curve to obtain a correlation analysis result; A stable segment logging curve that meets the preset conditions is determined according to the correlation analysis result.

5. The method for correcting the longitudinal wave velocity logging curve according to claim 1, characterized in that: The preset model training strategy includes the least squares method.

6. A longitudinal wave velocity logging curve correction device, characterized in that: The device comprises: A construction module, used for constructing input sample data and output sample data; A training module, used for iteratively training a deep feedforward neural network model according to the input sample data and the output sample data through a preset model training strategy to obtain a P-wave velocity logging curve correction model; The correction module is used to input all the expanded diameter section logging curves that meet the preset conditions in the expanded diameter section logging curve set into the P-wave velocity logging curve correction model to perform P-wave velocity logging curve correction and obtain the corrected expanded diameter section P-wave velocity logging curve.

7. The longitudinal wave velocity logging curve correction device according to claim 6, characterized in that: The building blocks include: A set construction unit is used to respectively construct a set of logging curves of a stable section and an expanded diameter section in a target layer section of a target well according to a preset construction strategy; The sample data determination unit is used to use all stable segment logging curves that meet preset conditions in the stable segment logging curve set as input sample data and use the stable segment P-wave velocity logging curve in the stable segment logging curve set as output sample data.

8. The longitudinal wave velocity logging curve correction device according to claim 7, characterized in that: The building blocks also include: A curve determination unit is used to delete the stable segment P-wave velocity logging curve from the stable segment logging curve set to obtain a target curve set; and to perform correlation analysis on each stable segment logging curve in the target curve set with the stable segment P-wave velocity logging curve to obtain a correlation analysis result; and to determine the stable segment logging curve that meets the preset condition according to the correlation analysis result.

9. A computer-readable storage medium, characterized in that: The computer program stored in the computer-readable storage medium, when executed by one or more processors, implements the steps of the method according to any one of claims 1 to 5.

10. An electronic device, characterized in that: The method comprises a memory and one or more processors, wherein a computer program is stored in the memory, and when the computer program is executed by the one or more processors, the steps of the method according to any one of claims 1 to 5 are performed.

Images