Method and device for predicting the time difference of shear waves

Abstract

A method and apparatus for predicting the time difference of shear waves, which relate to the field of oil exploration and production engineering. The method includes: obtaining sample well logging data as a training data set for a prediction model (101); pre-processing the training data set, performing data selection based on an importance analysis, and grouping the data based on kurtosis and asymmetry of the distribution to obtain a processed training data set (102); respectively inputting the processed training data set into a neural network formed by a combination of a CNN and an LSTM for training, in order to obtain a prediction model of the time difference of shear waves (103); obtaining well logging data for predicting the time difference of shear waves (104);pre-processing the well logging data and grouping the well logging data based on the kurtosis and skewness of the distribution to obtain processed well logging data (105); and accordingly using the processed well logging data as input data for a shear wave time difference prediction model to obtain a shear wave time difference (106). The proposed method and apparatus provide advantages of high data processing efficiency, high prediction accuracy and good regional applicability.

Description

Shear wave time difference prediction method and device Technical Field

[0001] The present invention relates to the technical field of petroleum exploration and development, and in particular to a shear wave time difference prediction method, a shear wave time difference prediction device, an electronic device and a machine-readable storage medium. Background Art

[0002] Shear wave logging data is a key parameter used in rock physics analysis, lithology identification, calculation of rock elastic parameters, reservoir description, and fluid identification. It plays a vital role in improving reservoir prediction accuracy. Conventional sonic logging can produce longitudinal and shear wave logging data, but the resulting shear waves are of poor quality or missing, insufficient to meet production needs. Dipole sonic logging instruments can produce better-quality shear wave data, but the acquisition cost is high and is limited to key wells or risky exploration wells. Most wells lack shear wave logging data. Well conditions, logging technology, and cost are the main reasons for this missing data, making accurate shear wave prediction particularly important.

[0003] Common methods for predicting shear waves include the empirical formula method and the rock physics model method. The empirical formula method calculates shear waves by analyzing the relationship between longitudinal waves and shear waves to obtain a fitted linear formula. This method is simple and convenient and can quickly predict shear waves, but the shear wave prediction accuracy of the empirical formula method is not high and there is a problem of poor regional applicability. The rock physics model method calculates shear waves by constructing a rock skeleton model and a fluid parameter model. This method can accurately predict shear waves, but the model requires more accurate parameters, such as rock mineral composition, porosity, pore structure, etc. Many parameters are difficult to collect, and it is not easy to establish an accurate rock physics model and the calculation efficiency is low. In summary, both the empirical formula method and the rock physics model method have certain limitations. Therefore, this application proposes a prediction method based on machine learning.

[0004] Summary of the Invention

[0005] The purpose of the embodiments of the present invention is to provide a shear wave time difference prediction method and device, which are used to solve the problems of low prediction accuracy, poor regional applicability and low calculation efficiency of the above-mentioned methods.

[0006] To achieve the above objectives, an embodiment of the present invention provides a shear wave time difference prediction method, comprising:

[0007] Obtaining well logging sample data as a training data set for the prediction model;

[0008] Preprocessing the training data set, screening the data based on importance analysis, and grouping the data based on kurtosis and skewness to obtain a processed training data set;

[0009] The processed training data sets are input into the neural network constructed by hybrid CNN and LSTM for training to obtain the shear wave time difference prediction model;

[0010] Obtaining well logging data for the shear wave time difference to be predicted;

[0011] Preprocessing the well logging data, grouping the well logging data based on kurtosis and skewness, and obtaining processed well logging data;

[0012] The processed logging data are respectively used as inputs of the shear wave time difference prediction model to obtain the shear wave time difference;

[0013] Preprocessing the training data set, screening the data based on importance analysis, and grouping the data based on kurtosis and skewness to obtain a processed training data set includes:

[0014] Performing data cleaning, data filtering, and data normalization on the training data set to obtain first training data;

[0015] Filtering data whose correlation coefficient with the shear wave time difference is greater than a first preset coefficient value from the first training data as second training data;

[0016] respectively calculating correlation coefficients between two different types of data in the second training data;

[0017] If there are two different types of data with a correlation coefficient greater than the second preset coefficient value, one type of data is selected from the two different types of data and used as the third training data with the other different types of data with a correlation coefficient less than or equal to the second preset coefficient value in the second training data;

[0018] Based on a preset kurtosis coefficient and a preset skewness coefficient, the third training data is divided into at least two groups of well logging data as the processed training data sets.

[0019] Optionally, the CNN neural network and the LSTM neural network in the shear wave time difference prediction model are connected through a Dropout layer.

[0020] Optionally, the logging data includes: natural gamma logging data, caliper logging data, natural potential logging data, resistivity logging data, neutron logging data, sonic logging data and density logging data.

[0021] Optionally, the correlation coefficient is calculated using the Pearson correlation coefficient calculation formula.

[0022] An embodiment of the present invention further provides a shear wave time difference prediction device, comprising:

[0023] A training data acquisition module is used to obtain well logging sample data as a training data set for the prediction model;

[0024] a first data processing module, configured to preprocess the training data set, perform data screening based on importance analysis, and group the data based on kurtosis and skewness to obtain a processed training data set;

[0025] The model training module is used to input the processed training data set into the neural network constructed by CNN and LSTM respectively for training to obtain the shear wave time difference prediction model;

[0026] An input data acquisition module is used to acquire the logging data of the shear wave time difference to be predicted;

[0027] a second data processing module, configured to pre-process the well logging data, group the well logging data based on kurtosis and skewness, and obtain processed well logging data;

[0028] A result output module is used to use the processed logging data as input of the shear wave time difference prediction model to obtain the shear wave time difference;

[0029] The first data processing module is specifically configured to:

[0030] Preprocessing the training data set, screening the data based on importance analysis, and grouping the data based on kurtosis and skewness to obtain a processed training data set includes:

[0031] Performing data cleaning, data filtering, and data normalization on the training data set to obtain first training data;

[0032] Filtering data whose correlation coefficient with the shear wave time difference is greater than a first preset coefficient value from the first training data as second training data;

[0033] respectively calculating correlation coefficients between two different types of data in the second training data;

[0034] If there are two different types of data with a correlation coefficient greater than the second preset coefficient value, one type of data is selected from the two different types of data and used as the third training data with the other different types of data with a correlation coefficient less than or equal to the second preset coefficient value in the second training data;

[0035] Based on a preset kurtosis coefficient and a preset skewness coefficient, the third training data is divided into at least two groups of well logging data as the processed training data sets.

[0036] Optionally, the CNN neural network and the LSTM neural network in the shear wave time difference prediction model are connected through a Dropout layer.

[0037] Optionally, the logging data includes: natural gamma logging data, caliper logging data, natural potential logging data, resistivity logging data, neutron logging data, sonic logging data and density logging data.

[0038] Optionally, the correlation coefficient is calculated using the Pearson correlation coefficient calculation formula.

[0039] An embodiment of the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned shear wave time difference prediction method when executing the computer program.

[0040] On the other hand, the present invention provides a machine-readable storage medium having stored thereon instructions for enabling a machine to execute the above-mentioned shear wave time difference prediction method.

[0041] This technical solution combines a neural network built by hybrid CNN and LSTM to construct a shear wave time difference prediction model. The logging data for the predicted shear wave time difference is preprocessed and the data after logging data is grouped based on kurtosis and skewness is input into the shear wave time difference prediction model to obtain the shear wave time difference. The calculation is simple and practical. It can accurately predict the shear wave time difference and provide the necessary parameters for rock physics analysis, lithology identification, rock elastic mechanical parameter calculation, reservoir description and fluid identification.

[0042] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following detailed description, they are used to explain the embodiments of the present invention, but do not constitute a limitation of the embodiments of the present invention. In the accompanying drawings:

[0044] FIG1 is a schematic flow chart of a shear wave time difference prediction method provided by the present invention;

[0045] FIG2 is a schematic diagram of the structure of the shear wave time difference prediction model provided by the present invention;

[0046] FIG3 is a schematic diagram of positions of different kurtosis provided by the present invention;

[0047] FIG4 is a schematic diagram of positions of different skewnesses provided by the present invention;

[0048] FIG5 is a schematic structural diagram of a shear wave time difference prediction device provided by the present invention;

[0049] FIG6 is a schematic diagram showing a comparison between the shear wave time difference obtained by the present invention and the shear wave time difference in the prior art.

[0050] Description of Reference Numerals

[0051] 10-training data acquisition module; 20-first data processing module;

[0052] 30-model training module; 40-input parameter acquisition module;

[0053] 50-second data processing module; 60-result output module. DETAILED DESCRIPTION

[0054] The following describes the specific implementation of the embodiment of the present invention in detail with reference to the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the embodiment of the present invention and is not used to limit the embodiment of the present invention.

[0055] In the embodiments of the present invention, unless otherwise specified, directional words such as "up, down, left, right" generally refer to the orientation or position relationship shown in the accompanying drawings, or the orientation or position relationship in which the product of the invention is usually placed when in use.

[0056] The terms "first", "second", "third", etc. are only used for distinction and description and should not be understood as indicating or implying relative importance.

[0057] In addition, terms such as "roughly" and "basically" are intended to illustrate that the relevant content does not require absolute precision, but rather can have certain deviations. For example, "roughly equal" does not only mean absolute equality. Since it is difficult to achieve absolute "equality" in actual production and operation, there are generally certain deviations. Therefore, in addition to absolute equality, "roughly equal" also includes the above-mentioned situation where there are certain deviations. Taking this as an example, in other cases, unless otherwise specified, terms such as "roughly" and "basic" have similar meanings to the above. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0058] Figure 1 is a flow chart of the shear wave time difference prediction method provided by the present invention; Figure 2 is a structural diagram of the shear wave time difference prediction model provided by the present invention; Figure 3 is a position diagram of different kurtosis provided by the present invention; Figure 4 is a position diagram of different skewness provided by the present invention; Figure 5 is a structural diagram of the shear wave time difference prediction device provided by the present invention; Figure 6 is a comparative diagram of the shear wave time difference obtained by the present invention through the scheme provided by the present invention and the shear wave time difference in the prior art.

[0059] As shown in FIG1 , this embodiment provides a shear wave time difference prediction method, including:

[0060] Step 101: Acquire well logging sample data as a training data set for a prediction model;

[0061] Step 102: preprocess the training data set, perform data screening based on importance analysis, and group the data based on kurtosis and skewness to obtain a processed training data set;

[0062] Step 103: Input the processed training data set into a neural network constructed by hybrid CNN and LSTM for training to obtain a shear wave time difference prediction model;

[0063] Step 104: obtaining the well logging data of the shear wave time difference to be predicted;

[0064] Step 105: pre-process the well logging data, group the well logging data based on kurtosis and skewness, and obtain processed well logging data;

[0065] Step 106: Use the processed logging data as inputs of the shear wave time difference prediction model to obtain the shear wave time difference.

[0066] Specifically, in step 101, the well logging sample data needs to be processed, including preprocessing, data screening based on importance analysis, and data grouping based on kurtosis and skewness, to ensure that the data format remains uniform, facilitate machine learning, achieve accurate identification of shear wave time difference prediction model data, and achieve precise prediction of shear wave time difference. In step 105, the well logging data is preprocessed and grouped based on kurtosis and skewness to obtain at least two sets of processed well logging data. The at least two sets of processed well logging data are used as input to the shear wave time difference prediction model, thereby obtaining more accurate shear wave time difference, reducing the amount of calculation in the prediction process, and improving efficiency. Furthermore, in this embodiment, the method steps of preprocessing the well logging data for predicting shear wave time difference and grouping the well logging data based on kurtosis and skewness are similar to the method steps of preprocessing the training data set and grouping the data based on kurtosis and skewness, and will not be repeated here.

[0067] Furthermore, the CNN neural network and the LSTM neural network in the shear wave time difference prediction model are connected through a Dropout layer.

[0068] Specifically, as shown in FIG2 , in this embodiment, the shear wave time difference prediction model is obtained by training a neural network constructed by mixing CNN and LSTM, wherein the CNN neural network and the LSTM neural network are connected via a Dropout layer, a structure used to reduce overfitting of the neural network. CNN neural network, also known as Convolutional Neural Networks (CNN), is a type of neural network, a feedforward neural network. Its weight-sharing network structure makes it more similar to a biological neural network, reducing the complexity of the network model and the number of weights. The CNN model structure includes three layers: convolution, pooling, and full connection. Its artificial neurons can respond to surrounding units within a portion of the coverage area, so it can consider the local characteristics of the data. The convolution layer convolves the input data with the aim of reducing the number of parameters and connections, thereby greatly reducing the number of model iterations and iteration time; the pooling layer, also known as the downsampling layer, is a commonly used component of convolutional neural networks. Its main function is to reduce the dimensionality of the data, remove redundant information, compress features, and simplify the network complexity to facilitate neural network learning; the fully connected layer usually appears in the last few layers and is used to perform weighted summation on the previously designed features. Its function is to map the distributed local features extracted by the previous convolution to the sample label space.

[0069] LSTM neural networks, also known as long short-term memory neural networks, are a type of recurrent neural network designed to address the long-term dependency issues common in general neural networks. They are suitable for processing and predicting important events in time series with very long intervals and delays. LSTM primarily consists of a cell state, a forget gate, an input gate, and an output gate. The cell state circulates the information stored in each cell. The forget gate is used to decide whether to delete certain information, primarily processing information transmitted from the previous time period and information input at the current time period. The input gate detects the presence of input and determines whether to enter the data into the cell state memory. The output gate outputs the result based on the cell state, which contains information from both the current and previous time periods.

[0070] Furthermore, the logging data includes: natural gamma logging data, caliper logging data, natural potential logging data, resistivity logging data, neutron logging data, sonic logging data and density logging data.

[0071] Furthermore, the training data set is preprocessed, data is screened based on importance analysis, and data is grouped based on kurtosis and skewness to obtain a processed training data set, including:

[0072] The training data set is cleaned, filtered, and normalized to obtain first training data.

[0073] Specifically, the training data set includes historical logging sample data. Data cleaning is the removal of outliers in the logging curve. Outliers may be caused by the logging environment or human error. Logging environment reasons include wellbore expansion or large well deviation, special reservoirs, instrument performance constraints, and instrument failures. These outliers will seriously affect the training of the neural network model. Conventional processing methods include deleting and replacing abnormal data. Data filtering is the smoothing of data to remove noise and mutation data in the data. Normalization processing is the current value minus the minimum value divided by the difference between the maximum and minimum values. Its purpose is to limit the data to a certain range, eliminate the adverse effects caused by singular sample data, and improve the convergence speed and accuracy of the model.

[0074] Furthermore, the training data set is preprocessed, data is screened based on importance analysis, and data is grouped based on kurtosis and skewness to obtain a processed training data set, further comprising:

[0075] Filtering data whose correlation coefficient with the shear wave time difference is greater than a first preset coefficient value from the first training data as second training data;

[0076] respectively calculating correlation coefficients between two different types of data in the second training data;

[0077] If there are two different types of data whose correlation coefficient is greater than the second preset coefficient value, any one type of data is filtered out from the two different types of data, and the different type of data with the remaining correlation coefficient in the second training data being less than or equal to the second preset coefficient value is used as the third training data.

[0078] Specifically, the size of the correlation coefficient can characterize the importance of the data and the degree of association between different data. Generally, the larger the correlation coefficient, the closer the association between the data. In a neural network, the quality of the output results of the neural network depends largely on the input data. Providing too much data to the machine learning model will lead to a decrease in prediction accuracy, an extension of training time and an increase in the possibility of data overfitting. Therefore, it is very necessary to select appropriate input data. Therefore, in this embodiment, by calculating the correlation coefficient between the input data and the prediction results of the shear wave time difference prediction model (i.e., shear wave time difference), it is possible to accurately determine the input data with the greatest degree of correlation with the prediction results, and determine the input data that is most important to the prediction results of the shear wave time difference prediction model (i.e., shear wave time difference), thereby realizing data screening, reducing the amount of invalid input data, and thereby reducing the amount of calculation and calculation time of the model, while improving prediction efficiency and ensuring that the prediction results of the shear wave time difference prediction model are more accurate.

[0079] In this embodiment, by calculating the correlation coefficient between the input data and the prediction result of the shear wave time difference prediction model (i.e., the shear wave time difference), the data with a correlation coefficient greater than the first preset coefficient value is screened out as the second training data. However, there may be some data with high similarity in the second training data. Using the data with high similarity as input data at the same time will cause repeated use of variables and data redundancy. Therefore, for different types of data in the second training data, the correlation coefficient of any two different types of data is calculated. If the correlation coefficient of two different types of data is greater than the second preset coefficient value, then for the two different types of data with such correlation coefficient greater than the second preset coefficient value, any one type of data is screened out from the two different types of data and merged with the remaining different types of data in the second training data with correlation coefficients less than or equal to the second preset coefficient value as the third training data.

[0080] For example: after screening by the correlation coefficient, the second training data is obtained. There are five groups of different types of data in the second training data (X1, X2, X3, X4 and X5). By calculating the correlation coefficients of any two groups of data in the five groups of data, X1 and X2, X1 and X3, X1 and X4, X1 and X5, X2 and X3, X2 and X4, X2 and X5, X3 and X4, X3 and X5, X4 and X5, among which only the correlation coefficient of data X2 and X3 is greater than the second preset coefficient value, then one group of type data can be selected from the data X2 and X3 (for example, select X3 type data) and the remaining data in the second training data (that is, the data X1, X4 and X5 whose correlation coefficients are less than or equal to the second preset coefficient value) as the third training data. Therefore, the third training data is (X1, X3, X4 and X5).

[0081] For another example, after screening by correlation coefficient, the second training data is obtained. There are five groups of data of different types in the second training data (X1, X2, X3, X4, and X5). By calculating the correlation coefficients of any two groups of data among the five groups of data, X1 and X2, X1 and X3, X1 and X4, X1 and X5, X2 and X3, X2 and X4, X2 and X5, X3 and X4, X3 and X5, and X4 and X5, where the correlation coefficients of data X2 and X3, and data X4 and X5 are both greater than the second preset coefficient value, then any one group of data from data X2 and X3 (for example, X2 type data) and any one group of data from X4 and X5 (for example, X5 type data) are selected together with the remaining data in the second training data (i.e., data with a correlation coefficient less than or equal to the second preset coefficient value: X1) as the third training data. Therefore, the third training data is (X1, X2, and X5).

[0082] Furthermore, the training data set is preprocessed, data is screened based on importance analysis, and data is grouped based on kurtosis and skewness to obtain a processed training data set, further comprising:

[0083] Based on a preset kurtosis coefficient and a preset skewness coefficient, the third training data is divided into at least two groups of well logging data as the processed training data sets.

[0084] Specifically, in this embodiment, the logging data is grouped by characterizing the peak sharpness of the longitudinal wave time difference curve and the degree of asymmetry of the data distribution, and each group of data is used as the input of the model, which can improve the prediction accuracy of the model. As shown in Figures 3 and 4, kurtosis, also known as kurtosis or kurtosis coefficient, is a characteristic number that characterizes the height of the peak of the probability density distribution curve at the mean value, that is, a statistic that describes the steepness of the distribution shape of all values ​​in the population, that is, kurtosis reflects the sharpness of the peak. Skewness, also known as skewness or skewness coefficient, is a measure of the direction and degree of skewness of the statistical data distribution, and is a digital characteristic of the degree of asymmetry of the statistical data distribution.

[0085] Furthermore, the correlation coefficient is calculated using the Pearson correlation coefficient calculation formula.

[0086] Specifically, the Pearson correlation coefficient, also known as the Pearson product-moment correlation coefficient, is widely used to measure the degree of correlation between two variables X and Y. Its value is between -1 and 1. The calculation formula of the Pearson correlation coefficient is:

[0087]

[0088] The Pearson correlation coefficient is the covariance of X and Y divided by the standard deviation of X times the standard deviation of Y.

[0089] As shown in FIG5 , this embodiment further provides a shear wave time difference prediction device, comprising:

[0090] The training data acquisition module 10 is used to obtain well logging sample data as a training data set for the prediction model;

[0091] A first data processing module 20 is configured to preprocess the training data set, perform data screening based on importance analysis, and group the data based on kurtosis and skewness to obtain a processed training data set;

[0092] The model training module 30 is used to input the processed training data set into the neural network constructed by the hybrid CNN and LSTM for training, so as to obtain the shear wave time difference prediction model;

[0093] An input data acquisition module 40 is used to acquire the well logging data of the shear wave time difference to be predicted;

[0094] A second data processing module 50 is configured to pre-process the well logging data, group the well logging data based on kurtosis and skewness, and obtain processed well logging data;

[0095] A result output module 60 is used to use the processed logging data as inputs of the shear wave time difference prediction model to obtain the shear wave time difference;

[0096] The first data processing module 20 is specifically configured to:

[0097] Preprocessing the training data set, screening the data based on importance analysis, and grouping the data based on kurtosis and skewness to obtain a processed training data set includes:

[0098] Performing data cleaning, data filtering, and data normalization on the training data set to obtain first training data;

[0099] Filtering data whose correlation coefficient with the shear wave time difference is greater than a first preset coefficient value from the first training data as second training data;

[0100] respectively calculating correlation coefficients between two different types of data in the second training data;

[0101] If there are two different types of data with a correlation coefficient greater than the second preset coefficient value, one type of data is selected from the two different types of data and used as the third training data with the other different types of data with a correlation coefficient less than or equal to the second preset coefficient value in the second training data;

[0102] Based on a preset kurtosis coefficient and a preset skewness coefficient, the third training data is divided into at least two groups of well logging data as the processed training data sets.

[0103] Furthermore, the CNN neural network and the LSTM neural network in the shear wave time difference prediction model are connected through a Dropout layer.

[0104] Furthermore, the logging data includes: natural gamma logging data, caliper logging data, natural potential logging data, resistivity logging data, neutron logging data, sonic logging data and density logging data.

[0105] Furthermore, the correlation coefficient is calculated using the Pearson correlation coefficient calculation formula.

[0106] This embodiment also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned shear wave time difference prediction method are implemented.

[0107] This embodiment further provides a machine-readable storage medium, on which instructions are stored, and the instructions are used to enable a machine to execute the above-mentioned shear wave time difference prediction method.

[0108] Example 1

[0109] A training data set is obtained, and the well logging sample data set is preprocessed, data is screened based on importance analysis, and data is grouped based on kurtosis and skewness to obtain a preprocessed training data set. The CNN neural network and the LSTM neural network are connected in series through a Dropout layer to form a new neural network structure (a neural network built by hybrid CNN and LSTM). The processed training data set is input into the neural network built by hybrid CNN and LSTM respectively, and a shear wave time difference prediction model is obtained by training, as follows:

[0110] The well logging data is cleaned, filtered, and normalized to obtain the first logging data. This data is formatted accurately and uniformly for machine learning. The number "-9999" is often found in well logging data and must be removed during data cleaning. Data filtering uses a median filter to remove spikes and glitches. Normalization limits the data to a range between 0 and 1, improving the model's convergence speed and accuracy.

[0111] The correlation between different types of data in the first logging data and the shear velocity is usually calculated using the Pearson correlation coefficient method. It can be found that the curves most correlated with the shear time difference (DTS) (the correlation coefficient is greater than the first preset coefficient value. In this embodiment, the first preset coefficient value is set to 0.15) are DTC, CNL, DEN, GR, and RI, respectively. DTC, CNL, DEN, GR, and RI are used as the second logging data. The comparison of the correlation coefficients of each data with the shear time difference (DTS) is shown in Table 1 below:

[0112] Table 1 Comparison of correlation coefficients

[0113] CNL10.990.810.810.520.170.11-0.18DEN0.9910.790.790.480.170.11-0. 12DTC0.810.79110.50.150.1-0.42DTS0.810.79110.50.150.1-0.42GR0.52 0.480.50.510.190.17-0.52RI0.170.170.150.150.1910.98-0.066RT0.110 .110.10.10.170.981-0.066SP-0.18-0.12-0.42-0.42-0.52-0.066-0.0661 CNLDENDTCDTSGRRIRTSP

[0114] Among them, the first column is the names of the eight input curves, from top to bottom they are CNL (compensated neutron), DEN (bulk density), DTC (longitudinal wave time difference), DTS (transverse wave time difference), GR (natural gamma), RI (shallow resistivity), RT (deep resistivity) and SP (spontaneous potential); the ninth row is the names of the eight input curves, from left to right they are CNL (compensated neutron), DEN (bulk density), DTC (longitudinal wave time difference), DTS (transverse wave time difference), GR (natural gamma), RI (shallow resistivity), RT (deep resistivity) and SP (spontaneous potential); the numbers are the Pearson correlation coefficients between the input curves. The larger the value, the higher the correlation, and the smaller the value, the lower the correlation.

[0115] If two highly correlated variables appear simultaneously in the input data, this can lead to variable duplication and data redundancy. Therefore, the correlation coefficients between the five data points (DTC, CNL, DEN, GR, and RI) are calculated. For example, if two data points (CNL and DEN) have correlation coefficients greater than the second preset coefficient, using both CNL and DEN as input variables in the model is equivalent to using the variable "porosity" (both CNL and DEN are porosity estimation curves) twice, which can easily lead to data redundancy and increase calculation time. Therefore, after comprehensive consideration, DTC, DEN, GR, and RI were selected as the third well logging data in this example.

[0116] The third logging data were grouped using kurtosis and skewness as indicators. Wells with kurtosis and skewness greater than 1 in the P-wave time difference between different wells were grouped into the first group, and wells with kurtosis and skewness less than 1 in the P-wave time difference between different wells were grouped into the second group, as shown in Table 2 below:

[0117] Table 2 Kurtosis and skewness grouped data table

[0118]

[0119]

[0120] The two sets of logging data were used as the processed training data sets and respectively input into the neural network constructed by hybrid CNN and LSTM for training to obtain the shear wave time difference prediction model.

[0121] A21-A24 are four new wells for which shear travel time prediction is needed. The processed logging data, obtained through preprocessing and grouping based on kurtosis and skewness, serve as input variables for the shear travel time prediction model. The preprocessing and grouping steps based on kurtosis and skewness are similar to those used in the training dataset and will not be repeated here. The prediction results of the present invention, the regional empirical formula, and the rock physics modeling method are shown in Figure 6. A comparison of the prediction accuracy of the different methods is shown in Table 3:

[0122] Table 3 Comparison of prediction accuracy of different methods

[0123] Well-signed intelligent prediction method, empirical formula method, rock physics modeling method, A21 93.83% 90.91% 92.46% A22 94.43% 91.25% 93.95% A23 94.51% 91.13% 91.14% A24 95.49% 92.46% 90.71%

[0124] The first track in Figure 6 shows the natural gamma (GR), natural potential (SP), and wellbore diameter (CALI). Natural gamma and natural potential represent changes in lithology, while wellbore diameter represents the quality of the wellbore. The second track is the depth track (Depth), which indicates the distance of the measured well section (i.e., the target layer) from the wellhead. The third track is a three-porosity curve, including the compressional time difference (DTC), bulk density (DEN), and compensated neutron (CNL) curves, which are usually used to calculate porosity and are used here to predict the shear wave time difference. The fourth track is a resistivity curve, including deep resistivity (RT), shallow resistivity (RI), and micro resistivity (RXO), which are usually used to identify oil, gas, and water layers and calculate saturation. Here, it is used to predict the shear wave time difference. The fifth track is the shear wave comparison, including the shear wave time difference (DTS) and the intelligent prediction method, which is used to compare the shear wave time difference obtained by the intelligent prediction method with the shear wave time difference actually measured. Track 6 shows a shear-wave comparison, including DTS and empirical formula methods, comparing the DTS obtained using the empirical formula method with the measured S-wave transit time. Track 7 shows a shear-wave comparison, including DTS and rock physics modeling methods, comparing the DTS obtained using the rock physics modeling method with the measured S-wave transit time.

[0125] By comparing Figure 6 and Table 3, it can be concluded that the shear wave time difference predicted by the method of the present application has the advantages of high accuracy, small error, and strong generalization ability compared with the shear wave time difference obtained by regional empirical formula and rock physics modeling method.

[0126] The above describes in detail the optional implementation methods of the embodiments of the present invention in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above implementation methods. Within the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the scope of protection of the embodiments of the present invention.

[0127] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any appropriate manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe various possible combinations.

[0128] Those skilled in the art will understand that all or part of the steps in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a program, which is stored in a storage medium and includes a number of instructions for causing a single-chip microcomputer, chip or processor to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program code.

[0129] In addition, various implementations of the embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the embodiments of the present invention, they should also be regarded as the contents disclosed in the embodiments of the present invention.

Claims

1. A shear wave time difference prediction method, characterized in that: include: Obtaining well logging sample data as a training data set for the prediction model; Preprocessing the training data set, screening the data based on importance analysis, and grouping the data based on kurtosis and skewness to obtain a processed training data set; The processed training data sets are input into the neural network constructed by hybrid CNN and LSTM for training to obtain the shear wave time difference prediction model; Obtaining well logging data for the shear wave time difference to be predicted; Preprocessing the well logging data, grouping the well logging data based on kurtosis and skewness, and obtaining processed well logging data; The processed logging data are respectively used as inputs of the shear wave time difference prediction model to obtain the shear wave time difference; Preprocessing the training data set, screening the data based on importance analysis, and grouping the data based on kurtosis and skewness to obtain a processed training data set includes: Performing data cleaning, data filtering, and data normalization on the training data set to obtain first training data; Filtering data whose correlation coefficient with the shear wave time difference is greater than a first preset coefficient value from the first training data as second training data; respectively calculating correlation coefficients between two different types of data in the second training data; If there are two different types of data with a correlation coefficient greater than the second preset coefficient value, one type of data is selected from the two different types of data and used as the third training data with the other different types of data with a correlation coefficient less than or equal to the second preset coefficient value in the second training data; Based on a preset kurtosis coefficient and a preset skewness coefficient, the third training data is divided into at least two groups of well logging data as the processed training data sets.

2. The method according to claim 1, characterized in that The CNN neural network and the LSTM neural network in the shear wave time difference prediction model are connected through a Dropout layer.

3. The method according to claim 1, characterized in that The logging data includes natural gamma ray logging data, caliper logging data, natural potential logging data, resistivity logging data, neutron logging data, sonic logging data and density logging data.

4. The method according to claim 1, wherein The correlation coefficient is calculated using the Pearson correlation coefficient calculation formula.

5. A shear wave time difference prediction device, characterized in that: include: A training data acquisition module is used to obtain well logging sample data as a training data set for the prediction model; a first data processing module, configured to preprocess the training data set, perform data screening based on importance analysis, and group the data based on kurtosis and skewness to obtain a processed training data set; The model training module is used to input the processed training data set into the neural network constructed by CNN and LSTM respectively for training to obtain the shear wave time difference prediction model; An input data acquisition module is used to acquire the logging data of the shear wave time difference to be predicted; a second data processing module, configured to pre-process the well logging data, group the well logging data based on kurtosis and skewness, and obtain processed well logging data; A result output module is used to use the processed logging data as input of the shear wave time difference prediction model to obtain the shear wave time difference; The first data processing module is specifically configured to: Performing data cleaning, data filtering, and data normalization on the training data set to obtain first training data; Filtering data whose correlation coefficient with the shear wave time difference is greater than a first preset coefficient value from the first training data as second training data; respectively calculating correlation coefficients between two different types of data in the second training data; If there are two different types of data with a correlation coefficient greater than the second preset coefficient value, one type of data is selected from the two different types of data and used as the third training data with the other different types of data with a correlation coefficient less than or equal to the second preset coefficient value in the second training data; Based on a preset kurtosis coefficient and a preset skewness coefficient, the third training data is divided into at least two groups of well logging data as the processed training data sets.

6. The device according to claim 5, characterized in that The CNN neural network and the LSTM neural network in the shear wave time difference prediction model are connected through a Dropout layer.

7. The device according to claim 5, characterized in that The logging data includes natural gamma ray logging data, caliper logging data, natural potential logging data, resistivity logging data, neutron logging data, sonic logging data and density logging data.

8. The device according to claim 5, characterized in that The correlation coefficient is calculated using the Pearson correlation coefficient calculation formula.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the shear wave time difference prediction method according to any one of claims 1 to 4 are implemented.

10. A machine-readable storage medium having instructions stored thereon, wherein the instructions are used to enable a machine to execute the shear wave time difference prediction method according to any one of claims 1 to 4.

Images