A dynamic time warping-based synthetic seismic record calibration method and system

Abstract

The present application relates to the technical field of oil and gas exploration, and particularly relates to a synthetic seismic record calibration method and system based on dynamic time planning. The method comprises making a synthetic seismic record and obtaining an actual seismic record, extracting a time sequence and creating a cost matrix, calculating the cumulative cost of each element in the cost matrix, and finding the path with the minimum cumulative cost as the optimal path by backtracking, adjusting the synthetic seismic record to align with the actual seismic record on the time axis according to the optimal path found, and obtaining the calibrated synthetic seismic record. The present application based on dynamic time planning can compensate for the time deviation caused by inaccurate velocity model or horizon deformation, thereby improving the time alignment between the synthetic and actual seismic records, improving the accuracy of well-to-seismic calibration, and allowing local time adjustment to realize intelligent synthetic seismic record calibration without human control.

Description

Technical Field

[0001] This invention relates to the field of oil and gas exploration technology, and in particular to a method and system for calibrating synthetic seismic records based on dynamic time programming. Background Technology

[0002] Accurate synthetic seismic record calibration is crucial for structural interpretation and reservoir prediction in oil exploration and development. The purpose of synthetic seismic record calibration is to precisely correlate reflection events in seismic data with stratigraphic interfaces, lithological variations, and other geological features in geological and well logging data, thereby improving the reliability and accuracy of seismic data interpretation. However, in practice, poor synthetic seismic record calibration results are often encountered, manifested as low correlation between synthetic and actual seismic records. This phenomenon can be caused by various factors, including velocity distortion during seismic data processing, well logging data quality issues, and inappropriate wavelet selection.

[0003] To improve the correlation coefficient between synthetic seismic records and actual seismic records, common methods include re-evaluating and correcting velocity models in seismic processing, improving well logging data quality, and optimizing wavelet selection. While these methods improve the accuracy of synthetic seismic record calibration to some extent, they also have limitations. The inhomogeneity and complexity of the subsurface medium make it difficult to perfectly establish velocity models, while systematic errors in well logging data limit the effectiveness of well logging curve correction. Furthermore, seismic velocities and well logging velocities are obtained through completely different measurement methods and inherently possess systematic differences; therefore, relying solely on seismic processing velocity correction, well logging curve correction, and wavelet optimization methods cannot solve the problem of accurate synthetic seismic record calibration. Summary of the Invention

[0004] The technical problem to be solved by the embodiments of the present invention is to provide a synthetic seismic record calibration method and system based on dynamic time planning, so as to solve the problem of accurate synthetic seismic record calibration in the prior art.

[0005] This invention discloses a method for calibrating synthetic seismic records based on dynamic time programming, comprising: S1. Obtain well logging curves and select a wavelet model to create synthetic seismic records, and obtain actual seismic records corresponding to the positions of the well logging curves; S2. Extract the first original time series corresponding to the synthetic seismic record and the second original time series corresponding to the actual seismic record, and create a cost matrix based on the first original time series and the second original time series; S3. Define a distance metric function based on dynamic time planning, calculate the cost of each element in the cost matrix using the distance metric function, and update the cumulative cost of the corresponding element by combining the cumulative cost of the element's predecessor position. S4. Starting from the element at the end of the cost matrix, trace backwards to find the path with the minimum cumulative cost as the optimal path; S5. Adjust the synthetic seismic record according to the found optimal path until the adjusted synthetic seismic record is aligned with the actual seismic record on the time axis, thus obtaining the calibrated synthetic seismic record.

[0006] Optionally, the step of acquiring well logging curves and selecting a wavelet model to create synthetic seismic records includes: The acoustic impedance of the formation is obtained based on seismic and well logging data, and the reflection coefficient is calculated based on the obtained acoustic impedance. By selecting a wavelet model and convolving the reflection coefficient with the selected wavelet model, the synthetic seismic record is obtained.

[0007] Optionally, methods for calculating acoustic impedance and reflection coefficient are included: The acoustic transit time and rock density are extracted from seismic and well logging data. The acoustic impedance is calculated based on the obtained acoustic transit time and rock density. The calculation function for the acoustic impedance is as follows:

[0008] In the formula, Indicates acoustic impedance. Indicates the time difference of sound waves. Indicates rock density; The acoustic impedance of two adjacent strata is obtained, and the reflection coefficient is calculated based on the acoustic impedance of the two adjacent strata. The calculation function of the reflection coefficient is as follows:

[0009] In the formula, Represents the reflection coefficient. Indicates the acoustic impedance of the first stratum. It is the acoustic impedance of the adjacent second stratum.

[0010] Optionally, this includes a method for creating the cost matrix: Let the length of the first original time series be n, and the length of the second original time series be m, and the creation size be... The cost matrix; Initialize all elements in the first column of the cost matrix to infinity, and represent them as D. The elements in the first row of the cost matrix are initialized to infinity and denoted as D. And set the elements at the valid initial positions in the cost matrix to D. .

[0011] Optionally, the method includes calculating the cost of each element in the cost matrix and updating the cumulative cost corresponding to that element: Define the distance metric function d The cost of each element in the cost matrix is ​​calculated using the distance metric function, and the expression for the distance metric function is as follows: d

[0012] In the formula, d Represents the distance metric function. This represents the i-th data point in the first original time series. This represents the j-th data point in the second original time series. This represents the straight-line distance between the i-th data point in the first original time series and the j-th data point in the second original time series. Obtain the cumulative cost of the three predecessor positions of the target element, and select the minimum cumulative cost; Based on the cost calculated for the target element and the selected minimum cumulative cost, the cumulative cost of the target element is calculated. The function expression for calculating the cumulative cost of the target element is as follows:

[0013] In the formula, This represents the target element containing the minimum cumulative cost from the first data point to the i-th data point in the first original time series, and the minimum cumulative cost from the first data point to the j-th data point in the second original time series; This represents the cumulative cost of the element located above the target element at the predecessor position. This represents the cumulative cost of the element located to the left of the target element. This represents the cumulative cost of the element located at the upper left predecessor position of the target element; Starting from position (1,1) in the cost matrix, calculate the cumulative cost of the corresponding element and fill it in until the position (n,m) in the cost matrix is ​​filled.

[0014] Optionally, the step of tracing backwards from the element at the end of the cost matrix to find the path with the minimum cumulative cost as the optimal path includes: Starting from the element at position (n, m) in the cost matrix, reverse tracing is performed, and each move is made along the top, left, and upper left of the current element position respectively; Record the position of each move during the reverse tracing process, and find the element corresponding to the minimum cumulative cost at each move position, until the element in the cost matrix is ​​reached. D ; The optimal path is obtained by connecting the selected elements through reverse tracing.

[0015] Optionally, adjusting the synthetic seismic record according to the found optimal path until the adjusted synthetic seismic record is aligned with the actual seismic record on the time axis includes: A first new time series is created for the synthetic seismic record, and a second new time series is created for the actual seismic record, wherein the lengths of the first new time series and the second new time series are the same as the length of the optimal path, based on the obtained optimal path. Traverse each element in the optimal path and fill the data points contained in the element into the corresponding first new time series or second new time series, so that the data points in the first new time series are aligned with the data points in the second new time series; If the i-th data point contained in the element is within the valid range of the first original time series, then the i-th data point is filled into the first new time series; if the i-th data point contained in the element is not within the valid range of the first original time series, then missing values ​​or placeholders are used to fill it. If the j-th data point contained in the element is within the valid range of the second original time series, then the j-th data point is filled into the second new time series. If the j-th data point contained in the element is not within the valid range of the first original time series, then missing values ​​or placeholders are used to fill it.

[0016] Optionally, this includes methods for assessing the degree of match between the aligned synthetic seismic record and the actual seismic record: Select a time window by extracting attributes or targeting the reservoir inversion; Within the selected time window, the correlation coefficient between the first new time series and the second new time series is calculated, and the degree of matching between the aligned synthetic seismic record and the actual seismic record is evaluated based on the calculated correlation coefficient. The function expression for calculating the correlation coefficient is as follows:

[0017] In the formula, Represents the correlation coefficient. This represents the average value of the first new time series. This represents the average value of the second new time series.

[0018] Optionally, this includes a method for iteratively optimizing the correlation coefficient: Based on the evaluation results of the matching degree between the aligned synthetic seismic record and the actual seismic record, the distance metric function, the constraint of the optimal path, and the size of the time window are adjusted, and steps S2-S5 are repeated until the adjustment range is less than the preset threshold or the preset number of iterations is reached.

[0019] This invention also discloses a system for calibrating synthetic seismic records, employing the aforementioned synthetic seismic record calibration method. The system comprises: The seismic record acquisition module is used to acquire well logging curves, select a wavelet model to create synthetic seismic records, and acquire the actual seismic records corresponding to the positions of the well logging curves. The cost matrix building module is used to extract the first original time series corresponding to the synthetic seismic record and the second original time series corresponding to the actual seismic record, and to create a cost matrix based on the first original time series and the second original time series. The cumulative cost calculation module is used to define a distance metric function based on dynamic time planning, calculate the cost of each element in the cost matrix through the distance metric function, and update the cumulative cost of the corresponding element by combining the cumulative cost of the element's predecessor position. The optimal path finding module is used to trace backwards from the element at the end of the cost matrix to find the path with the minimum cumulative cost as the optimal path; The synthetic seismic record calibration module is used to adjust the synthetic seismic record according to the found optimal path until the adjusted synthetic seismic record is aligned with the actual seismic record on the time axis, thus obtaining the calibrated synthetic seismic record.

[0020] Compared with the prior art, the synthetic seismic record calibration method and system based on dynamic time programming provided in this invention have the following advantages: The intelligent synthetic seismic record calibration method based on dynamic time programming can compensate for time deviations caused by inaccurate velocity models or stratigraphic deformation, thereby improving the time alignment between synthetic and actual seismic records and enhancing the accuracy of well-seismic calibration. It can more accurately correlate stratigraphic information in well logging data with reflection events in seismic data, which is crucial for subsequent stratigraphic interface interpretation, stratigraphic thickness estimation, and stratigraphic attribute prediction. Furthermore, unlike the global linear stretching method, the dynamic time programming algorithm can perform local time adjustments. This means that even if only some stratigraphic layers have velocity errors or deformations, these local time deviations can be adjusted, achieving intelligent synthetic seismic record calibration without human intervention. Attached Figure Description

[0021] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a schematic block diagram illustrating the steps of the synthetic seismic record calibration method provided in an embodiment of the present invention; Figure 2 Figure (a) is a sound wave time difference curve provided in an embodiment of the present invention; Figure (b) is a rock density curve provided in an embodiment of the present invention; Figure (c) is a sound wave impedance curve provided in an embodiment of the present invention; Figure (d) is a reflection coefficient distribution map provided in an embodiment of the present invention; Figure (e) is a synthetic seismic record distribution map provided in an embodiment of the present invention; and Figure (f) is an actual seismic record distribution map provided in an embodiment of the present invention. Figure 3 To align the time series plots of the synthetic seismic record and the actual seismic record before alignment; Figure 4 This is a time series diagram of aligned synthetic seismic records and actual seismic records provided in an embodiment of the present invention; Figure 5 A correlation distribution map of the synthetic seismic record and the actual seismic record before alignment; Figure 6 The correlation distribution diagram between the aligned synthetic seismic record and the actual seismic record is provided for embodiments of the present invention. Detailed Implementation

[0022] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0023] This invention discloses a method for calibrating synthetic seismic records based on dynamic time programming, comprising: S1. Obtain well logging curves and select wavelet model to create synthetic seismic records, and obtain actual seismic records corresponding to the positions of well logging curves; S2. Extract the first original time series corresponding to the synthetic seismic record and the second original time series corresponding to the actual seismic record, and create a cost matrix based on the first and second original time series; S3. Define a distance metric function based on dynamic time planning, calculate the cost of each element in the cost matrix using the distance metric function, and update the cumulative cost of the corresponding element by combining the cumulative cost of the element's predecessor position. S4. Starting from the element at the end of the cost matrix, trace backwards to find the path with the minimum cumulative cost as the optimal path; S5. Adjust the synthetic seismic record according to the found optimal path until the adjusted synthetic seismic record is aligned with the actual seismic record on the time axis, and obtain the calibrated synthetic seismic record.

[0024] Through the implementation of the above-described synthetic seismic record calibration method, the intelligent synthetic seismic record calibration method based on dynamic time programming can compensate for time deviations caused by inaccurate velocity models or stratigraphic deformation, thereby improving the time alignment between synthetic and actual seismic records and enhancing the accuracy of well-seismic calibration. It can more accurately correlate formation information in well logging data with reflection events in seismic data, which is crucial for subsequent stratigraphic interface interpretation, formation thickness estimation, and formation attribute prediction. Furthermore, unlike the global linear stretching method, the dynamic time programming algorithm can perform local time adjustments. This means that even if only some strata have velocity errors or formation deformations, these local time deviations can be adjusted, achieving intelligent synthetic seismic record calibration without human intervention.

[0025] Furthermore, well logging curves are acquired and a wavelet model is selected to create a synthetic seismic record, including: The acoustic impedance of the formation is obtained from seismic and well logging data, and the reflection coefficient is calculated based on the obtained acoustic impedance. By selecting a wavelet model and convolving the reflection coefficient with the selected wavelet model, a synthetic seismic record is obtained.

[0026] Furthermore, this includes methods for calculating acoustic impedance and reflection coefficient: Acoustic transit time and rock density are extracted from seismic and well logging data. Acoustic impedance is then calculated based on these data. The formula for calculating acoustic impedance is:

[0027] In the formula, Indicates acoustic impedance. Indicates the time difference of sound waves. Indicates rock density; The acoustic impedance of two adjacent strata is obtained, and the reflection coefficient is calculated based on the acoustic impedance of the two adjacent strata. The calculation function of the reflection coefficient is as follows:

[0028] In the formula, Represents the reflection coefficient. Indicates the acoustic impedance of the first stratum. It is the acoustic impedance of the adjacent second stratum.

[0029] Through the implementation of the above-described synthetic seismic record calibration method, the creation of synthetic seismic records is a crucial process for combining seismic and well logging data, and is also the foundation for seismic data interpretation and well-seismic calibration. Acoustic impedance (AI) is calculated using sonic transit time and rock density from well logging data. The reflection coefficient (RC) represents the intensity of sonic reflection at formation interfaces, and can be calculated from the difference in acoustic impedance between adjacent formations. A suitable wavelet model is selected; this can be a theoretical wavelet, such as a zero-phase waveform, a minimum-phase waveform, or a wavelet extracted from actual seismic data. Convolving the reflection coefficient sequence with the selected wavelet generates the synthetic seismic record. Figure 2 Figure (a) shows the acoustic time difference curve, (b) shows the rock density curve, (c) shows the acoustic impedance curve, (d) shows the reflection coefficient distribution, (e) shows the distribution of synthetic seismic records, and (f) shows the distribution of actual seismic records.

[0030] Furthermore, this includes methods for creating cost matrices: Set the length of the first original time series to n, and the length of the second original time series to m, and create a size of... The cost matrix; Initialize all elements in the first column of the cost matrix to infinity, and represent them as D. Initialize all elements in the first row of this matrix to infinity, and represent it as D. And set the elements at the effective initial positions in the cost matrix to D. .

[0031] Through the implementation of the above-described synthetic seismic record calibration method, synthetic seismic records and actual seismic records can be represented as two sequences: synthetic seismic record X, with a length of n, and actual seismic record Y, with a length of m. To handle boundary cases, a sequence of size m is established. The cost matrix is ​​modified by adding an extra row and column to handle boundary conditions between synthetic and actual seismic records, respectively. To ensure the matching process always begins within the valid range of both sequences, the first row of the cost matrix, D... And the first column D Initialize to infinity (or a sufficiently large number) to indicate that matching cannot begin outside of the synthetic or actual seismic records; that is, ensure that the matching process always begins from within both sequences to obtain accurate matching results. Simultaneously, the effective initial location element D is... Set to 0 to indicate that the two sequences have not yet started matching at the valid initial position.

[0032] Furthermore, this includes methods for calculating the cost of each element in the cost matrix and updating the cumulative cost of the corresponding element: Define the distance metric function d The cost of each element in the cost matrix is ​​calculated using a distance metric function. The expression for the distance metric function is as follows: d

[0033] In the formula, d Represents the distance metric function. This represents the i-th data point in the first original time series. This represents the j-th data point in the second original time series. This represents the straight-line distance between the i-th data point in the first original time series and the j-th data point in the second original time series. Obtain the cumulative cost of the three predecessor positions of the target element, and select the minimum cumulative cost; Based on the cost calculated for the target element and the selected minimum cumulative cost, the cumulative cost of the target element is calculated. The function expression for calculating the cumulative cost of the target element is as follows:

[0034] In the formula, This represents the target element containing the minimum cumulative cost from the first data point to the i-th data point in the first original time series, and the minimum cumulative cost from the first data point to the j-th data point in the second original time series; This represents the cumulative cost of the element located above the target element at the predecessor position. This represents the cumulative cost of the element located to the left of the target element. This represents the cumulative cost of the element located at the upper left predecessor position of the target element; Starting from position (1,1) in the cost matrix, calculate the cumulative cost of the corresponding element and fill it in until the position (n,m) in the cost matrix is ​​filled.

[0035] In implementing the above-described synthetic seismic record calibration method, the calculation of the cost matrix is ​​a core step in the dynamic time programming algorithm. The cost matrix is ​​used to record the cumulative distance of element matching between sequence pairs, facilitating the finding of an optimal path to align two sequences, thereby allowing subsequent aligned sequences to scale non-linearly over time.

[0036] First, define a distance metric function. This is used to measure the distance between the i-th data point in the first original time series X and the j-th data point in the second original time series Y. Euclidean distance is typically used as the standard distance metric; it is a metric based on spatial geometry that measures the straight-line distance between two points. Then, each element in the cost matrix needs to be filled, starting from position (1,1) and continuing up to position (n,m). Each element... The minimum cumulative distances from the starting point of the first original time series to the i-th data point, and from the starting point of the second original time series to the j-th data point, are recorded. The existence of the cost matrix allows the algorithm to track these minimum cumulative costs and find the optimal alignment path between the two series. Therefore, all possible paths are considered, enabling it to adapt to time scaling between series and find alignments that maintain high correlation even with velocity mismatches. This means that regardless of time delays or shortening between series, the algorithm can find the optimal alignment path to maintain correlation between them. By utilizing such a cost matrix, the cumulative distances of different paths can be calculated and compared, ultimately finding the optimal correspondence between synthetic and actual seismic records. This can be used to generate an alignment matrix to more intuitively illustrate the relationship between the two series. The alignment matrix provides a visual way to show how each element in one series corresponds and matches with elements in another series.

[0037] Furthermore, starting from the element at the end of the cost matrix, we trace backwards to find the path with the minimum cumulative cost as the optimal path, including: Starting from the element at position (n, m) in the cost matrix, trace backwards, and move upwards, to the left, and to the upper left of the current element position each time. Record the position of each move during the reverse tracing process, and find the element corresponding to the minimum cumulative cost at each move position, until the element in the cost matrix is ​​reached. D ; The optimal path is obtained by connecting the selected elements through reverse tracing.

[0038] Through the implementation of the above-described synthetic seismic record calibration method, the step of finding the optimal path aims to determine a path from the starting point to the ending point. This path, represented by a cost matrix, indicates the optimal alignment of the two time series. The optimal path is the path with the lowest cost, which is traced back using the cost matrix. Starting from the bottom right corner of the cost matrix, i.e., the element at position (n, m), tracing back to the element at position (m)... In each step, the position with the lowest cumulative cost among the three possible forward drive positions will be selected. This indicates moving up one step. This means move one step to the left. This indicates moving one step diagonally. During the tracing process, each movement point is recorded. These points will form the optimal path, representing the best alignment between the first and second original time series, until tracing back to the element point D in the top left corner of the cost matrix. .

[0039] The above tracing process demonstrates how to align two time series in a non-linear manner, aligning their similarities along the time axis. The optimal path guarantees that the total distance of the entire sequence alignment (i.e., the cost of mismatch between sequences) is minimized. In this way, an optimal alignment scheme can be found to minimize the differences between the sequences.

[0040] Furthermore, the synthetic seismic record is adjusted according to the found optimal path until the adjusted synthetic seismic record is aligned with the actual seismic record on the time axis, including: To create a new first time series for the synthetic seismic record and a new second time series for the actual seismic record, the lengths of the first and second time series are made the same as the length of the optimal path, based on the obtained optimal path. Traverse each element in the optimal path and fill the data points contained in the element into the corresponding first new time series or second new time series, so that the data points in the first new time series are aligned with the data points in the second new time series. If the i-th data point contained in the checked element is within the valid range of the first original time series, then the i-th data point is filled into the first new time series; if the i-th data point contained in the checked element is not within the valid range of the first original time series, then missing values ​​or placeholders are used to fill it. If the j-th data point contained in the checked element is within the valid range of the second original time series, then the j-th data point is filled into the second new time series. If the j-th data point contained in the checked element is not within the valid range of the first original time series, then missing values ​​or placeholders are used to fill it.

[0041] Through the implementation of the above-described synthetic seismic record calibration method, path transformation converts the calculated optimal path into a correspondence between elements in two sequences. The aim is to identify which elements match and determine how to align the two sequences using insertion and deletion operations.

[0042] As mentioned above, the optimal path obtained using dynamic time planning algorithms is typically a list of paired indices, such as [(0,0),(0,1),(1,2)……(n,m)]. Two new sequences are created from this optimal path, with the same length as the optimal path. Each element in the new sequence corresponds to a data point in the original sequence. By traversing the coordinates of each element in the optimal path, we can guide how to populate the new sequences. For example, if a point in the path is... This means that the i-th data point of the first original time series is aligned with the j-th data point of the second original time series. Path transformation allows for clearer identification of similar patterns in the two time series and accurate mapping between them. This adjusts for local temporal deviations between data points in the series to find the optimal global alignment. Path transformation more clearly represents the matching relationship between the two series, as well as their similarities and differences. Through this process, accurate and comprehensive alignment results can be obtained, helping to more clearly present the potential correspondence between synthetic seismic records and actual seismic records. Figure 3 This is a time series diagram of the synthetic seismic record and the actual seismic record before alignment. Figure 4 It is a time series diagram of the synthesized seismic record and the actual seismic record after dynamic time planning alignment.

[0043] Furthermore, this includes methods for assessing the degree of matching between the aligned synthetic seismic record and the actual seismic record: Select a time window by extracting attributes or targeting the reservoir inversion; Within the selected time window, the correlation coefficient between the first and second new time series is calculated, and the degree of matching between the aligned synthetic seismic record and the actual seismic record is evaluated based on the calculated correlation coefficient. The function expression for calculating the correlation coefficient is as follows:

[0044] In the formula, Represents the correlation coefficient. This represents the average value of the first new time series. This represents the average value of the second new time series.

[0045] By implementing the above-described synthetic seismic record calibration method, the degree of matching between the synthetic seismic record and the actual seismic record is evaluated by calculating the correlation coefficient between the first and second new time series. Here, the correlation coefficient refers to the Pearson correlation coefficient, which is an indicator of the linear correlation between two variables, with values ​​ranging from -1 to 1. When the correlation coefficient is close to 1, it indicates a strong positive linear correlation between the two series, meaning their trends are roughly the same. Conversely, when the correlation coefficient is close to -1, it indicates a strong negative linear correlation between the two series, meaning their trends are opposite. When the correlation coefficient is close to 0, it indicates no significant linear correlation between the two series. Figure 5 This is a correlation distribution map between the pre-aligned synthetic seismic record and the actual seismic record, indicating a correlation coefficient of 0.79. Figure 6 This is a correlation distribution diagram of the synthetic seismic records and actual seismic records after alignment using dynamic time programming, showing a correlation coefficient of 0.94. It is evident that the correlation between the synthetic and actual seismic records is significantly improved after alignment using the dynamic time programming algorithm.

[0046] Furthermore, this includes methods for iteratively optimizing the correlation coefficient: Based on the evaluation results of the matching degree between the aligned synthetic seismic record and the actual seismic record, adjust the distance metric function, the constraint of the optimal path, and the size of the time window, and repeat steps S2-S5 until the adjustment range is less than the preset threshold or the preset number of iterations is reached.

[0047] By implementing the above-described synthetic seismic record calibration method, and through iterative optimization of the correlation coefficient, the matching error can be further reduced or the correlation coefficient can be increased, thereby gradually improving the accuracy of synthetic seismic record calibration. This helps to improve the interpretation, analysis, and prediction capabilities of seismic data, and provides a more reliable basis for decision-making in seismic exploration and related fields.

[0048] This invention also discloses a system for calibrating synthetic seismic records, employing the aforementioned synthetic seismic record calibration method. The system comprises: The seismic record acquisition module is used to acquire well logging curves, select wavelet models to create synthetic seismic records, and acquire actual seismic records corresponding to the well logging curve positions. The cost matrix building module is used to extract the first original time series corresponding to the synthetic seismic record and the second original time series corresponding to the actual seismic record, and to create a cost matrix based on the first and second original time series. The cumulative cost calculation module is used to define a distance metric function based on dynamic time planning, calculate the cost of each element in the cost matrix through the distance metric function, and update the cumulative cost of the corresponding element by combining the cumulative cost of the element's predecessor position. The optimal path finding module is used to trace backwards from the element at the end of the cost matrix to find the path with the minimum cumulative cost as the optimal path. The synthetic seismic record calibration module is used to adjust the synthetic seismic record according to the found optimal path until the adjusted synthetic seismic record is aligned with the actual seismic record on the time axis, thus obtaining the calibrated synthetic seismic record.

[0049] Based on the above-described synthetic seismic record calibration method, this invention also discloses a computer-readable storage medium, which, when executed by a processor, implements the steps of the above-described synthetic seismic record calibration method.

[0050] Based on the above-described synthetic seismic record calibration method, the present invention also discloses a computer 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, it implements the steps of the above-described synthetic seismic record calibration method.

[0051] This invention is described based on flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to specific embodiments. It should be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing device, generate instructions for implementing the flowcharts and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A system that specifies functions in one or more boxes.

[0052] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including an instruction set implemented in a process. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0053] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0054] It should be understood that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Those skilled in the art can modify the technical solutions described in the above embodiments, or make equivalent substitutions for some of the technical features; and all such modifications and substitutions should fall within the protection scope of the appended claims of the present invention.

Claims

1. A method for calibrating synthetic seismic records based on dynamic time programming, characterized in that, The synthetic seismic record calibration method includes: S1. Obtain well logging curves and select a wavelet model to create synthetic seismic records, and obtain actual seismic records corresponding to the positions of the well logging curves; S2. Extract the first original time series corresponding to the synthetic seismic record and the second original time series corresponding to the actual seismic record, and create a cost matrix based on the first original time series and the second original time series; S3. Define a distance metric function based on dynamic time planning, calculate the cost of each element in the cost matrix using the distance metric function, and update the cumulative cost of the corresponding element by combining the cumulative cost of the element's predecessor position. S4. Starting from the element at the end of the cost matrix, trace backwards to find the path with the minimum cumulative cost as the optimal path; S5. Adjust the synthetic seismic record according to the found optimal path until the adjusted synthetic seismic record is aligned with the actual seismic record on the time axis, thus obtaining the calibrated synthetic seismic record.

2. The synthetic seismic record calibration method based on dynamic time programming according to claim 1, characterized in that, The process of acquiring well logging curves and selecting a wavelet model to create synthetic seismic records includes: The acoustic impedance of the formation is obtained based on seismic and well logging data, and the reflection coefficient is calculated based on the obtained acoustic impedance. By selecting a wavelet model and convolving the reflection coefficient with the selected wavelet model, the synthetic seismic record is obtained.

3. The synthetic seismic record calibration method based on dynamic time programming according to claim 2, characterized in that, Including methods for calculating acoustic impedance and reflection coefficient: The acoustic transit time and rock density are extracted from seismic and well logging data. The acoustic impedance is calculated based on the obtained acoustic transit time and rock density. The calculation function for the acoustic impedance is as follows: In the formula, Indicates acoustic impedance. Indicates the time difference of sound waves. Indicates rock density; The acoustic impedance of two adjacent strata is obtained, and the reflection coefficient is calculated based on the acoustic impedance of the two adjacent strata. The calculation function of the reflection coefficient is as follows: In the formula, Represents the reflection coefficient. Indicates the acoustic impedance of the first stratum. It is the acoustic impedance of the adjacent second stratum.

4. The synthetic seismic record calibration method based on dynamic time programming according to claim 1, characterized in that, Including methods for creating cost matrices: Let the length of the first original time series be n, and the length of the second original time series be m, and the creation size be... The cost matrix; Initialize all elements in the first column of the cost matrix to infinity, and represent them as D. The elements in the first row of the cost matrix are initialized to infinity and denoted as D. And set the elements at the valid initial positions in the cost matrix to D. .

5. The synthetic seismic record calibration method based on dynamic time programming according to claim 4, characterized in that, This includes a method for calculating the cost of each element in the cost matrix and updating the cumulative cost of the corresponding element: Define the distance metric function d The cost of each element in the cost matrix is ​​calculated using the distance metric function, and the expression for the distance metric function is as follows: d In the formula, d Represents the distance metric function. This represents the i-th data point in the first original time series. This represents the j-th data point in the second original time series. This represents the straight-line distance between the i-th data point in the first original time series and the j-th data point in the second original time series. Obtain the cumulative cost of the three predecessor positions of the target element, and select the minimum cumulative cost; Based on the cost calculated for the target element and the selected minimum cumulative cost, the cumulative cost of the target element is calculated. The function expression for calculating the cumulative cost of the target element is as follows: In the formula, This represents the target element containing the minimum cumulative cost from the first data point to the i-th data point in the first original time series, and the minimum cumulative cost from the first data point to the j-th data point in the second original time series; This represents the cumulative cost of the element located above the target element at the predecessor position. This represents the cumulative cost of the element located to the left of the target element. This represents the cumulative cost of the element located at the upper left predecessor position of the target element; Starting from position (1,1) in the cost matrix, calculate the cumulative cost of the corresponding element and fill it in until the position (n,m) in the cost matrix is ​​filled.

6. The synthetic seismic record calibration method based on dynamic time programming according to claim 5, characterized in that, The step of tracing backwards from the element at the end of the cost matrix to find the path with the minimum cumulative cost as the optimal path includes: Starting from the element at position (n, m) in the cost matrix, reverse tracing is performed, and each move is made along the top, left, and upper left of the current element position respectively; Record the position of each move during the reverse tracing process, and find the element corresponding to the minimum cumulative cost at each move position, until the element in the cost matrix is ​​reached. D ; The optimal path is obtained by connecting the selected elements through reverse tracing.

7. The synthetic seismic record calibration method based on dynamic time programming according to claim 6, characterized in that, The step of adjusting the synthetic seismic record according to the found optimal path until the adjusted synthetic seismic record is aligned with the actual seismic record on the time axis includes: A first new time series is created for the synthetic seismic record, and a second new time series is created for the actual seismic record, wherein the lengths of the first new time series and the second new time series are the same as the length of the optimal path, based on the obtained optimal path. Traverse each element in the optimal path and fill the data points contained in the element into the corresponding first new time series or second new time series, so that the data points in the first new time series are aligned with the data points in the second new time series; If the i-th data point contained in the element is within the valid range of the first original time series, then the i-th data point is filled into the first new time series; if the i-th data point contained in the element is not within the valid range of the first original time series, then missing values ​​or placeholders are used to fill it. If the j-th data point contained in the element is within the valid range of the second original time series, then the j-th data point is filled into the second new time series. If the j-th data point contained in the element is not within the valid range of the first original time series, then missing values ​​or placeholders are used to fill it.

8. The synthetic seismic record calibration method based on dynamic time programming according to claim 7, characterized in that, This includes methods for assessing the degree of match between aligned synthetic seismic records and actual seismic records: Select a time window by extracting attributes or targeting the reservoir inversion; Within the selected time window, the correlation coefficient between the first new time series and the second new time series is calculated, and the degree of matching between the aligned synthetic seismic record and the actual seismic record is evaluated based on the calculated correlation coefficient. The function expression for calculating the correlation coefficient is as follows: In the formula, Represents the correlation coefficient. This represents the average value of the first new time series. This represents the average value of the second new time series.

9. The synthetic seismic record calibration method based on dynamic time programming according to claim 8, characterized in that, This includes methods for iteratively optimizing the correlation coefficient: Based on the evaluation results of the matching degree between the aligned synthetic seismic record and the actual seismic record, the distance metric function, the constraint of the optimal path, and the size of the time window are adjusted, and steps S2-S5 are repeated until the adjustment range is less than the preset threshold or the preset number of iterations is reached.

10. A system for calibrating synthetic seismic records, employing the synthetic seismic record calibration method according to any one of claims 1-9, characterized in that, The system includes: The seismic record acquisition module is used to acquire well logging curves, select a wavelet model to create synthetic seismic records, and acquire the actual seismic records corresponding to the positions of the well logging curves. The cost matrix building module is used to extract the first original time series corresponding to the synthetic seismic record and the second original time series corresponding to the actual seismic record, and to create a cost matrix based on the first original time series and the second original time series. The cumulative cost calculation module is used to define a distance metric function based on dynamic time planning, calculate the cost of each element in the cost matrix through the distance metric function, and update the cumulative cost of the corresponding element by combining the cumulative cost of the element's predecessor position. The optimal path finding module is used to trace backwards from the element at the end of the cost matrix to find the path with the minimum cumulative cost as the optimal path; The synthetic seismic record calibration module is used to adjust the synthetic seismic record according to the found optimal path until the adjusted synthetic seismic record is aligned with the actual seismic record on the time axis, thus obtaining the calibrated synthetic seismic record.

Images