Reservoir classification method and device based on logging curve
By calculating the pattern distance, morphological distance and structural distance of the logging curve and determining the similarity in combination with these distances, the problem of difficulty in effectively utilizing logging curve information in the prior art is solved, and accurate reservoir identification and classification are achieved.
Patent Information
- Application Number
- CN202410348848.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-26
- Publication Date
- 2025-06-24
AI Technical Summary
The prior art is difficult to effectively utilize the local and overall information of the logging curve in the interpretation of petroleum logging treatment, resulting in the inability to accurately identify and classify similar reservoirs.
By pre-processing and converting the logging curve into a string, the pattern distance, morphological distance and structural distance are calculated, combining these distances to determine the similarity of the logging curves, identifying and classifying reservoirs.
This method can effectively utilize the local and overall information of the logging curve to accurately identify and classify reservoirs, solving the classification errors caused by the lack of unified parameters and logging curve drift in the prior art.
Smart Images
Figure CN120197044A_ABST
Abstract
Description
Technical Field
[0001] This document relates to the technical field of oil logging processing and interpretation, and particularly to a reservoir classification method and device based on logging curves. Background Art
[0002] In the field of oil logging processing and interpretation, the clustering method represents the formation sequence as features without sequence properties and then classifies these features. The commonly used method in current logging is to calculate physical property parameters based on logging curves and then use the K-means method for clustering. This kind of method has two disadvantages. One is that this method generally takes a single well as the processing unit. If you want to know which reservoirs in a certain formation of all wells are similar, all wells must have the same parameters as the measurement standard. In engineering applications, there are often some logging curves without normal data, resulting in these wells not being able to participate in the classification. The other is that this method measures the distance based on sampling points and uses local information. If you want to determine whether the reservoirs of two formations are similar, the existing K-means clustering method based on physical property parameter measurement is not applicable.
[0003] Therefore, in order to determine similar reservoirs in the same formation in the work area, it is an urgent problem to implement a method that can comprehensively utilize local information and global information. Summary of the Invention
[0004] In view of the above problems, a reservoir classification method and device based on logging curves are proposed. This method can comprehensively utilize the local information and global information of logging curves to identify reservoirs and classify reservoirs.
[0005] In a first aspect, the present application provides a reservoir classification method based on logging curves, and the method includes:
[0006] Preprocess the logging curves and convert the processed curves into strings;
[0007] According to the string, calculate the pattern distance, shape distance, and structure distance respectively;
[0008] Determine the numerical value of the similarity of the logging curves according to the pattern distance, shape distance, and structure distance;
[0009] Determine the category of the reservoir according to the numerical value of the similarity.
[0010] In a second aspect, an embodiment of the present invention also provides a reservoir classification device based on logging curves. The device includes: a memory and a processor; the memory is used to save the program for performing reservoir classification based on logging curves, and the processor is used to read and execute the program for performing reservoir classification based on logging curves to execute the method according to any one of the above embodiments.
[0011] In a third aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a data processing program is stored, and the data processing program is executed by a processor to perform the reservoir classification method based on logging curves according to any one of the above embodiments.
[0012] Compared with the related art, the present application provides a reservoir classification method and device based on logging curves. The method includes: preprocessing the logging curves and converting the processed curves into strings; calculating the pattern distance, morphological distance, and structural distance respectively according to the strings; determining the numerical value of the similarity of the logging curves according to the pattern distance, morphological distance, and structural distance; and determining the category of the reservoir according to the numerical value of the similarity. The present application can comprehensively utilize the local information and overall information of the logging curves to identify the reservoir and classify the reservoir.
[0013] Other features and advantages of the present application will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by implementing the present application. Other advantages of the present application can be realized and obtained through the solutions described in the specification and the drawings. Description of the Drawings
[0014] The drawings are used to provide an understanding of the technical solutions of the present application, and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the technical solutions of the present application and do not constitute a limitation to the technical solutions of the present application.
[0015] Figure 1 It is a flowchart of the reservoir classification method based on logging curves according to an embodiment of the present application;
[0016] Figure 2 It is a schematic diagram of the reservoir classification device based on logging curves according to an embodiment of the present application;
[0017] Figure 3 It is a schematic diagram of the morphological distance calculation method in some exemplary embodiments;
[0018] Figure 4 It is a schematic diagram of the planar distribution range of salt layers in some exemplary embodiments;
[0019] Figure 5 It is a schematic diagram of the statistical chart structure times in some exemplary embodiments;
[0020] Figure 6 It is a schematic diagram of extracting key points of logging curves in some exemplary embodiments;
[0021] Figure 7 It is a schematic diagram of calculating the morphological distance by DTW in some exemplary embodiments;
[0022] Figure 8 Schematic diagram of a similarity matrix obtained based on pattern distance, morphological distance, and structural distance in some exemplary embodiments;
[0023] Figure 9 Schematic diagram of visualizing the clustering results of the AP algorithm for logging curves in some exemplary embodiments;
[0024] Figure 10 Flowchart of a reservoir classification method based on logging curves in some exemplary embodiments;
[0025] Figure 11 Schematic diagram of clustering results in some exemplary embodiments. Detailed implementation manners
[0026] This application describes multiple embodiments, but the description is exemplary rather than restrictive, and it will be obvious to those of ordinary skill in the art that there can be more embodiments and implementation solutions within the scope of the embodiments described in this application. Although many possible feature combinations are shown in the drawings and discussed in the detailed implementation manners, many other combination ways of the disclosed features are also possible. Unless specifically restricted, any feature or element of any embodiment can be combined with any other feature or element in any other embodiment, or can replace any other feature or element in any other embodiment.
[0027] This application includes and contemplates combinations with features and elements known to those of ordinary skill in the art. The disclosed embodiments, features, and elements of this application can also be combined with any conventional features or elements to form unique inventive solutions defined by the claims. Any feature or element of any embodiment can also be combined with features or elements from other inventive solutions to form another unique inventive solution defined by the claims. Therefore, it should be understood that any feature shown and / or discussed in this application can be implemented alone or in any suitable combination. Therefore, the embodiments are not subject to other restrictions except those made according to the appended claims and their equivalent replacements. In addition, various modifications and changes can be made within the scope of protection of the appended claims.
[0028] In addition, when describing representative embodiments, the specification may have presented the method and / or process as a specific sequence of steps. However, to the extent that the method or process does not depend on the specific order of the steps described herein, the method or process should not be limited to the specific order of steps described. As those of ordinary skill in the art will understand, other step orders are possible. Therefore, the specific order of steps set forth in the specification should not be construed as a limitation on the claims. In addition, the claims directed to the method and / or process should not be limited to performing their steps in the order written, as those skilled in the art can readily understand that these orders can vary and still remain within the spirit and scope of the embodiments of the present application.
[0029] An embodiment of the present invention provides a reservoir classification method based on logging curves, as Figure 1 shown, the method includes steps S100 - S130:
[0030] S100: Preprocess the logging curves and convert the processed curves into strings;
[0031] S110: Calculate the pattern distance, morphological distance, and structural distance respectively according to the string;
[0032] S120: Determine the numerical value of the similarity of the logging curves according to the pattern distance, morphological distance, and structural distance;
[0033] S130: Determine the category of the reservoir according to the numerical value of the similarity.
[0034] In an exemplary embodiment, the preprocessing of the logging curves includes: data cleaning, data normalization, and abnormal data detection and processing. In this step, the specific implementation processes of data cleaning, data normalization, and abnormal data detection and processing are as follows:
[0035] ① Data preprocessing: Clean the data, remove obvious errors or damaged records (e.g., invalid values such as -9999), and correct the drift and calibration errors of the logging instrument.
[0036] ② Data normalization: For example, take the logarithm of the resistivity logging curve Rt to facilitate subsequent algorithm processing of the data.
[0037] ③ Abnormal data detection and processing: Apply the interquartile range (IQR) method to screen data points outside 4 times the interquartile range, and interpolate and replace them with surrounding data. Of course, other suitable algorithms and parameters can be selected according to the characteristics of the data in the work area.
[0038] In an exemplary embodiment, the process of converting the processed curve into a string can be as follows: Convert the processed curve into a string according to the time series symbolic aggregation approximation method. The converted string, such asFigure 3 as shown
[0039] In an exemplary embodiment, the calculation process of the pattern distance is as follows:
[0040] Step 1: After converting the processed curve into a string, calculate the minimum edit distance between the strings of any two wells respectively;
[0041] In this embodiment, for the two wells to be compared, determine the common curves related to reservoir division of the two wells: gamma ray GR and resistivity Rt. Taking the gamma ray GR curve as an example, the curve is reduced to a string by the symbolic aggregate approximation method for time series (SAX), and the minimum edit distance ED between the strings can be calculated by the minimum edit distance ED method. In this embodiment, the minimum edit distance ED method is a common method for measuring the difference between strings. For example, for string A = "abd" and B = "acd", the edit distance is 1, and at least one edit is required to make them equal.
[0042] If for these two wells, the common curves related to reservoir division of the two wells are determined to be gamma ray GR and resistivity Rt. The minimum edit distance of the resistivity Rt needs to be calculated again using the same method.
[0043] Step 2: Take the minimum edit distance as the pattern distance of the curve.
[0044] In this step, for the two wells to be compared, if the common curves related to reservoir division of the two wells are gamma ray GR and resistivity Rt, after calculating the minimum edit distance of the gamma ray GR log curve and the minimum edit distance of the resistivity Rt log curve according to the above steps, calculate the average value of the minimum edit distance of the gamma ray GR log curve and the minimum edit distance of the resistivity Rt log curve as the final minimum edit distance.
[0045] In an exemplary embodiment, the calculation process of the morphological distance is as follows:
[0046] Step 1: For any two wells, determine the common curves related to reservoir division of the two wells;
[0047] Step 2: For each pair of common curves, use the dynamic time warping method (DTW) to calculate the minimum cost value for aligning the first i points of the first curve and the first j points of the second curve;
[0048] Step 3: Take the average value of the minimum cost values of multiple pairs of common curves as the morphological distance between any two wells.
[0049] In an exemplary embodiment, the dynamic time warping (DTW) method is:
[0050]
[0051] The specific implementation of the dynamic time warping (DTW) method is as follows. Figure 4 As shown, find an alignment method to align the cost between the upper curve and the lower curve in Figure 4 . In this method, the morphological distance of the logging curve is defined as the minimum cost required to align the two curves.
[0052] In an exemplary embodiment, the calculation process of the structural distance is as follows:
[0053] First step: For any two wells, determine the common curves related to reservoir division of the two wells;
[0054] Second step: Establish corresponding graph structure feature vectors according to each pair of curves;
[0055] Third step: Calculate the Manhattan distance between the graph structure feature vectors;
[0056] Count Figure 5 The number of occurrences of the 6 graph structures in the logging curve of another well. For example, the statistical result is (19, 2, 4, 5, 1, 6). The meaning of the statistical result is as follows: Figure 5 The 6 graph structures in appear 19 times, 2 times, 4 times, 5 times, 1 time, and 6 times in sequence in the logging curve of another well. The same method is used for statistical analysis of another well for comparison. For example, the statistical result is (10, 3, 2, 5, 1, 4).
[0057] According to the above statistical results, calculate the structural distance between the logging curves of these two wells as: abs(19 - 10) + abs(2 - 3) + abs(4 - 2) + abs(5 - 5) + abs(1 - 1) + abs(6 - 4) = 14, that is, |19 - 10| + |2 - 3| + |4 - 2| + |5 - 5| + |1 - 1| + |6 - 4| = 14.
[0058] Fourth step: Take the Manhattan distance as the structural distance.
[0059] In an exemplary embodiment, establishing corresponding graph structure feature vectors according to each pair of curves includes:
[0060] First step: Use the Douglas - Peucker algorithm to calculate the key points of each pair of common curves respectively;
[0061] In this step, the Douglas - Peucker algorithm can retain the key morphology of the curve, extract the key points of the curve, and is an algorithm that ignores details. Calculate the key points of the logging curve. The calculation method of the key points is as shown in the appendix Figure 5 . The Douglas - Peucker algorithm can be used.
[0062] As shown in Figure 6 After extracting the key points of the logging curve from the original logging curve using the Douglas - Peucker algorithm, the curve is as shown in Figure 6 the dashed curve in. The points in the dashed curve represent the extracted key points. While retaining the curve shape, the dashed curve simplifies the curve, facilitating subsequent calculations.
[0063] Step 2: For each well, select 4 adjacent key points from the key points, move 2 of the 4 adjacent key points each time, and count the number of occurrences of each graph structure;
[0064] Figure 5 The first column in is the number of the graph structure, the second column is the graph structure, and a connection between two points indicates that the amplitude levels of these two points are visible. The third column represents: If the amplitudes of 4 adjacent key points satisfy this inequality relationship, the relationship of these 4 points is mapped to the graph structure in this row.
[0065] Step 3: Establish a graph - structure feature vector based on the number of occurrences of each graph structure; for example: The number of occurrences of 6 graph structures is concatenated to form the feature vector (19, 2, 4, 5, 1, 6).
[0066] Count Figure 5 the number of occurrences of the 6 graph structures in the logging curve of another well. For example: The statistical result is (19, 2, 4, 5, 1, 6), indicating that Figure 5 the 6 graph structures in have occurred 19 times, 2 times, 4 times, 5 times, 1 time, and 6 times in sequence in the logging curve of another well. The same method is used to count for another well being compared. For example, the statistical result is (10, 3, 2, 5, 1, 4).
[0067] In an exemplary embodiment, the method for determining the similarity value of logging curves based on the pattern distance, shape distance, and structure distance includes:
[0068] Step 1: Sum the pattern distance, the shape distance, and the structure distance to determine the total distance;
[0069] Step 2: Take the opposite of the total distance to determine the value of the total similarity between two wells.
[0070] In an exemplary embodiment, the method for determining the reservoir category based on the similarity value includes:
[0071] Step 1: Form a similarity matrix with the similarity between any two wells;
[0072] Step 2: Cluster the similarity matrix using the affinity propagation algorithm to obtain the clustering result;
[0073] In this step, the Affinity Propagation (AP) algorithm is a clustering algorithm based on "information passing" between data points, which aims to identify cluster centers (referred to as "exemplar points") and corresponding clusters by considering the "similarity" between data points.
[0074] In this step, the specific implementation process of clustering is as follows:
[0075] Step 1. Initialize similarity: The similarity matrix is generally constructed based on negative Euclidean distance or other metrics, representing the degree of similarity between data points. In this step, the similarity calculated in the previous step is initialized.
[0076] Step 2. Update responsibility: Update the responsibility of each pair of data points according to the current membership and similarity matrix.
[0077] Step 3. Update membership: Update the membership according to the current responsibility.
[0078] Step 4. Iterative update: Continuously iterate to update the responsibility and membership until the selection of cluster centers converges (i.e., the changes in responsibility and membership are less than a certain threshold, or the preset number of iterations is reached).
[0079] Step 5. Select cluster centers: The cluster centers are composed of data points with the highest sum of membership and responsibility.
[0080] Step 6. Assign clusters: Each data point is assigned to the nearest cluster center to form the final clustering result; as Figure 11 shown, the schematic diagram of the final clustering result.
[0081] Step 3. Determine the category of the reservoir according to the clustering result.
[0082] Combined with the following problems existing in the actual application of ordinary clustering based on physical property parameters, this embodiment has the following technical effects:
[0083] First: Due to the different qualities of well logging data of each well, it is impossible to establish a unified physical property parameter feature vector. For example, the well logging curves related to permeability of some wells are of poor quality, resulting in the inability to calculate permeability. Then, either these wells do not participate in clustering, or the permeability cannot be included in the feature vector.
[0084] Second: Using the same physical property parameters as the feature vector for different wells will be affected by the overall drift of the parameters, resulting in a large error in the results.
[0085] Based on the above problems, the present application determines the reservoir similarity between two wells from the similarities in the pattern distance, alignment cost, and structural relationship of the original logging curves, eliminating the calculation of physical property parameters and avoiding the influence caused by the overall drift of the logging curves. Moreover, since the similarity is calculated pairwise, the problem of missing logging curves can be effectively addressed. When calculating, only the curves that both wells have are taken, and there is no need to unify the parameters.
[0086] In a second aspect, an embodiment of the present invention further provides a reservoir classification device based on logging curves, as Figure 2 shown. The device includes: a memory 200 and a processor 210; the memory is used to store a program for performing reservoir classification based on logging curves, and the processor is used to read and execute the program for performing reservoir classification based on logging curves, and execute the method described in any one of the above embodiments.
[0087] In a third aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a data processing program is stored, and the data processing program is executed by a processor to perform the reservoir classification method based on logging curves described in any one of the above embodiments.
[0088] Example 1
[0089] This example shows the method flow for classifying reservoirs through logging curves, as Figure 10 shown. The specific steps are as follows:
[0090] Step 1: Input logging curves;
[0091] Step 2: Identify outliers and normalize the data;
[0092] In step S2, outliers are removed and the data is normalized. To remove outliers, the identification of outliers needs to be provided. Normalization mainly changes the absolute values into relative values, removes the influence of drift, and performs different preprocessing on different logging curves to make the logging response closer to linearity.
[0093] Step 3: Calculate the pattern distance using ED;
[0094] Calculate the pattern distance between any two wells through their logging curves.
[0095] In step S3, the specific process of calculating the pattern distance is as follows:
[0096] First, change the trends of the logging curves of the two wells into letter sequences;
[0097] Then, use the minimum edit distance to find the pattern distance.
[0098] In this step, if there are more than two common curves between two wells, the minimum edit distance needs to be calculated for each common curve separately; after calculating the minimum edit distance of each common curve, the average value of the minimum edit distances of multiple common curves is taken as the pattern distance.
[0099] Step 4: Calculate the morphological distance using DTW.
[0100] In step S4, calculate the morphological distance, extract the key points of the logging curves, calculate the alignment cost using the DTW algorithm, and calculate the morphological distance; as Figure 7 shown, the calculation process of calculating the morphological distance using DTW.
[0101] Step 5: Calculate the structural distance.
[0102] In step S5, the process of calculating the structural distance includes:
[0103] S51: Use the Douglas-Peucker algorithm to calculate the key points of each pair of common curves respectively;
[0104] S52: For each well, from the key points, for 4 adjacent key points on this curve, move 2 points each time, and count the number of occurrences of each graph structure respectively;
[0105] S53: Establish a graph structure feature vector according to the number of occurrences of each graph structure;
[0106] S54: Calculate the Manhattan distance between the graph structure feature vectors;
[0107] S55: Take the Manhattan distance as the structural distance.
[0108] Step 6: Add the pattern distance, morphological distance, and structural distance as the total distance, take the opposite number of the total distance as the total similarity, and the total similarities between two wells form a similarity matrix.
[0109] In step S6, add the pattern distance, morphological distance, and structural distance and then take the opposite number to get the similarity, and the similarities between two wells form a similarity matrix. As Figure 8 shown, obtain the similarity matrix according to the pattern distance, morphological distance, and structural distance.
[0110] Step 7: Take the similarity matrix as the input and perform clustering using AP (Affinity Propagation algorithm);
[0111] In step S7, take the similarity matrix as the input and perform clustering using the AP algorithm (Affinity Propagation algorithm) and output the result. As Figure 9 shown, the visualization graph of the clustering result of the logging curve AP algorithm, with the first category on the left and the second category on the right.
[0112] Those of ordinary skill in the art can understand that all or some of the steps in the methods disclosed above, and the functional modules / units in systems and devices, can be implemented as software, firmware, hardware, and their appropriate combinations. In the hardware implementation, the division between the functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, one physical component can have multiple functions, or one function or step can be executed by several physical components in cooperation. Some components or all components can be implemented as software executed by a processor, such as a digital signal processor or a microprocessor, or be implemented as hardware, or be implemented as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include a computer storage medium (or non-transitory medium) and a communication medium (or transitory medium). As is well known to those of ordinary skill in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data. Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those of ordinary skill in the art that a communication medium typically contains computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and can include any information delivery medium.
Claims
1. A reservoir classification method based on well logging curves, characterized in that: The method comprises: Preprocess the logging curve and convert the processed curve into a character string; According to the character string, the pattern distance, the morphological distance and the structural distance are calculated respectively; Determining a similarity value of a well logging curve according to the pattern distance, the morphological distance and the structural distance; The category of the reservoir is determined according to the similarity value.
2. The reservoir classification method based on well logging curves according to claim 1, characterized in that: The well logging curve is preprocessed, including: Cleaning data, data normalization, and abnormal data detection and processing.
3. The reservoir classification method based on well logging curves according to claim 2 is characterized in that: Convert the processed curve into a string: Convert the processed curve into a string using the time series symbol aggregation approximation method.
4. The reservoir classification method based on well logging curves according to claim 3 is characterized in that: The calculation process of the pattern distance is: After converting the processed curve into a string, the minimum edit distance of the string between any two wells is calculated; The minimum edit distance is taken as the pattern distance of the curve.
5. The reservoir classification method based on well logging curves according to claim 1, characterized in that: The calculation process of the morphological distance is: For any two wells, determine the common curve between the two wells that is relevant to reservoir division; For each pair of common curves, the minimum cost value of each pair of common curves is calculated using the dynamic time warping method; The average of the minimum cost values of multiple pairs of common curves is taken as the morphological distance between any two wells.
6. The reservoir classification method based on well logging curves according to claim 5, characterized in that: The dynamic time warping method is: In the above formula, DTW(i, j) represents the minimum cost value of the Q curve and the S curve, i represents the sampling point of the Q curve, j represents the sampling point of the S curve, Q[i] represents the amplitude of the i-th sampling point of the Q curve, and S[j] represents the amplitude of the j-th point.
7. The reservoir classification method based on well logging curves according to claim 1, characterized in that: The calculation process of the structural distance is: For any two wells, determine the common curves of the two wells related to reservoir division; According to each pair of curves, a corresponding graph structure feature vector is established; Calculating the Manhattan distance between the feature vectors of the graph structure; The Manhattan distance is taken as the structural distance.
8. The reservoir classification method based on well logging curves according to claim 7 is characterized in that: The step of establishing a corresponding graph structure feature vector according to each pair of curves includes: The Douglas Peucker algorithm is used to calculate the key points of each pair of common curves separately; For each well, move 2 points at a time from the 4 adjacent key points in the curve, and count the number of times each graph structure appears; A graph structure feature vector is established according to the number of times each graph structure appears.
9. The reservoir classification method based on well logging curves according to claim 1, characterized in that: Determining the numerical value of the similarity of the logging curves according to the pattern distance, the morphological distance and the structural distance includes: Determine a total distance by summing the pattern distance, the morphological distance and the structural distance; The total distance is negated to determine the value of the total similarity between the two wells.
10. The reservoir classification method based on well logging curves according to claim 9, characterized in that: Determining the category of the reservoir according to the numerical value of the similarity includes: The similarities between any two wells form a similarity matrix; Clustering the similarity matrix using an affinity propagation algorithm to obtain a clustering result; The category of the reservoir is determined according to the clustering result.
11. A reservoir classification device based on well logging curves, characterized in that: The device comprises: a memory and a processor; the memory is used to store a program for performing reservoir classification based on well logging curves, and the processor is used to read and execute the program for performing reservoir classification based on well logging curves, and execute the method described in any one of claims 1-10.
12. A computer-readable storage medium having a data processing program stored thereon, wherein the data processing program is executed by a processor to implement the reservoir classification method based on well logging curves as described in any one of claims 1 to 10.