Method, system and equipment for identifying thin sand layer and medium

Abstract

The invention provides a method, a system, equipment and a medium for identifying a thin sand layer, overcomes the defect of low identification precision caused by the fact that the resolution of a natural gamma curve is smaller than the thickness of the thin sand layer when a conventional single well is used for identifying the thin sand layer, and overcomes the defect that the geological constraint of an adjacent well is not considered when the thin sand layer is identified only on the single well in the prior art. According to the thin sand layer identification method, the natural gamma resolution and adjacent well geological information constraint comparison are improved based on the deconvolution model, the good use effect is achieved, the workload of manual thin sand layer identification is reduced, the thin sand layer identification efficiency is improved, resources needed by calculation are low, and the method has the good popularization advantage.

Description

Technical Field

[0001] The present invention belongs to the technical field of oil and gas exploration and development, specifically to the technology of fine processing of logging data, and particularly relates to a method, system, device and medium for identifying thin sand layers. Background Art

[0002] With the improvement of the development degree of onshore complex oil and gas reservoirs, the requirements for refined geological research are increasing day by day. The thin sand layer reservoir is to find the reservoir under the extreme condition of "finding oil in thin and poor layers". The conventional logging curve stratification ability can only reach about 0.6m, which has become the "bottleneck" restricting the fine interpretation and precise development of thin sand layers in oilfields. Therefore, it is urgent to solve the problems such as improving the identification accuracy of thin sand layer reservoirs, characterizing the internal heterogeneity of oil layers, the uneven vertical distribution of oil and gas, and description. One of the keys to solving these problems is to obtain logging information with high vertical resolution; onshore old oilfields have a long development history, and the early logging data are difficult to meet the current needs of fine geological research, which seriously restricts the in-depth research on the geological characteristics of oil reservoirs and the formulation of development adjustment measures in the later stage of development.

[0003] Therefore, there is an urgent need for a method for high-resolution processing of logging curves and identifying thin sand layers under the constraint of adjacent well information of thin sand layers to solve the above problems. Summary of the Invention

[0004] Aiming at the problems existing in the prior art, the present invention provides a method, system, device and medium for identifying thin sand layers to achieve accurate division of thin sand layers.

[0005] The present invention is realized through the following technical solutions:

[0006] A method for identifying thin sand layers includes the following steps:

[0007] S1: Respectively extract the natural gamma curves of the target well and the comparison well and perform deconvolution processing to obtain deconvolution-generated curves;

[0008] S2: Obtain a difference curve and a preliminary division result of thin sand layers based on the natural gamma curve and the deconvolution-generated curve;

[0009] S3: Based on the Euclidean distance of the difference curves of the target well and the comparison well, use the dynamic time warping algorithm to match the deconvolution matching results between the target well and the comparison well;

[0010] S4: Obtain the identification result of thin sand layers of the target well based on the deconvolution matching result and the preliminary division result.

[0011] Further, summarize the well coordinate information of the well to be identified to form a triple data set, and select the target well and the comparison well within the triple data set.

[0012] Further, the triple dataset formed by summarizing the well coordinate information of the wells to be recognized is:

[0013] S = {(x1, y1, w1), (x2, y2, w2),..., (x i , y i , w i ),..., (x N , y N , w N )};

[0014] Among them, w i represents the well name of the i-th well; x i represents the abscissa of the i-th well; y i represents the ordinate of the i-th well; (x i , y i , w i ) is a triple composed of well coordinates and well names, N represents the number of wells in the dataset, and S represents the dataset.

[0015] Further, the target well is randomly selected and located at the edge of the triple dataset.

[0016] Further, the target well is the well that selects the nearest neighbor of the target well through traversal calculation.

[0017] Further, the process of selecting the well nearest to the target well as the comparison well through traversal calculation is:

[0018] Assume that the selected well w i is used as the target well, then the nearest neighbor well w j is found by traversing the dataset S through the Euclidean distance, then:

[0019]

[0020] Among them, w j is the nearest neighbor comparison well for solving the target well w i , and the argmin function returns the well coordinates that satisfy the minimum Euclidean distance in the set.

[0021] Further, the process of separately extracting the natural gamma curves of the target well and the comparison well and performing deconvolution processing to obtain the deconvolution-generated curve is:

[0022] The natural gamma curves GR i and GR j of the target well and the comparison well are respectively:

[0023]

[0024]

[0025] Among them, gr i p is the data of the p-th sampling point of the target well, P is the number of sampling points of the target well, and gr i q is the data of the q-th sampling point of the comparison well, and Q is the number of sampling points of the comparison well;

[0026] Perform deconvolution operation on the natural gamma curve. The deconvolution calculation results of a single sampling point of the target well and the comparison well, gr i p* and gr i q* are as follows:

[0027]

[0028] Among them, α = [α1, α2,..., α k is the deconvolution operator, k is the number of deconvolution operators, and k is taken as an odd number;

[0029] Then the deconvolution results of the target well and the comparison well and are:

[0030]

[0031] Since the deconvolution operation will include the first sampling points, the data of the first and last sampling points of the deconvolution result are set to 0.

[0032] Furthermore, the process of obtaining the difference curve and the preliminary division result of the thin sand layer based on the natural gamma curve and the deconvolution-generated curve is as follows:

[0033] Generate a difference curve from the difference between the natural gamma curve and the deconvolution-generated curve. The positive part of the difference curve is used as the preliminary division result of the thin sand layer, which represents the developed section of the thin sand layer.

[0034] Furthermore, the process of generating a difference curve from the difference between the natural gamma curve and the deconvolution-generated curve is as follows:

[0035]

[0036] Among them, is the difference curve of the target well, is the difference curve of the comparison well;

[0037] The positive part of the difference curve is used as the preliminary division result of the thin sand layer, which represents the developed section of the thin sand layer. Then, the part of the difference curve greater than zero is assigned a value of 1, and the part less than or equal to zero is assigned a value of 0. The part of the difference curve equal to 1 is used as the preliminary division result of the thin sand layer. Assume that the depth range of the thin sand layer is

[0038] Further, the process of using the dynamic time warping algorithm to match the deconvolution matching results between the target well and the comparison well is as follows:

[0039] Preset a matrix grid P of P×Q. The matrix element (p, q) is the Euclidean distance D between the deconvolution result of the target well and the deconvolution result of the comparison well in the difference curve pq Among the paths from the starting point P[1][1] to the ending point P[P][Q], search for the path L with the smallest sum of matrix elements min (p, q);

[0040] The search rule is solved by a recursive algorithm, and the process is as follows:

[0041] Initial condition:

[0042] L min (1, 1) = P[1][1]

[0043] Recursive rule:

[0044] L min (p, q) = min(L min (p, q - 1), L min (p - 1, q), L min (p - 1, q - 1)}

[0045] + P[p][q]

[0046] L min (p, q) represents the shortest path from the starting point P[1][1] to the matrix element P[p][q];

[0047] L min (p, q) in the matching p and q are the interpreted thin sand layer results of the target well and the comparison well. Assume that the depth range of the sampling points matched by the target well is Then the identified depth of the thin sand layer in the target well is:

[0048]

[0049]

[0050] Among them, is the top depth of the thin sand layer, is the bottom depth of the thin sand layer.​

[0051] Further, the process of obtaining the thin sand layer identification result of the target well based on the deconvolution matching result and the preliminary division result is as follows:

[0052] Calculate the average value of the deconvolution matching result and the preliminary division result, and obtain the thin sand layer identification result of the target well based on the average value.

[0053] Further, the process of calculating the average value of the deconvolution matching result and the preliminary division result and obtaining the thin sand layer identification result of the target well based on the average value is as follows:

[0054] Calibrate all wells in the dataset as target wells and match the corresponding deconvolution identification results, which are used as the initial division results;

[0055] Randomly select another well as the target well again and repeat the comparison of adjacent well information. At this time, the initial division result is the division result after the previous round of iteration. Now assume that after z iterations, the error e between the thin sand layer identification result and the result of the previous round of iteration z is within the preset range [-ε, ε]. It is considered that the thin sand layer identification result no longer changes, and the algorithm converges at this time. Otherwise, perform the (z + 1)-th iteration until the algorithm converges. The division scheme at the time of convergence is the thin sand layer identification result;

[0056] Among them, the judgment in the z-th iteration is:

[0057]

[0058] Among them, e z represents the root mean square error of the thin sand layer identification results after the z-th iteration and the (z - 1)-th iteration; ε represents the minimum error value, which is an empirical parameter; z represents the current number of iterations; Z max represents the maximum number of iterations allowed by the method, which is an empirical parameter; R represents the final thin sand layer identification result; r z represents the thin sand layer identification result after the z-th iteration.

[0059] The first processing module is configured to separately extract the natural gamma curves of the target well and the comparison well and perform deconvolution processing to obtain deconvolution-generated curves;

[0060] The second processing module is configured to obtain difference curves and preliminary thin sand layer division results based on the natural gamma curves and the deconvolution-generated curves;

[0061] The third processing module is configured to match the deconvolution matching results between the target well and the comparison well by using the dynamic time warping algorithm based on the Euclidean distance of the difference curves of the target well and the comparison well;

[0062] An output module, configured to obtain the identification result of thin sand layers of a target well based on the deconvolution matching result and the preliminary division result.

[0063] A computer device includes 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 a method for identifying thin sand layers are implemented.

[0064] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of a method for identifying thin sand layers are implemented.

[0065] Compared with the prior art, the present invention has the following beneficial technical effects:

[0066] The present invention provides a method, system, device, and medium for identifying thin sand layers, including the following steps: respectively extracting the natural gamma curves of a target well and a comparison well and performing deconvolution processing to obtain deconvolution-generated curves; obtaining a difference curve and a preliminary division result of thin sand layers based on the natural gamma curve and the deconvolution-generated curve; using the dynamic time warping algorithm to match the deconvolution matching results between the target well and the comparison well based on the Euclidean distance of the difference curves of the target well and the comparison well; obtaining the identification result of thin sand layers of the target well based on the deconvolution matching result and the preliminary division result; this application supplements the deficiency that the resolution of the natural gamma curve is less than the thickness of the thin sand layer when identifying thin sand layers in a conventional single well, resulting in low identification accuracy. At the same time, compared with the defect that only the geological constraints of adjacent wells are not considered when identifying thin sand layers in a single well in the prior art, a method for identifying thin sand layers based on improving the natural gamma resolution and comparing adjacent well geological information constraints using a deconvolution model is proposed, obtaining good usage effects, reducing the workload of manually identifying thin sand layers, improving the efficiency of identifying thin sand layers, and requiring low computing resources, having good promotion advantages. Description of the Drawings

[0067] Figure 1 Shows a flowchart of a method for identifying thin sand layers according to an embodiment of the present invention;

[0068] Figure 2 Shows another flowchart of a method for identifying thin sand layers according to an embodiment of the present invention;

[0069] Figure 3 Shows a schematic diagram of the single-well GR deconvolution result and the difference curve indicating thin sand layers according to an embodiment of the present invention;

[0070] Figure 4 Shows a schematic diagram of the multi-well thin sand layer division result connecting wells according to an embodiment of the present invention. Detailed Embodiments

[0071] The present invention will be further described in detail below in conjunction with specific embodiments, which are explanations of the present invention rather than limitations.

[0072] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.

[0073] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0074] The present invention provides a method for identifying thin sand layers, as Figure 1 shown, including the following steps:

[0075] S1: Respectively extract the natural gamma curves of the target well and the comparison well and perform deconvolution processing to obtain the deconvolution-generated curves;

[0076] S2: Obtain the difference curve and the preliminary division result of the thin sand layer based on the natural gamma curve and the deconvolution-generated curve;

[0077] S3: Based on the Euclidean distance of the difference curves of the target well and the comparison well, use the dynamic time warping algorithm to match the deconvolution matching results between the target well and the comparison well;

[0078] S4: Obtain the identification result of the thin sand layer of the target well based on the deconvolution matching result and the preliminary division result.

[0079] Furthermore, this embodiment also provides step S6: Repeat steps S1-S4 to obtain the identification and division of the thin sand layers of all wells in the well data set.

[0080] Preferably, in this embodiment, the well coordinate information of the well to be identified is summarized to form a triple data set, and the target well and the comparison well are selected within the triple data set.

[0081] Specifically, the triple dataset formed by summarizing the well coordinate information of the wells to be recognized is as follows:

[0082] S = {(x1, y1, w1), (x2, y2, w2),..., (x i , y i , w i ),..., (x N , y N , w N )};

[0083] Among them, w i represents the well name of the i-th well; x i represents the abscissa of the i-th well; y i represents the ordinate of the i-th well; (x i , y i , w i ) is a triple composed of well coordinates and well names, N represents the number of wells in the dataset, and S represents the dataset.

[0084] Preferably, in the embodiment of the present application, the target well is randomly selected and located at the edge of the triple dataset; the target well is to select the well closest to the target well through traversal calculation.

[0085] Specifically, the process of selecting the well closest to the target well as the comparison well through traversal calculation is as follows:

[0086] Assume that the selected well w i is used as the target well, then the nearest neighbor well w i is found by traversing the dataset S through the Euclidean distance, then:

[0087]

[0088] Among them, w j is the nearest neighbor comparison well for solving the target well w i , and the argmin function returns the well coordinates that satisfy the minimum Euclidean distance in the set.

[0089] Preferably, the process of separately extracting the natural gamma curves of the target well and the comparison well and performing deconvolution processing to obtain the deconvolution-generated curve is as follows:

[0090] The natural gamma curves GR i and GR j of the target well and the comparison well are respectively:

[0091]

[0092] Among them, gr i pData of the p-th sampling point of the target well, P is the number of sampling points of the target well, gr i q Data of the q-th sampling point of the comparison well, Q is the number of sampling points of the comparison well;

[0093] Perform deconvolution operation on the natural gamma curve, and the deconvolution calculation results gr i p* and gr i q* are as follows:

[0094]

[0095] Among them, α = [α1, α2,..., α k is the deconvolution operator, k is the number of deconvolution operators, and k is an odd number;

[0096] Then the deconvolution results of the target well and the comparison well and are:

[0097]

[0098] Since the deconvolution operation will include the first sampling points, the data of the first and last sampling points of the deconvolution result are set to 0.

[0099] Preferably, in this embodiment, the process of obtaining the difference curve and the preliminary division result of thin sand layers based on the natural gamma curve and the deconvolution-generated curve is as follows:

[0100] Generate a difference curve from the difference between the natural gamma curve and the deconvolution-generated curve. The positive part of the difference curve is used as the preliminary division result of thin sand layers, which represents the developed section of thin sand layers.

[0101] Specifically, the process of generating a difference curve from the difference between the natural gamma curve and the deconvolution-generated curve is as follows:

[0102]

[0103] Among them, is the difference curve of the target well, is the difference curve of the comparison well;

[0104] The positive part of the difference curve is used as the preliminary division result of thin sand layers, which represents the developed section of thin sand layers. Then, the part greater than zero in the difference curve is assigned a value of 1, the part less than or equal to zero is assigned a value of 0, and the part equal to 1 in the difference curve is used as the preliminary division result of thin sand layers. Assume that the depth range of thin sand layers is

[0105] Preferably, the process of using the dynamic time warping algorithm to match the deconvolution matching results between the target well and the comparison well is as follows:

[0106] Preset a matrix grid P of P×Q. The matrix element (p, q) is the Euclidean distance D between the difference curve and the deconvolution result of the target well and the deconvolution result of the comparison well pq , and in the path from the starting point P[1][1] to the ending point P[P][Q], search for the path L with the smallest sum of matrix elements min (p, q);

[0107] The search rule is solved by a recursive algorithm, and the process is as follows:

[0108] Initial condition:

[0109] L min (1, 1) = P[1][1]

[0110] Recursive rule:

[0111] v min (p, q) = min{v min (p, q - 1), L min (p - 1, q), L min (p - 1, q - 1)}

[0112] + P[p][q]

[0113] L min (p, q) represents the shortest path from the starting point P[1][1] to the matrix element P[p][q];

[0114] L min (p, q) in the matching p and q are the interpreted thin sand layer results of the target well and the comparison well. Assume that the depth range of the sampling points matched by the target well is Then the identified depth of the thin sand layer of the target well is:

[0115]

[0116]

[0117] Among them, is the top depth of the thin sand layer, is the bottom depth of the thin sand layer.

[0118] Preferably, in this embodiment, the process of obtaining the identified result of the thin sand layer of the target well based on the deconvolution matching result and the preliminary division result is:

[0119] Calculate the average of the deconvolution matching result and the preliminary division result, and obtain the thin sand layer identification result of the target well based on the average value.

[0120] Specifically, the process of calculating the average of the deconvolution matching result and the preliminary division result and obtaining the thin sand layer identification result of the target well based on the average value is as follows:

[0121] Calibrate all wells in the dataset as target wells and match the corresponding deconvolution identification results, which are used as the initial division results;

[0122] Randomly select another well as the target well again and repeat the comparison of adjacent well information. At this time, the initial division result is the division result after the previous round of iteration. Now assume that after z iterations, the error e between the thin sand layer identification result and the result of the previous round of iteration z is within the preset range [-ε, ε]. It is considered that the thin sand layer identification result no longer changes. At this time, the algorithm converges. Otherwise, perform the (z + 1)-th iteration until the algorithm converges. The division scheme at the time of convergence is the thin sand layer identification result;

[0123] Among them, the judgment in the z-th iteration is:

[0124]

[0125] Among them, e z represents the root mean square error between the thin sand layer identification results after the z-th iteration and the (z - 1)-th iteration; ε represents the minimum error value, which is an empirical parameter; z represents the current number of iterations; Z max represents the maximum number of iterations allowed by the method, which is an empirical parameter; R represents the final thin sand layer identification result; r z represents the thin sand layer identification result after the z-th iteration.

[0126] This application takes the clastic rock reservoir of a certain oilfield as an example to further elaborate on the present invention in detail;

[0127] Another method for identifying thin sand layers provided by the present invention is as Figure 2 shown, Figure 2 shows a flowchart of another method for identifying thin sand layers according to an embodiment of the present invention.

[0128] The well dataset in this embodiment consists of triples composed of well coordinates and well names. The data instance triples are as follows:

[0129] S = {(14724XXX, 45276XX, HD1X1), (14724XXX, 45287XX, HD1X2),...}

[0130] Among them, assuming that HD1X1 is the target well, the Euclidean distance is traversed through the dataset S to find the nearest neighbor well HD1X2, and the Euclidean distance between the two is as follows:

[0131]

[0132] The natural gamma ray curve GR of the target well and the comparison well i and GR j The example is as follows:

[0133] GR i =[111.35, 110.96,..., 112.056,..., 113.639]

[0134] GR j =[115.24, 115.976,..., 115.31,..., 113.336]

[0135] Among them, the number of sampling points of the target well is 133, and the number of sampling points of the comparison well is 129.

[0136] Perform deconvolution operation on the natural gamma ray curve. The deconvolution operator is [-5, 7, -3, 7, -5]. Calculate the difference between the natural gamma ray curve and the deconvolution-generated curve to obtain the difference curve. The positive part of the difference curve indicates the thin sand layer, such as Figure 3 shown.

[0137] Adopt the dynamic time warping algorithm to match the thin sand layer identification results of the target well and the comparison well. First, construct a 133×129 matrix grid P. The matrix elements are the Euclidean distances between the deconvolution results of the target well and the comparison well. Among the paths from the starting point P[1][1] to the ending point P

[133]

[129] , search for the path with the smallest sum of matrix elements. The search rule is solved by a recursive algorithm. The solved result is the interpreted thin sand layer result of the target well and the comparison well. Apply the above process to all wells in the dataset, and then perform the next processing on the same well. If the thin sand layer division result changes, repeat the above process until the thin sand layer division results of all wells no longer change. At this time, the algorithm converges, and the result is the thin sand layer identification result, such as Figure 4 shown.

[0138] The present invention provides a system for identifying thin sand layers, including:

[0139] The first processing module is configured to separately extract the natural gamma ray curves of the target well and the comparison well and perform deconvolution processing to obtain deconvolution-generated curves;

[0140] The second processing module is configured to obtain a difference curve and a preliminary thin sand layer division result based on the natural gamma ray curve and the deconvolution-generated curve;

[0141] A third processing module, configured to match the deconvolution matching results between the target well and the comparison well by using a dynamic time warping algorithm based on the Euclidean distance of the difference curves of the target well and the comparison well;

[0142] An output module, configured to obtain the thin sand layer identification result of the target well based on the deconvolution matching result and the preliminary division result.

[0143] In another embodiment of the present invention, a computer device is provided. The computer device includes a processor and a memory. The memory is used to store a computer program. The computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of a method for identifying thin sand layers.

[0144] In another embodiment of the present invention, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a computer device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, and this storage space stores the operating system of the terminal. And, one or more instructions suitable for being loaded and executed by the processor are also stored in this storage space. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the method for identifying thin sand layers in the above embodiments.

[0145] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0146] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for realizing the functions specified in Figure 1 one or more of the flows Figure 1 or blocks or the combination of blocks.

[0147] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing devices to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that realizes the functions specified in Figure 1 one or more of the flows Figure 1 or blocks or the combination of blocks.

[0148] These computer program instructions can also be loaded onto a computer or other programmable data processing devices, such that a series of operation steps are executed on the computer or other programmable devices to generate a computer-implemented process, so that the instructions executed on the computer or other programmable devices provide steps for realizing the functions specified in Figure 1 one or more of the flows Figure 1 or blocks or the combination of blocks.

[0149] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or equivalently replace some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying thin sand layers, characterized in that, Including the following steps: S1: Extract the natural gamma curves of the target well and the comparison well respectively and perform deconvolution processing to obtain the deconvolution-generated curves; S2: Obtain the difference curve and the preliminary division result of thin sand layers based on the natural gamma curve and the deconvolution-generated curve; S3: Based on the Euclidean distance of the difference curves of the target well and the comparison well, adopt the dynamic time warping algorithm to match the deconvolution matching results between the target well and the comparison well; S4: Obtain the identification result of thin sand layers of the target well based on the deconvolution matching result and the preliminary division result.

2. The method for identifying thin sand layers according to claim 1, wherein Summarize the well coordinate information of the well to be identified to form a triple dataset, and select the target well and the comparison well within the triple dataset.

3. The method for identifying thin sand layers according to claim 2, wherein The triple dataset formed by summarizing the well coordinate information of the well to be identified is: S = {(x1, y1, w1), (x2, y2, w2), …, (x i , y i , w i ), …, (x N , y N , w N )}; Among them, w i represents the well name of the i-th well; x i represents the abscissa of the i-th well; y i represents the ordinate of the i-th well; (x i , y i , w i ) is a triple composed of well coordinates and well names, N represents the number of wells in the dataset, and S represents the dataset.

4. The method for identifying thin sand layers according to claim 1, wherein, The target well is randomly selected and located at the edge of the triple dataset.

5. The method for identifying thin sand layers according to claim 1, wherein The target well is the well that selects the nearest neighbor of the target well through traversal calculation.

6. The method for identifying thin sand layers according to claim 5, wherein The process of selecting the well nearest to the target well as the comparison well through traversal calculation is: Suppose well w is selected i as the target well, then the nearest neighbor well w is found by traversing the dataset S using the Euclidean distance j , then: Among them, w j is the nearest neighbor comparison well for the target well w i . The argmin function returns the well coordinates in the set that satisfy the minimum Euclidean distance.

7. The method for identifying thin sand layers according to claim 1, wherein The process of respectively extracting the natural gamma curves of the target well and the comparison well and performing deconvolution processing to obtain the deconvolution-generated curves is: Natural gamma ray curve GR of the target well and the comparison well i and GR j are respectively as follows: Among them, gr i p is the data of the p-th sampling point of the target well, P is the number of sampling points of the target well, gr i q is the data of the q-th sampling point of the comparison well, Q is the number of sampling points of the comparison well; Perform deconvolution on the natural gamma ray curve. The deconvolution calculation results gr for individual sampling points of the target well and the comparison well are as follows: i p* and gr i q* are as follows: where α = [α1, α2, …, α k is the deconvolution operator, k is the number of deconvolution operators, and k is an odd number; The deconvolution results of the target well and the comparison well and are as follows: Since the deconvolution operation will include the first sampling points, the data of the first and last sampling points of the deconvolution result are set to 0.

8. The method for identifying thin sand layers according to claim 1, characterized in that, The process of obtaining the difference curve and the preliminary division result of thin sand layers based on the natural gamma curve and the deconvolution-generated curve is: Generate a difference curve from the difference between the natural gamma curve and the deconvolution-generated curve. The positive part in the difference curve is used as the preliminary division result of thin sand layers, which represents the developed section of thin sand layers.

9. The method for identifying thin sand layers according to claim 8, wherein The process of generating a difference curve from the difference between the natural gamma curve and the deconvolution-generated curve is: Among them, is the difference curve of the target well, is the difference curve of the comparison well; The positive part of the difference curve is used as the preliminary division result of thin sand layers, which represents the developed sections of thin sand layers. Then, the part of the difference curve greater than zero is assigned a value of 1, and the part less than or equal to zero is assigned a value of 0. The part of the difference curve equal to 1 is used as the preliminary division result of thin sand layers. Assume that the depth range of the thin sand layer is 10. The method for identifying thin sand layers according to claim 1, characterized in that, The process of adopting the dynamic time warping algorithm to match the deconvolution matching results between the target well and the comparison well is: Preset a matrix grid \(P\) of \(P\times Q\), and the matrix element \((p, q)\) is the Euclidean distance \(D\) between the difference curve and the deconvolution result of the target well and the deconvolution result of the comparison well Among the paths from the starting point \(P[1][1]\) to the ending point \(P[P][Q]\), search for the path \(L\) with the minimum sum of matrix elements pq \((p, q)\); min (p,q); The search rule is solved by a recursive algorithm, and the process is as follows: Initial condition: L min (1,1) = P[1][1] Recursive rule: L min (p,q) = min{L min (p,q - 1), L min (p - 1,q), L min (p - 1,q - 1)} + P[p][q] L min (p, q) represents the shortest path from the starting point P[1][1] to the matrix element P[p][q]; L min (p, q) The matching p and q in it are the interpreted thin sand layer results of the target well and the comparison well. Assume that the depth range of the sampling points matched by the target well is Then the identification depth of the thin sand layer of the target well is: Among them, is the top depth of the thin sand layer, is the bottom depth of the thin sand layer.

11. The method for identifying thin sand layers according to claim 1, characterized in that, The process of obtaining the identification result of thin sand layers of the target well based on the deconvolution matching result and the preliminary division result is: Take the average value of the deconvolution matching result and the preliminary division result, and obtain the identification result of thin sand layers of the target well based on the average value.

12. The method for identifying thin sand layers according to claim 11, wherein, The process of taking the average value of the deconvolution matching result and the preliminary division result and obtaining the identification result of thin sand layers of the target well based on the average value is: Calibrate all wells in the dataset as target wells and match the corresponding deconvolution identification results, which are used as the initial division results; Randomly select another well as the target well and repeat the comparison of adjacent well information. The initial partitioning result at this time is the partitioning result after the previous iteration. Now assume that after a total of z iterations, the error e between the thin sand layer identification result and the result of the previous iteration z Within the preset range [-ε, ε], it is determined that the thin sand layer identification result no longer changes. At this time, the algorithm converges. Otherwise, perform the (z + 1)-th iteration until the algorithm converges. The partitioning scheme at the time of its convergence is the thin sand layer identification result; Among them, the judgment in the z-th iteration is: Among them, e z represents the root mean square error of the thin sand layer recognition results after the z-th iteration and the z-1-th iteration; ε represents the minimum error value, which is an empirical parameter; z represents the current iteration number; Z max represents the maximum number of iterations allowed by the method, which is an empirical parameter; R represents the final recognition result of the thin sand layer; r z represents the recognition result of the thin sand layer after the z-th iteration.

13. A system for identifying thin sand layers, characterized in that, A method for identifying thin sand layers according to any one of claims 1 to 12 includes: The first processing module is configured to respectively extract the natural gamma curves of the target well and the comparison well and perform deconvolution processing to obtain the deconvolution-generated curves; The second processing module is configured to obtain the difference curve and the preliminary division result of thin sand layers based on the natural gamma curve and the deconvolution-generated curve; The third processing module is configured to, based on the Euclidean distance of the difference curves of the target well and the comparison well, adopt the dynamic time warping algorithm to match the deconvolution matching results between the target well and the comparison well; The output module is configured to obtain the identification result of thin sand layers of the target well based on the deconvolution matching result and the preliminary division result.

14. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of a method for identifying thin sand layers according to any one of claims 1 to 12.

15. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, the steps of a method for identifying thin sand layers according to any one of claims 1 to 12 are implemented.

Images