Logging lithofacies intelligent prediction method and system based on virtual rock core lithofacies sample
By establishing virtual core lithophase samples and geological statistics models, and training logging lithophase machine learning models, the problem of sparse deep carbonate lithophase data is solved, and high-precision intelligent prediction of complex lithophase reservoirs is achieved , provides technical support for oil and gas exploration.
Patent Information
- Application Number
- CN202310802594.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-06-30
- Publication Date
- 2024-12-31
AI Technical Summary
The coreing cost of deep carbonate lithophagos is high, resulting in sparse core lithophagos data, which cannot directly guide logging lithophagos, affecting the exploration and development of complex lithophagos.
By using finite core lithophagometry description data to establish a one-dimensional core lithophagometry and logging curve geological statistical model, simulate virtual core lithophagometry samples, train logging lithophagometer machine learning model, and perform transfer learning through real drilling data , establish a logging lithophase prediction model to achieve high-precision intelligent prediction of coreless wells and long sections.
The problems of scarce core lithophagometry data and lack of sample data for well logging prediction models are solved, and intelligent prediction of complex lithophagometry formations are realized, providing technical support for the exploration and evaluation of complex lithophagometry reservoirs.
Smart Images
Figure CN119227486A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of oil and gas exploration, and in particular relates to a well logging lithofacies intelligent prediction method and system based on virtual core lithofacies samples. Background Art
[0002] For deep carbonate lithofacies, due to objective reasons such as high coring costs, the core lithofacies description data is scarce and cannot directly guide logging lithofacies prediction, which seriously affects the work of dominant lithofacies prediction. By using limited core lithofacies description data to establish a one-dimensional core lithofacies and logging curve geostatistical model to simulate virtual core lithofacies samples, the logging lithofacies machine learning model training is realized, and the core lithofacies and logging data of the actual drilling coring section are used as the initial model for transfer learning to establish a logging lithofacies prediction model, which can achieve high-precision intelligent prediction of logging lithofacies in large quantities of wells and long well sections without coring, which is of great significance for guiding the exploration and development of complex lithofacies reservoirs. Summary of the invention
[0003] The purpose of the present invention is to provide a method and system for intelligent prediction of well logging lithofacies based on virtual core lithofacies samples in order to solve the above problems.
[0004] The present invention achieves the above-mentioned purpose through the following technical solutions:
[0005] A well logging lithofacies intelligent prediction method based on virtual core lithofacies samples comprises the following steps:
[0006] Acquire virtual sample data, wherein the virtual sample data includes virtual core lithofacies data and corresponding virtual logging curve data;
[0007] Constructing a well logging lithofacies machine learning model, and training the well logging lithofacies machine learning model through the virtual sample data to obtain a trained well logging lithofacies machine learning model;
[0008] Obtaining core lithofacies data and corresponding logging curve data of the actual drilling section and performing transfer learning training on the trained logging lithofacies machine learning model to obtain a final model;
[0009] The well logging curve data of the uncored section and the layer section without core lithofacies data are obtained and input into the final model to obtain the predicted core lithofacies data of the layer section.
[0010] As a further optimization solution of the present invention, the specific process of obtaining virtual sample data is as follows:
[0011] Acquire core lithofacies data and well logging curve data of the coring section, and construct a lithofacies-well logging curve geostatistical model based on the core lithofacies data and well logging curve data of the coring section;
[0012] The lithofacies-well logging curve geostatistical model is used to simulate and generate a plurality of virtual core lithofacies data and corresponding virtual logging curve data as virtual sample data.
[0013] As a further optimization scheme of the present invention, the lithofacies-logging curve geostatistical model includes a percentage function of each lithofacies and a variation function of each lithofacies.
[0014] As a further optimization scheme of the present invention, the model structure of the well logging lithofacies machine learning model includes an input layer, an LSTM layer, a fully connected layer, a pooling layer and an output layer connected in sequence.
[0015] As a further optimization scheme of the present invention, the model parameters of the well logging lithofacies machine learning model are solved using the Adam algorithm.
[0016] As a further optimization scheme of the present invention, the model structure of the trained well logging lithofacies machine learning model includes an input layer, an LSTM layer, a fully connected layer, a pooling layer and an output layer connected in sequence.
[0017] As a further optimization scheme of the present invention, the model parameters of the trained well logging lithofacies machine learning model are solved using the Adam algorithm.
[0018] A well logging lithofacies intelligent prediction system based on virtual core lithofacies samples, comprising:
[0019] An acquisition module is used to acquire virtual sample data, core lithofacies data and corresponding logging curve data of a real drilling section, and logging curve data of an uncored section and a layer section without core lithofacies data, wherein the virtual sample data includes virtual core lithofacies data and corresponding virtual logging curve data;
[0020] Model building module, used to build well logging lithofacies machine learning models;
[0021] A training module is used to train the well logging lithofacies machine learning model through the virtual sample data to obtain a trained well logging lithofacies machine learning model; and is also used for the core lithofacies data of the actual drilling section and the corresponding logging curve data to perform transfer learning training on the trained well logging lithofacies machine learning model to obtain a final model;
[0022] The prediction module is used to input the logging curve data of the uncored section and the layer section without core lithofacies data into the final model to obtain the predicted core lithofacies data of the layer section.
[0023] An electronic device comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;
[0024] Memory, used to store computer programs;
[0025] The processor is used to implement the well logging lithofacies intelligent prediction method based on virtual core lithofacies samples when executing the program stored in the memory.
[0026] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the well logging lithofacies intelligent prediction method based on virtual core lithofacies samples.
[0027] The beneficial effects of the present invention are:
[0028] The present invention utilizes limited core lithofacies description data to establish a one-dimensional core lithofacies and logging curve geostatistical model to simulate virtual core lithofacies samples, thereby realizing the training of logging lithofacies machine learning model, and uses the actual drilling coring section core lithofacies and logging data as the initial model for transfer learning to establish a logging lithofacies prediction model. This solves the problems of scarcity of core lithofacies data due to high coring costs and lack of sample data for logging lithofacies prediction models, and realizes intelligent prediction of logging lithofacies in complex lithofacies formations, providing technical support for the exploration and evaluation of complex lithology reservoirs and the detailed description of oil and gas reservoirs. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] Figure 1 It is a flowchart of the present invention;
[0030] Figure 2 It is a schematic diagram of logging data and core description lithology of a coring section of an exploration well A in a certain area of the Tarim Basin in an embodiment of the present invention;
[0031] Figure 3 It is a schematic diagram of a one-dimensional geostatistical model of core lithofacies and well logging curves of the present invention;
[0032] Figure 4 is a schematic diagram of virtual core lithofacies and curve samples in an embodiment of the present invention;
[0033] Figure 5 is a schematic diagram of the structure of a machine learning model in an embodiment of the present invention;
[0034] Figure 6 is a schematic diagram of an example of training a well logging lithofacies machine learning model in an embodiment of the present invention;
[0035] Figure 7 It is a schematic diagram of the intelligent prediction results of well logging lithofacies in the coring section of an exploration well A in a certain area of the Tarim Basin in an embodiment of the present invention;
[0036] Figure 8 Schematic diagram of the lithofacies intelligent prediction results of well logging in a B exploration well without coring in a certain area of the Tarim Basin in an embodiment of the present invention;
[0037] Fig. 9 It is a system structure block diagram of the present invention;
[0038] Fig.10 It is a structural block diagram of an electronic device of the present invention. DETAILED DESCRIPTION
[0039] The present application is further described in detail below in conjunction with the accompanying drawings. It is necessary to point out here that the following specific implementation methods are only used to further illustrate the present application and cannot be understood as limiting the scope of protection of the present application. Technical personnel in this field can make some non-essential improvements and adjustments to the present application based on the above application content.
[0040] like Figure 1 As shown, a well logging lithofacies intelligent prediction method based on virtual core lithofacies samples comprises the following steps:
[0041] S1: Acquire virtual sample data, where the virtual sample data includes virtual core lithofacies data and corresponding virtual logging curve data;
[0042] S2: constructing a well logging lithofacies machine learning model, and training the well logging lithofacies machine learning model through the virtual sample data to obtain a trained well logging lithofacies machine learning model;
[0043] S3: Obtaining core lithofacies data and corresponding logging curve data of the actual drilling section and performing transfer learning training on the trained logging lithofacies machine learning model to obtain a final model;
[0044] S4: Obtain the logging curve data of the uncored section and the layer section without core lithofacies data and input them into the final model to obtain the predicted core lithofacies data of the layer section.
[0045] The specific process of obtaining virtual sample data is as follows:
[0046] Acquire core lithofacies data and well logging curve data of the coring section, and construct a lithofacies-well logging curve geostatistical model based on the core lithofacies data and well logging curve data of the coring section;
[0047] The lithofacies-well logging curve geostatistical model is used to simulate and generate a plurality of virtual core lithofacies data and corresponding virtual logging curve data as virtual sample data.
[0048] The lithofacies-well logging curve geostatistical model includes a percentage function of each lithofacies and a variation function of each lithofacies.
[0049] The model structures of the well logging lithofacies machine learning model and the trained well logging lithofacies machine learning model both include an input layer, an LSTM layer, a fully connected layer, a pooling layer and an output layer connected in sequence; the model parameters are all solved using the Adam algorithm.
[0050] In this embodiment, a complex carbonate reservoir in a certain area of the Tarim Basin is taken as an example to perform intelligent prediction of well logging lithofacies. The specific process is as follows:
[0051] Step 1: Use the core lithofacies data and logging data of the coring section to establish a one-dimensional geostatistical model of lithofacies and logging curves.
[0052] Specifically, Figure 2 As shown in the figure, Well A is a typical exploration well in a certain area of the Tarim Basin. There is a relatively long well section of core data in the illustrated layer section, and core lithofacies analysis has been carried out. The rightmost column in the figure is the core lithofacies description result, among which the core lithofacies include limestone, dolomitic limestone, dolomite, lime dolomite, gypsum, and gypsum dolomite.
[0053] Using the core lithofacies data and logging curves of the coring section, a one-dimensional geostatistical model of lithofacies (discrete attributes) and curves (continuous attributes) can be established. Figure 3 As shown in Figure 1, the one-dimensional geostatistical model of lithofacies includes the percentage of each lithofacies (a) and the variation function of each lithofacies (b). Similarly, for the one-dimensional geostatistical model of the curve (continuous attribute), (c) is the GR curve data histogram and probability density curve under each lithofacies, and (d) is the variation function of the GR curve under the dolomite facies.
[0054] Step 2: Implement geostatistical simulation of lithofacies-well logging curves and establish a virtual core lithofacies sample library.
[0055] Specifically, the established one-dimensional geostatistical model can be used to simulate and generate multiple virtual core lithofacies data (discrete attributes) and their corresponding virtual logging curves (continuous attributes). Figure 4 As shown, in Figure 1 In the coring section of Well A, the lithofacies column on the left is the virtual core lithofacies generated by geostatistical model simulation, and the remaining columns are virtual logging curves simulated by geostatistics based on the virtual core lithofacies. Figure 2 and Figure 4 , virtual wells can effectively expand samples based on geostatistical models.
[0056] Step 3: Using the virtual sample library as training data, a well logging lithofacies machine learning model is trained to predict core lithofacies using well logging curves.
[0057] Specifically, the training of the well logging lithofacies machine learning model mainly includes the construction of the machine learning model structure and the determination of the machine learning parameters. Figure 5As shown in the figure, in this step, the machine learning model structure used includes input layer, long short-term memory artificial neural network layer (long short-term memory, LSTM), fully connected layer, pooling layer and output layer. The machine learning model parameters are solved using the adaptive moment estimation algorithm (Adaptive moment estimation, Adam) method. The training data used is the virtual core lithofacies sample library. Figure 6 For this method Figure 4 The rightmost column in the figure shows the core lithofacies results predicted by the machine learning model. The predicted results in the rightmost column are highly matched with the sample data in the leftmost column, indicating that the machine learning model can be effectively closed.
[0058] Step 4: Use the machine learning model trained with the virtual sample library as the initial model, and the core lithofacies-logging data of the actual drilling section as the training data to achieve transfer learning of the logging lithofacies machine learning model.
[0059] Specifically, the well logging lithofacies machine learning model trained based on the virtual sample library is used as the initial model, and the core lithofacies of the actual drilling coring section and the logging curves of the corresponding depth are used as input to perform transfer learning on the well logging lithofacies machine learning model. The implementation of transfer learning uses the same machine learning model as the well logging lithofacies machine learning model trained based on the virtual sample library ( Figure 5 ) and the model parameter solving method (Adam), the difference mainly lies in the difference between the initial model and the training data. Figure 7 The well logging lithofacies machine learning model trained for transfer learning is Figure 2 The prediction results of the coring section of the A exploration well in the figure. The first column on the right side of the figure shows the lithofacies predicted by the machine learning model, and the second column on the right side shows the lithofacies description of the coring section. The high similarity between the two verifies the prediction ability of the machine learning model.
[0060] Step 5: For the uncored sections and the layers without core lithofacies description data, the well logging data is used as input, and the well logging lithofacies prediction model after transfer learning is used to realize the lithofacies intelligent prediction of a large number of wells and long well sections.
[0061] Specifically, the well logging lithofacies machine learning model established in the first four steps is applied to the B exploration well with no coring data in the same work area as the A well to perform intelligent lithofacies prediction. Figure 8As shown in the figure, the lithofacies results predicted in the rightmost column realize the lithofacies transition characteristics from deep to shallow and from dolomite to limestone in the vertical direction. Based on the predicted lithofacies transition trend, the thin-layer lithology changes in dolomite and limestone can also be described in detail. The gypsum and gypsum dolomite encountered in Well A changed to limestone in Well B, revealing the lateral changes of lithofacies, guiding the lateral prediction and understanding of lithofacies, and providing a research basis for the subsequent characterization of dominant sedimentary facies belts and lithofacies belts.
[0062] like Fig. 9 As shown, a well logging lithofacies intelligent prediction system based on virtual core lithofacies samples includes:
[0063] The acquisition module 11 is used to acquire virtual sample data, core lithofacies data and corresponding logging curve data of the actual drilling section, and logging curve data of the uncored section and the layer section without core lithofacies data, wherein the virtual sample data includes virtual core lithofacies data and corresponding virtual logging curve data;
[0064] A model building module 12, used to build a well logging lithofacies machine learning model;
[0065] The training module 13 is used to train the well logging lithofacies machine learning model through the virtual sample data to obtain a trained well logging lithofacies machine learning model; and is also used for the core lithofacies data of the actual drilling section and the corresponding logging curve data to perform transfer learning training on the trained well logging lithofacies machine learning model to obtain a final model;
[0066] The prediction module 14 is used to input the logging curve data of the uncored section and the layer section without core lithofacies data into the final model to obtain the predicted core lithofacies data of the layer section.
[0067] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.
[0068] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of the present invention. Ordinary technicians in this field can understand and implement it without paying creative work.
[0069] In the above-mentioned embodiment, any multiple of all modules can be combined in one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least some functions of one or more of these modules can be combined with at least some functions of other modules and implemented in one module. At least one of all modules can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation modes of software, hardware and firmware or in a suitable combination of any of them. Alternatively, at least one of all modules can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be executed.
[0070] like Fig.10 As shown, an electronic device provided by an embodiment of the present disclosure includes a processor 1110, a communication interface 1120, a memory 1130 and a communication bus 1140, wherein the processor 1110, the communication interface 1120, and the memory 1130 communicate with each other through the communication bus 1140;
[0071] Memory 1130, used for storing computer programs;
[0072] The processor 1110 is used to implement the well logging lithofacies intelligent prediction method based on virtual core lithofacies samples when executing the program stored in the memory 1130.
[0073] The communication bus 1140 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus 1140 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.
[0074] The communication interface 1120 is used for communication between the above electronic device and other devices.
[0075] The memory 1130 may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory 1130 may also be at least one storage device located away from the processor 1110.
[0076] The above-mentioned processor 1110 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.
[0077] The embodiment of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the above-mentioned method for intelligent prediction of well logging lithofacies based on virtual core lithofacies samples is implemented.
[0078] The computer-readable storage medium may be included in the device / apparatus described in the above embodiment; or it may exist independently without being assembled into the device / apparatus. The above computer-readable storage medium carries one or more programs. When the above one or more programs are executed, the well logging lithofacies intelligent prediction method based on virtual core lithofacies samples according to the embodiment of the present disclosure is implemented.
[0079] According to an embodiment of the present disclosure, a computer-readable storage medium may be a non-volatile computer-readable storage medium, such as but not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.
[0080] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.
Claims
1. A well logging lithofacies intelligent prediction method based on virtual core lithofacies samples, characterized in that: The following steps are involved: Acquire virtual sample data, wherein the virtual sample data includes virtual core lithofacies data and corresponding virtual logging curve data; Constructing a well logging lithofacies machine learning model, and training the well logging lithofacies machine learning model through the virtual sample data to obtain a trained well logging lithofacies machine learning model; Obtaining core lithofacies data and corresponding logging curve data of the actual drilling section and performing transfer learning training on the trained logging lithofacies machine learning model to obtain a final model; The well logging curve data of the uncored section and the layer section without core lithofacies data are obtained and input into the final model to obtain the predicted core lithofacies data of the layer section.
2. The method for intelligent prediction of well logging lithofacies based on virtual core lithofacies samples according to claim 1, characterized in that: The specific process of obtaining virtual sample data is as follows: Acquire core lithofacies data and well logging curve data of the coring section, and construct a lithofacies-well logging curve geostatistical model based on the core lithofacies data and well logging curve data of the coring section; The lithofacies-well logging curve geostatistical model is used to simulate and generate a plurality of virtual core lithofacies data and corresponding virtual logging curve data as virtual sample data.
3. The method for intelligent prediction of well logging lithofacies based on virtual core lithofacies samples according to claim 1, characterized in that: The lithofacies-logging curve geostatistical model includes a percentage function of each lithofacies and a variation function of each lithofacies.
4. The method for intelligent prediction of well logging lithofacies based on virtual core lithofacies samples according to claim 1, characterized in that: The model structure of the well logging lithofacies machine learning model includes an input layer, an LSTM layer, a fully connected layer, a pooling layer and an output layer connected in sequence.
5. The method for intelligent prediction of well logging lithofacies based on virtual core lithofacies samples according to claim 1, characterized in that: The model parameters of the well logging lithofacies machine learning model are solved using the Adam algorithm.
6. The method for intelligent prediction of well logging lithofacies based on virtual core lithofacies samples according to claim 1, characterized in that: The model structure of the trained well logging lithofacies machine learning model includes an input layer, an LSTM layer, a fully connected layer, a pooling layer and an output layer connected in sequence.
7. The method for intelligent prediction of well logging lithofacies based on virtual core lithofacies samples according to claim 1, characterized in that: The model parameters of the trained well logging lithofacies machine learning model are solved using the Adam algorithm.
8. A well logging lithofacies intelligent prediction system based on virtual core lithofacies samples, characterized in that: include: An acquisition module is used to acquire virtual sample data, core lithofacies data and corresponding logging curve data of a real drilling section, and logging curve data of an uncored section and a layer section without core lithofacies data, wherein the virtual sample data includes virtual core lithofacies data and corresponding virtual logging curve data; Model building module, used to build well logging lithofacies machine learning models; A training module is used to train the well logging lithofacies machine learning model through the virtual sample data to obtain a trained well logging lithofacies machine learning model; and is also used for the core lithofacies data of the actual drilling section and the corresponding logging curve data to perform transfer learning training on the trained well logging lithofacies machine learning model to obtain a final model; The prediction module is used to input the logging curve data of the uncored section and the layer section without core lithofacies data into the final model to obtain the predicted core lithofacies data of the layer section.
9. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory, used to store computer programs; The processor is used to implement the well logging lithofacies intelligent prediction method based on virtual core lithofacies samples as described in any one of claims 1 to 7 when executing the program stored in the memory.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for intelligent prediction of well logging lithofacies based on virtual core lithofacies samples according to any one of claims 1 to 7 is implemented.