Oil well logging prediction method and prediction device based on big data

Abstract

The present invention discloses a petroleum logging prediction method and prediction device based on big data, which are applied to a petroleum logging prediction system. The method includes: obtaining petroleum logging big data; creating a prediction model based on the petroleum logging big data; obtaining oil and gas monitoring metadata; processing the oil and gas monitoring metadata based on the prediction model to generate corresponding petroleum logging prediction results. By adopting a distributed data acquisition and data analysis method, a more accurate analysis of petroleum logging can be performed based on a large amount of data generated in the historical petroleum logging process, thereby achieving a more accurate petroleum logging prediction effect, meeting people's growing demand for prediction accuracy; on the other hand, by adopting a physical model combined with a machine model to predict and analyze petroleum logging data, the speed of data analysis and processing can be effectively improved, and the prediction efficiency can be improved. At the same time, based on the training and analysis of big data, the prediction accuracy can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of petroleum well logging, and in particular to a petroleum well logging prediction method based on big data and a petroleum well logging prediction device based on big data. Background Art

[0002] As human life needs continue to increase, people's demand for living materials is also increasing. In order to meet people's growing energy needs, well logging technology is also constantly developing and improving. Since its birth, well logging technology has undergone many revolutionary reforms.

[0003] With the continuous development of science and technology and the continuous improvement of environmental protection concepts, people have put forward higher requirements and more demands on oil well logging technology. However, in the actual application of existing oil well logging technology, technicians have found that the existing technology has at least the following technical problems:

[0004] On the one hand, the large amount of data acquired and accumulated in the historical logging process has not been fully mined and utilized due to its multi-format and multi-scale characteristics.

[0005] On the other hand, the existing logging technology analyzes the logging data of a single well to analyze the reservoir characteristics of the reservoir, and then determines the credibility of the analysis data and further predicts and calculates the logging data based on the personal knowledge reserves and actual work experience of the technicians. However, the above analysis and calculation process based on expert experience has low interpretation accuracy, especially when identifying lithology for complex composition formations, the detection error is large, which is difficult to meet the needs of exploration and development. Summary of the invention

[0006] In order to overcome the above-mentioned technical problems existing in the prior art, an embodiment of the present invention provides an oil well logging prediction method based on big data. By adopting a distributed data acquisition method and a data analysis and processing method based on a prediction model, rapid and accurate analysis and prediction of oil well logging data are achieved, thus meeting the actual needs of users.

[0007] In order to achieve the above-mentioned purpose, an embodiment of the present invention provides an oil well logging prediction method based on big data, which is applied to an oil well logging prediction system. The method includes: obtaining oil well logging big data; creating a prediction model based on the oil well logging big data; obtaining oil and gas monitoring metadata; processing the oil and gas monitoring metadata based on the prediction model to generate corresponding oil well logging prediction results.

[0008] Preferably, the acquisition of oil well logging big data includes: determining a data source and a data format corresponding to the data source; acquiring a data analysis tool corresponding to the data format, wherein the data analysis tool is connected to the oil well logging prediction system in a distributed manner; generating a data import instruction; based on the data import instruction, controlling the data analysis tool to acquire raw data from the data source, and performing data analysis on the raw data based on the data format to generate corresponding oil well logging big data.

[0009] Preferably, the creating a prediction model based on the oil well logging big data includes: creating a physical algorithm model and a machine learning model based on the oil well logging big data respectively; and creating a prediction model based on the physical algorithm model and the machine learning model.

[0010] Preferably, the creation of a physical algorithm model based on the oil well logging big data includes: obtaining an initial neural network model and simulated logging data; calculating and determining characteristic information of the simulated logging data and the oil well logging big data; and training the initial neural network model based on the characteristic information to generate the physical algorithm model.

[0011] Preferably, the initial neural network model is trained based on the characteristic information to generate the physical algorithm model, including: S231) training the initial neural network model based on the characteristic information to determine the maximum correlation coefficient and the minimum correlation coefficient of the initial neural network model; S232) determining an intermediate model based on the maximum correlation coefficient and the minimum correlation coefficient; S233) obtaining the output result of the intermediate model based on the simulated logging data, and judging whether the deviation between the output result and the oil logging big data meets the preset deviation requirement; S2341) if so, using the intermediate model as the physical algorithm model; S2342) otherwise, adjusting the simulated logging data to obtain adjusted data, training the intermediate model based on the adjusted data to obtain a new intermediate model, and continuing to execute step S233).

[0012] Preferably, the method also includes: determining a first neural network model and a second neural network model based on the physical algorithm model; performing forward training on the first neural network model based on the oil well logging big data to obtain a forward training model; performing reverse training on the second neural network model based on the oil well logging big data to obtain a reverse training model; obtaining a preset encoder and a preset decoder, and generating a corresponding automatic decoder based on the preset encoder and the preset decoder; optimizing the physical algorithm model based on the forward training model, the reverse training model and the automatic decoder to obtain an optimized physical algorithm model.

[0013] Preferably, the creation of a machine learning model based on the oil well logging big data includes: generating a first training data set Xnew based on the oil well logging big data; obtaining a preset tree classification rule tree; generating a machine learning algorithm predict based on the first training data set Xnew and the preset tree classification rule tree, the machine learning algorithm predict being represented as: Ynew = predict(tree,Xnew), where Ynew is represented as a prediction result; or generating a corresponding data matrix X based on the simulated logging data; obtaining a response matrix Y corresponding to the data matrix X; generating a machine learning ens based on the data matrix X and the response matrix Y, the machine learning algorithm ens being represented as: ens = fitensemble(X, Y, model, numberens, learners), where numberens is represented as attribute information of the data matrix X, and learners is represented as attribute information of the response matrix Y.

[0014] Preferably, the method also includes: generating corresponding simulation data based on the oil and gas monitoring metadata; optimizing the prediction model based on the simulation data to obtain an optimized model; processing the oil and gas monitoring metadata based on the optimized model to generate corresponding oil well logging prediction results.

[0015] Preferably, the oil well logging prediction system includes multiple distributed computing nodes, and the oil and gas monitoring metadata is processed based on the prediction model to generate corresponding oil well logging prediction results, including: generating at least one prediction task based on the oil and gas monitoring metadata; determining the matching distributed computing node corresponding to each prediction task; executing the corresponding prediction task based on the prediction model through the matching distributed computing node to generate corresponding oil well logging prediction results.

[0016] Correspondingly, an embodiment of the present invention also provides an oil well logging prediction device based on big data, which includes: a big data unit for acquiring oil well logging big data; a model creation unit for creating a prediction model based on the oil well logging big data; a metadata acquisition unit for acquiring oil and gas monitoring metadata; and a prediction unit for processing the oil and gas monitoring metadata based on the prediction model to generate corresponding oil well logging prediction results.

[0017] Preferably, the big data unit includes: a data source determination module, used to determine the data source and the data format corresponding to the data source; a parsing tool determination module, used to obtain a data parsing tool corresponding to the data format, and the data parsing tool is connected to the oil well logging prediction system in a distributed manner; an instruction generation module, used to generate data import instructions; a data import module, used to control the data parsing tool to obtain raw data from the data source based on the data import instructions, and perform data parsing on the raw data based on the data format to generate corresponding oil well logging big data.

[0018] Preferably, the model creation unit includes: an intermediate model creation module, used to create a physical algorithm model and a machine learning model based on the oil well logging big data; and a prediction model creation module, used to create a prediction model based on the physical algorithm model and the machine learning model.

[0019] Preferably, the intermediate model creation module includes a first model creation module, and the first model creation module is used to: obtain an initial neural network model and simulated logging data; calculate and determine characteristic information of the simulated logging data and the oil logging big data; train the initial neural network model based on the characteristic information to generate the physical algorithm model.

[0020] Preferably, the initial neural network model is trained based on the characteristic information to generate the physical algorithm model, including: S231) training the initial neural network model based on the characteristic information to determine the maximum correlation coefficient and the minimum correlation coefficient of the initial neural network model; S232) determining an intermediate model based on the maximum correlation coefficient and the minimum correlation coefficient; S233) obtaining the output result of the intermediate model based on the simulated logging data, and judging whether the deviation between the output result and the oil logging big data meets the preset deviation requirement; S2341) if so, using the intermediate model as the physical algorithm model; S2342) otherwise, adjusting the simulated logging data to obtain adjusted data, training the intermediate model based on the adjusted data to obtain a new intermediate model, and continuing to execute step S233).

[0021] Preferably, the first model creation module is also used to: determine the first neural network model and the second neural network model based on the physical algorithm model; perform forward training on the first neural network model based on the oil well logging big data to obtain a forward training model; perform reverse training on the second neural network model based on the oil well logging big data to obtain a reverse training model; obtain a preset encoder and a preset decoder, and generate a corresponding automatic decoder based on the preset encoder and the preset decoder; optimize the physical algorithm model based on the forward training model, the reverse training model and the automatic decoder to obtain an optimized physical algorithm model.

[0022] Preferably, the intermediate model creation module includes a second model creation module, which is used to: generate a first training data set Xnew based on the oil well logging big data; obtain a preset tree classification rule tree; generate a machine learning algorithm predict based on the first training data set Xnew and the preset tree classification rule tree, and the machine learning algorithm predict is characterized by: Ynew = predict(tree,Xnew), where Ynew is characterized as the prediction result; or generate a corresponding data matrix X based on the simulated well logging data; obtain a response matrix Y corresponding to the data matrix X; generate a machine learning ens based on the data matrix X and the response matrix Y, and the machine learning algorithm ens is characterized by: ens =fitensemble(X, Y, model, numberens, learners), where numberens is characterized as the attribute information of the data matrix X, and learners is characterized as the attribute information of the response matrix Y.

[0023] Preferably, the device also includes a model optimization unit, which is used to: generate corresponding simulation data based on the oil and gas monitoring metadata; optimize the prediction model based on the simulation data to obtain an optimized model; process the oil and gas monitoring metadata based on the optimized model to generate corresponding oil well logging prediction results.

[0024] Preferably, the oil well logging prediction system includes multiple distributed computing nodes, and the prediction unit is used to: generate at least one prediction task based on the oil and gas monitoring metadata; determine the matching distributed computing node corresponding to each prediction task; and execute the corresponding prediction task based on the prediction model through the matching distributed computing node to generate the corresponding oil well logging prediction result.

[0025] On the other hand, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which implements the method provided by the embodiment of the present invention when the program is executed by a processor.

[0026] Through the technical solution provided by the present invention, the present invention has at least the following technical effects:

[0027] By adopting distributed data collection and data analysis methods, we can conduct more accurate analysis of oil logging based on the large amount of data generated in the historical oil logging process, thereby achieving more accurate oil logging prediction results and meeting people's growing demand for prediction accuracy. On the other hand, by using physical models combined with machine models to predict and analyze oil logging data, we can effectively improve the speed of data analysis and processing, improve prediction efficiency, and at the same time, based on big data training and analysis, we can effectively improve prediction accuracy.

[0028] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0030] Figure 1 This is a specific implementation flow chart of the oil well logging prediction method based on big data provided by an embodiment of the present invention;

[0031] Figure 2 It is a specific implementation flow chart of obtaining oil well logging big data in the oil well logging prediction method based on big data provided by an embodiment of the present invention;

[0032] Figure 3 It is a specific implementation flow chart of generating a physical algorithm model in the oil well logging prediction method based on big data provided by an embodiment of the present invention;

[0033] Figure 4 It is a structural schematic diagram of a petroleum well logging prediction device based on big data provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0035] The terms "system" and "network" in the embodiments of the present invention can be used interchangeably. "Multiple" means two or more than two. In view of this, "multiple" can also be understood as "at least two" in the embodiments of the present invention. "And / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / ", unless otherwise specified, generally indicates that the previous and subsequent associated objects are in an "or" relationship. In addition, it should be understood that in the description of the embodiments of the present invention, the words "first", "second", etc. are only used to distinguish the purpose of description, and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying order.

[0036] See also Figure 1 The embodiment of the present invention provides a petroleum well logging prediction method based on big data, which is applied to a petroleum well logging prediction system. The method includes:

[0037] S10) Obtaining oil well logging big data;

[0038] S20) creating a prediction model based on the oil well logging big data;

[0039] S30) obtaining oil and gas monitoring metadata;

[0040] S40) Processing the oil and gas monitoring metadata based on the prediction model to generate corresponding oil well logging prediction results.

[0041] In a possible implementation, firstly, oil logging big data is obtained. In the prior art, oil logging data can be obtained through various channels, such as point logging hand-drawing, photoelectric recording pen, tracing and printing, tape recording, and disk recording, etc., to form the data basis for oil logging prediction. However, there are various existing data acquisition channels, and the processing terminal can often only process data in a single format or a single channel. Therefore, it is also necessary to optimize the acquisition process of logging big data.

[0042] See also Figure 2 In an embodiment of the present invention, the step of obtaining petroleum well logging big data includes:

[0043] S11) determining a data source and a data format corresponding to the data source;

[0044] S12) obtaining a data analysis tool corresponding to the data format, wherein the data analysis tool is connected to the oil well logging prediction system in a distributed manner;

[0045] S13) generating a data import instruction;

[0046] S14) Based on the data import instruction, the data analysis tool is controlled to obtain raw data from the data source, and the raw data is analyzed based on the data format to generate corresponding oil well logging big data.

[0047] For example, in a possible implementation, when performing an operation of importing oil well logging big data, first determine the data source and the data format corresponding to the data source, and then obtain a data analysis tool corresponding to the data format. For example, when importing data recorded in a photoelectric record table, the corresponding photoelectric conversion tool is used to import the corresponding data into the oil well logging prediction system in a specific format. Since different regions use different data collection methods, in order to meet the data import requirements of each region, the above-mentioned data analysis tool is connected to the oil well logging prediction system in a distributed manner, and the analyzed data is imported into the oil well logging prediction system. After determining the data analysis tool, a corresponding data import instruction is generated, and then based on the data import instruction, the data analysis tool is controlled to obtain raw data from the corresponding data source, and the raw data is analyzed according to the data format to generate the corresponding oil well logging big data.

[0048] In an embodiment of the present invention, by accessing various types of data analysis tools in a distributed manner, the oil logging data of various regions can be imported into the oil logging prediction system in real time and accurately, so that the oil logging prediction system can perform accurate prediction operations based on the acquired oil logging big data, thereby improving the accuracy of oil logging prediction.

[0049] However, after obtaining the above oil well logging big data, if the traditional manual processing method is used for data analysis and processing, it will not meet people's needs for data processing efficiency, and the deviation of manual analysis and processing will greatly reduce the accuracy of data processing. Therefore, in order to solve the above technical problems, in an embodiment of the present invention, the oil well logging data is automatically analyzed by further creating a prediction model, which can effectively improve the prediction efficiency and prediction accuracy.

[0050] However, in actual application, due to specific technical problems in the field of oil well logging technology, such as strong data diversity, rapid data changes, and large data volume, the use of traditional prediction models or single prediction models cannot quickly and accurately process oil well logging data, nor can it predict subsequent oil well logging data.

[0051] In order to solve the above technical problems, in an embodiment of the present invention, the creation of a prediction model based on the oil well logging big data includes: creating a physical algorithm model and a machine learning model based on the oil well logging big data respectively; and creating a prediction model based on the physical algorithm model and the machine learning model.

[0052] In an embodiment of the present invention, by combining a physical algorithm model with a machine learning model and further generating a prediction model for oil well logging, it is possible to simultaneously predict oil well logging from the aspects of physical properties and theoretical data (simulation data), thereby effectively improving the prediction speed, efficiency and accuracy.

[0053] In an embodiment of the present invention, the creation of a physical algorithm model based on the oil well logging big data includes: obtaining an initial neural network model and simulated logging data; calculating and determining characteristic information of the simulated logging data and the oil well logging big data; training the initial neural network model based on the characteristic information to generate the physical algorithm model.

[0054] See also Figure 3 Further, in an embodiment of the present invention, the training of the initial neural network model based on the feature information to generate the physical algorithm model includes:

[0055] S231) training the initial neural network model based on the feature information to determine the maximum correlation coefficient and the minimum correlation coefficient of the initial neural network model;

[0056] S232) determining an intermediate model based on the maximum correlation coefficient and the minimum correlation coefficient;

[0057] S233) obtaining an output result of the intermediate model based on the simulated well logging data, and determining whether a deviation between the output result and the petroleum well logging big data meets a preset deviation requirement;

[0058] S2341) If yes, use the intermediate model as the physical algorithm model;

[0059] S2342) Otherwise, adjust the simulated logging data to obtain adjusted data, train the intermediate model based on the adjusted data to obtain a new intermediate model, and continue to execute step S233).

[0060] There are many physical algorithms in the prior art. After comparative analysis, technicians determined that the algorithm model based on neural network can perform the best analysis on oil logging technology. Therefore, in one possible implementation, the initial neural network model and simulated logging data are first obtained. For example, the initial neural network model is a general neural network model. The technicians can pre-set certain simulated logging data according to the actual oil logging data, and then calculate and determine the characteristic information between the simulated logging data and the oil logging big data. For example, in the process of training through the above-mentioned initial neural network model, the simulated logging data and the oil logging big data are respectively input into the initial neural network model, and the corresponding characteristic values ​​are determined, and then the initial neural network model is continued to be trained according to the characteristic information, so as to generate a physical algorithm model.

[0061] For example, the initial neural network model is iteratively trained multiple times through the characteristic information, and the maximum correlation coefficient and the minimum correlation coefficient of the initial neural network model are finally determined, and the intermediate model is determined according to the maximum correlation coefficient and the minimum correlation coefficient. At this time, the input can be further improved, and the output result of the intermediate model when the above-mentioned simulated logging data is used as input data is obtained. Then, it can be determined whether the deviation between the output result and the oil logging big data meets the preset deviation requirement. For example, in an embodiment of the present invention, the preset deviation requirement can be a comparison result of the error value of the cost function of the intermediate model and a preset error value (for example, 0.001 ohm-meter). When the error value of the cost function is less than the preset error value, it can be determined that the deviation of the intermediate model meets the preset deviation requirement.

[0062] In one embodiment, it is determined that the deviation of the intermediate model does not meet the preset deviation requirement, so the above-mentioned simulated logging data is adjusted to obtain adjusted data, and then the adjusted data is used as new input data to further train the intermediate model to obtain a new intermediate model, and then it is continued to be judged whether the output result of the new intermediate model meets the above-mentioned preset deviation requirement, and the calculation is continuously iterated until the output result of the intermediate model meets the above-mentioned preset deviation requirement, and the finally determined intermediate model is used as the physical algorithm model.

[0063] In the embodiment of the present invention, by using a physical algorithm model based on a neural network model to predict oil well logging data, the prediction efficiency and prediction accuracy can be effectively improved. However, in actual application, technicians have found that since the maximum and minimum values ​​of oil well logging data are often fixed or in the same range, in the process of using a single model for training, a large number of (e.g., millions of) forward calculations may be required to obtain a more accurate physical algorithm model, which cannot meet people's actual needs.

[0064] In order to solve the above technical problems, in an embodiment of the present invention, the method also includes: determining a first neural network model and a second neural network model based on the physical algorithm model; performing forward training on the first neural network model based on the oil well logging big data to obtain a forward training model; performing reverse training on the second neural network model based on the oil well logging big data to obtain a reverse training model; obtaining a preset encoder and a preset decoder, and generating a corresponding automatic decoder based on the preset encoder and the preset decoder; optimizing the physical algorithm model based on the forward training model, the reverse training model and the automatic decoder to obtain an optimized physical algorithm model.

[0065] In one possible implementation, after the above-mentioned physical algorithm model is preliminarily determined, the first neural network model and the second neural network model are further determined, for example, two identical physical algorithm models are copied as the first neural network model and the second neural network model, and then the above-mentioned oil well logging big data are used to perform forward training and reverse training on the above-mentioned first neural network model and the second neural network model respectively, and the corresponding forward training model and reverse training model are obtained, and then the preset encoder and the preset decoder are obtained, and the corresponding automatic decoder is generated, and then the forward training model and the reverse training model are fused through the automatic decoder to optimize the physical algorithm model and generate an optimized physical algorithm model.

[0066] In an embodiment of the present invention, by adopting forward calculation model and reverse calculation model for calculation and optimizing the initially acquired physical algorithm model, the convergence speed of the physical algorithm model in the actual application process can be further improved, the generation speed and prediction accuracy of the physical algorithm model can be effectively improved, and the actual needs of users can be met.

[0067] On the other hand, in order to achieve further in-depth application of well logging technology and realize accurate analysis and prediction of technologies such as lithology classification, well logging data repair, well logging formation parameter inversion, formation shear wave extraction and virtual well logging curves, a machine learning model is further created.

[0068] In an embodiment of the present invention, the creation of a machine learning model based on the oil well logging big data includes: generating a first training data set Xnew based on the oil well logging big data; obtaining a preset tree classification rule tree; generating a machine learning algorithm predict based on the first training data set Xnew and the preset tree classification rule tree, the machine learning algorithm predict being represented as: Ynew = predict(tree,Xnew), where Ynew is represented as a prediction result; or generating a corresponding data matrix X based on the simulated logging data; obtaining a response matrix Y corresponding to the data matrix X; generating a machine learning ens based on the data matrix X and the response matrix Y, the machine learning algorithm ens being represented as: ens = fitensemble(X, Y, model, numberens, learners), where numberens is represented as attribute information of the data matrix X, and learners is represented as attribute information of the response matrix Y.

[0069] In an embodiment of the present invention, the method also includes: generating corresponding simulation data based on the oil and gas monitoring metadata; optimizing the prediction model based on the simulation data to obtain an optimized model; processing the oil and gas monitoring metadata based on the optimized model to generate corresponding oil well logging prediction results.

[0070] In an embodiment of the present invention, simulation data is further generated based on the actually collected oil and gas monitoring metadata, and the generated prediction model is further verified and optimized through the simulation data, so that the prediction accuracy of the prediction model can be further improved.

[0071] However, in actual applications, the amount of oil logging data from different data sources (such as global data sources) is relatively large. Therefore, if a centralized data processing mode is adopted, the processing load of the oil logging prediction system will be greatly increased, the processing cost will be increased, and the processing efficiency will be reduced.

[0072] In order to solve the above technical problems, in an embodiment of the present invention, the oil well logging prediction system includes multiple distributed computing nodes, and the oil and gas monitoring metadata is processed based on the prediction model to generate corresponding oil well logging prediction results, including: generating at least one prediction task based on the oil and gas monitoring metadata; determining the matching distributed computing node corresponding to each prediction task; executing the corresponding prediction task based on the prediction model through the matching distributed computing node to generate corresponding oil well logging prediction results.

[0073] In a possible implementation, the oil well logging prediction system adopts a distributed data storage and processing system based on Hadoop, in which multiple distributed computing nodes are set up. After obtaining the oil and gas monitoring metadata, at least one corresponding prediction task is generated. For example, the corresponding prediction task can be generated according to the collection location, collection time and other information of the oil and gas monitoring metadata, and then the matching distributed computing node corresponding to each prediction task is determined. For example, the corresponding matching distributed computing node can be assigned to each prediction according to the node type of each distributed computing node and the task processing time required for each task, and then each matching distributed computing node is controlled to perform the corresponding prediction task based on the above-mentioned livestock prediction model and generate the corresponding oil well logging prediction result.

[0074] In the embodiment of the present invention, by adopting a distributed system to analyze and process a large amount of collected oil logging metadata, the system load of the oil logging prediction system can be effectively balanced, data congestion and task queuing can be avoided, the computing power of each distributed computing node can be effectively utilized, and the prediction efficiency can be improved.

[0075] The following is a description of the oil well logging prediction device based on big data provided by an embodiment of the present invention in conjunction with the accompanying drawings.

[0076] See also Figure 4 Based on the same inventive concept, an embodiment of the present invention provides an oil well logging prediction device based on big data, the device comprising: a big data unit, used to obtain oil well logging big data; a model creation unit, used to create a prediction model based on the oil well logging big data; a metadata acquisition unit, used to obtain oil and gas monitoring metadata; a prediction unit, used to process the oil and gas monitoring metadata based on the prediction model to generate corresponding oil well logging prediction results.

[0077] In an embodiment of the present invention, the big data unit includes: a data source determination module, which is used to determine the data source and the data format corresponding to the data source; a parsing tool determination module, which is used to obtain a data parsing tool corresponding to the data format, and the data parsing tool is connected to the oil well logging prediction system in a distributed manner; an instruction generation module, which is used to generate data import instructions; a data import module, which is used to control the data parsing tool to obtain raw data from the data source based on the data import instructions, and perform data parsing on the raw data based on the data format to generate corresponding oil well logging big data.

[0078] In an embodiment of the present invention, the model creation unit includes: an intermediate model creation module, which is used to create a physical algorithm model and a machine learning model based on the oil logging big data; and a prediction model creation module, which is used to create a prediction model based on the physical algorithm model and the machine learning model.

[0079] In an embodiment of the present invention, the intermediate model creation module includes a first model creation module, which is used to: obtain an initial neural network model and simulated logging data; calculate and determine characteristic information of the simulated logging data and the oil logging big data; train the initial neural network model based on the characteristic information to generate the physical algorithm model.

[0080] In an embodiment of the present invention, the initial neural network model is trained based on the feature information to generate the physical algorithm model, including: S231) training the initial neural network model based on the feature information to determine the maximum correlation coefficient and the minimum correlation coefficient of the initial neural network model; S232) determining an intermediate model based on the maximum correlation coefficient and the minimum correlation coefficient; S233) obtaining an output result of the intermediate model based on the simulated logging data, and determining whether the deviation between the output result and the oil logging big data meets a preset deviation requirement; S2341) if so, using the intermediate model as the physical algorithm model; S2342) otherwise, adjusting the simulated logging data to obtain adjusted data, training the intermediate model based on the adjusted data to obtain a new intermediate model, and continuing to execute step S233).

[0081] In an embodiment of the present invention, the first model creation module is also used to: determine a first neural network model and a second neural network model based on the physical algorithm model; perform forward training on the first neural network model based on the oil well logging big data to obtain a forward training model; perform reverse training on the second neural network model based on the oil well logging big data to obtain a reverse training model; obtain a preset encoder and a preset decoder, and generate a corresponding automatic decoder based on the preset encoder and the preset decoder; optimize the physical algorithm model based on the forward training model, the reverse training model and the automatic decoder to obtain an optimized physical algorithm model.

[0082] In an embodiment of the present invention, the intermediate model creation module includes a second model creation module, which is used to: generate a first training data set Xnew based on the oil well logging big data; obtain a preset tree classification rule tree; generate a machine learning algorithm predict based on the first training data set Xnew and the preset tree classification rule tree, and the machine learning algorithm predict is characterized by: Ynew = predict(tree,Xnew), where Ynew is characterized as a prediction result; or generate a corresponding data matrix X based on the simulated well logging data; obtain a response matrix Y corresponding to the data matrix X; generate a machine learning ens based on the data matrix X and the response matrix Y, and the machine learning algorithm ens is characterized by: ens = fitensemble(X, Y, model, numberens, learners), where numberens is characterized as attribute information of the data matrix X, and learners is characterized as attribute information of the response matrix Y.

[0083] In an embodiment of the present invention, the device also includes a model optimization unit, which is used to: generate corresponding simulation data based on the oil and gas monitoring metadata; optimize the prediction model based on the simulation data to obtain an optimized model; process the oil and gas monitoring metadata based on the optimized model to generate corresponding oil well logging prediction results.

[0084] In an embodiment of the present invention, the oil well logging prediction system includes multiple distributed computing nodes, and the prediction unit is used to: generate at least one prediction task based on the oil and gas monitoring metadata; determine the matching distributed computing node corresponding to each prediction task; and execute the corresponding prediction task based on the prediction model through the matching distributed computing node to generate the corresponding oil well logging prediction result.

[0085] Furthermore, an embodiment of the present invention also provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method described in the embodiment of the present invention is implemented.

[0086] The optional implementation modes of the embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above implementation modes. Within the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical scheme of the embodiments of the present invention, and these simple modifications all belong to the protection scope of the embodiments of the present invention.

[0087] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe various possible combinations.

[0088] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including a number of instructions to enable a single-chip microcomputer, a chip or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0089] In addition, various implementation modes of the embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the embodiments of the present invention, they should also be regarded as the contents disclosed by the embodiments of the present invention.

Claims

1. A petroleum logging prediction method based on big data, applied to a petroleum logging prediction system, characterized in that: The method comprises: Obtaining oil well logging big data; A prediction model is created based on the oil well logging big data, including: Creating a physical algorithm model and a machine learning model based on the oil well logging big data; Creating a prediction model based on the physical algorithm model and the machine learning model; Wherein, physical algorithm models are created based on the oil well logging big data, including: Obtaining initial neural network model and simulated logging data; Calculating and determining characteristic information of the simulated well logging data and the petroleum well logging big data; The initial neural network model is trained based on the feature information to generate the physical algorithm model, including: S231) training the initial neural network model based on the feature information to determine the maximum correlation coefficient and the minimum correlation coefficient of the initial neural network model; S232) determining an intermediate model based on the maximum correlation coefficient and the minimum correlation coefficient; S233) obtaining an output result of the intermediate model based on the simulated well logging data, and determining whether a deviation between the output result and the petroleum well logging big data meets a preset deviation requirement; S2341) If yes, use the intermediate model as the physical algorithm model; S2342) otherwise, adjusting the simulated logging data to obtain adjusted data, training the intermediate model based on the adjusted data to obtain a new intermediate model, and continuing to execute step S233); Obtaining oil and gas monitoring metadata; The oil and gas monitoring metadata is processed based on the prediction model to generate corresponding oil well logging prediction results.

2. The method according to claim 1, characterized in that The obtaining of oil well logging big data includes: Determine a data source and a data format corresponding to the data source; Acquire a data analysis tool corresponding to the data format, wherein the data analysis tool is connected to the oil well logging prediction system in a distributed manner; Generate data import instructions; Based on the data import instruction, the data analysis tool is controlled to obtain raw data from the data source, and the raw data is analyzed based on the data format to generate corresponding oil well logging big data.

3. The method according to claim 1, characterized in that The method further comprises: Determine a first neural network model and a second neural network model based on the physical algorithm model; Performing forward training on the first neural network model based on the oil well logging big data to obtain a forward training model; Performing reverse training on the second neural network model based on the oil well logging big data to obtain a reverse training model; Obtain a preset encoder and a preset decoder, and generate a corresponding automatic decoder based on the preset encoder and the preset decoder; The physical algorithm model is optimized based on the forward training model, the reverse training model and the automatic decoder to obtain an optimized physical algorithm model.

4. The method according to claim 1, characterized in that: The step of creating a machine learning model based on the petroleum well logging big data includes: Generate a first training data set Xnew based on the oil well logging big data; Get the preset tree classification rule tree; A machine learning algorithm predict is generated based on the first training data set Xnew and the preset tree classification rule tree. The machine learning algorithm predict is characterized as follows: Ynew = predict(tree,Xnew), Where Ynew represents the prediction result; or Generate a corresponding data matrix X based on the simulated logging data; Obtain a response matrix Y corresponding to the data matrix X; A machine learning algorithm ens is generated based on the data matrix X and the response matrix Y. The machine learning algorithm ens is characterized as follows: ens = fitensemble(X,Y,model,numberens,learners), Among them, numberens represents the attribute information of the data matrix X, and learners represents the attribute information of the response matrix Y.

5. The method according to claim 1, characterized in that The method further comprises: Generate corresponding simulation data based on the oil and gas monitoring metadata; Optimizing the prediction model based on the simulation data to obtain an optimized model; The oil and gas monitoring metadata is processed based on the optimized model to generate corresponding oil well logging prediction results.

6. The method according to claim 1, characterized in that The oil well logging prediction system includes a plurality of distributed computing nodes, and the oil and gas monitoring metadata is processed based on the prediction model to generate corresponding oil well logging prediction results, including: generating at least one prediction task based on the oil and gas monitoring metadata; Determine the matching distributed computing node corresponding to each prediction task; The matching distributed computing nodes execute corresponding prediction tasks based on the prediction model to generate corresponding oil well logging prediction results.

7. A petroleum logging prediction device based on big data, applied to a petroleum logging prediction system, characterized in that: The device comprises: Big data unit, used to obtain oil well logging big data; A model creation unit is used to create a prediction model based on the oil well logging big data, and the model creation unit includes: An intermediate model creation module, used to create a physical algorithm model and a machine learning model based on the oil well logging big data; A prediction model creation module, used to create a prediction model based on the physical algorithm model and the machine learning model; The intermediate model creation module includes a first model creation module, and the first model creation module is used to: Obtaining initial neural network model and simulated logging data; Calculating and determining characteristic information of the simulated well logging data and the petroleum well logging big data; The initial neural network model is trained based on the feature information to generate the physical algorithm model, including: S231) training the initial neural network model based on the feature information to determine the maximum correlation coefficient and the minimum correlation coefficient of the initial neural network model; S232) determining an intermediate model based on the maximum correlation coefficient and the minimum correlation coefficient; S233) obtaining an output result of the intermediate model based on the simulated well logging data, and determining whether a deviation between the output result and the petroleum well logging big data meets a preset deviation requirement; S2341) If yes, use the intermediate model as the physical algorithm model; S2342) otherwise, adjusting the simulated logging data to obtain adjusted data, training the intermediate model based on the adjusted data to obtain a new intermediate model, and continuing to execute step S233); A metadata acquisition unit, used for acquiring oil and gas monitoring metadata; The prediction unit is used to process the oil and gas monitoring metadata based on the prediction model to generate corresponding oil well logging prediction results.

8. The device according to claim 7, characterized in that The big data unit includes: A data source determination module, used to determine a data source and a data format corresponding to the data source; A parsing tool determination module, used to obtain a data parsing tool corresponding to the data format, wherein the data parsing tool is connected to the oil well logging prediction system in a distributed manner; An instruction generation module, used to generate data import instructions; The data import module is used to control the data analysis tool to obtain raw data from the data source based on the data import instruction, and to perform data analysis on the raw data based on the data format to generate corresponding oil well logging big data.

9. The device according to claim 7, characterized in that The first model creation module is also used for: Determine a first neural network model and a second neural network model based on the physical algorithm model; Performing forward training on the first neural network model based on the oil well logging big data to obtain a forward training model; Performing reverse training on the second neural network model based on the oil well logging big data to obtain a reverse training model; Obtain a preset encoder and a preset decoder, and generate a corresponding automatic decoder based on the preset encoder and the preset decoder; The physical algorithm model is optimized based on the forward training model, the reverse training model and the automatic decoder to obtain an optimized physical algorithm model.

10. The device according to claim 7, characterized in that The intermediate model creation module includes a second model creation module, and the second model creation module is used to: Generate a first training data set Xnew based on the oil well logging big data; Get the preset tree classification rule tree; A machine learning algorithm predict is generated based on the first training data set Xnew and the preset tree classification rule tree. The machine learning algorithm predict is characterized as follows: Ynew = predict(tree,Xnew), Where Ynew represents the prediction result; or Generate a corresponding data matrix X based on the simulated logging data; Obtain a response matrix Y corresponding to the data matrix X; A machine learning algorithm ens is generated based on the data matrix X and the response matrix Y. The machine learning algorithm ens is characterized as follows: ens = fitensemble(X,Y,model,numberens,learners), Among them, numberens represents the attribute information of the data matrix X, and learners represents the attribute information of the response matrix Y.

11. The device according to claim 7, characterized in that The device also includes a model optimization unit, which is used to: Generate corresponding simulation data based on the oil and gas monitoring metadata; Optimizing the prediction model based on the simulation data to obtain an optimized model; The oil and gas monitoring metadata is processed based on the optimized model to generate corresponding oil well logging prediction results.

12. The device according to claim 7, characterized in that The oil well logging prediction system includes a plurality of distributed computing nodes, and the prediction unit is used for: generating at least one prediction task based on the oil and gas monitoring metadata; Determine the matching distributed computing node corresponding to each prediction task; The matching distributed computing nodes execute corresponding prediction tasks based on the prediction model to generate corresponding oil well logging prediction results.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method described in any one of claims 1 to 6 is implemented.

Images