A production intelligent decision-making system and method for oil and gas fields
By designing an intelligent production decision-making system for the oil and gas field, the existing IT architecture cannot adapt to production needs has been solved, integrated data governance and intelligent decision-making have been achieved, digital and intelligent development of the oil and gas industry has been improved, and cost reduction and increase effect.
Patent Information
- Application Number
- CN202210209642.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-03-03
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2042-03-03
AI Technical Summary
The existing IT architecture cannot adapt to production needs in oilfield exploration and development, resulting in inaccurate data, untimely processing and inaccurate results, and inaccurate production decisions.
A production intelligent decision-making system for the oil and gas field was designed, including oil and gas multi-source heterogeneous data governance module, production intelligent decision-making platform business module, intelligent algorithm component library, functional module containerized packaging and automation management module, as well as models in different scenarios Custom development modules, through these modules, we realize integrated data governance, intelligent decision-making, intelligent computing and customized model development.
实现了油气领域数据一体化治理,增强了行业知识的紧密联系,解决了业务场景关联少的问题,通过智能决策系统提高了油气行业的数字化和智能化发展,实现了降本增效的目的。
Smart Images

Figure CN114676978B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of oil and gas development, and in particular to an intelligent production decision-making system and method for the oil and gas field. Background Art
[0002] The integration of big data, artificial intelligence and the oil and gas industry is the era of "new infrastructure" in the energy industry; one of the cores is also a major demand for the intelligent industrial upgrade of oil and gas companies. At present, traditional oil and gas companies have accumulated rich data resources, but they are faced with problems such as many types of data, complex structures, loose industry knowledge connections, and few business scenarios. Big data technology can process and quickly extract massive amounts of oil and gas data, and artificial intelligence algorithms can professionally integrate and accurately mine data. Therefore, it is urgent to carry out innovative research on big data and artificial intelligence technology in the oil and gas field, develop a smart production decision-making platform for the oil and gas field, and form an integrated framework based on the pipeline aggregation, integration, processing and fusion of multi-source heterogeneous data bodies of oil and gas, integrating multi-scenario production control and prediction technologies based on big data and artificial intelligence, and realizing the digitalization and intelligent transformation of the oil and gas industry and the cost reduction and efficiency improvement of oil and gas companies.
[0003] In the digital age, the oil industry's oilfield exploration and development sector has long been implementing a data business model compiled in the IT era, called POSC. It has been implemented as a data business model for oilfield data construction and information management systems, handling the operational relationship between data and business.
[0004] However, in oilfield exploration and development, the existing IT architecture can no longer meet production needs and faces problems such as a large number of data types, complex structures, loose connections with industry knowledge, and few business scenario associations. Summary of the invention
[0005] The technical problem to be solved by the present invention is to provide a production intelligent decision-making system and method for the oil and gas field, which can solve the problem that the above-mentioned existing IT architecture can no longer adapt to production needs, the obtained data is inaccurate, the data processing is not timely, the processing results are inaccurate, and normal production decisions cannot be met.
[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:
[0007] On the one hand, a production intelligent decision-making system for the oil and gas field is provided, the system comprising:
[0008] The oil and gas multi-source heterogeneous data governance module is used to achieve integrated data governance in the oil and gas field;
[0009] The business module of the oil and gas production intelligent decision-making platform is used to realize intelligent decision-making, intelligent storage, transportation, gathering and transmission, intelligent operation and sales, and system management and maintenance of oil and gas production;
[0010] Intelligent algorithm component library, which is used to provide basic algorithms and customized intelligent algorithms developed based on specific scenarios;
[0011] Functional module containerization and automation management modules are used to containerize oil and gas multi-source heterogeneous data, intelligent algorithm component library, and intelligent service component library, and to perform unified container scheduling and management;
[0012] Model custom development modules for different scenarios are used to build specialized models for different scenarios.
[0013] In an optional embodiment, the production intelligence decision module includes:
[0014] The equipment layer is used to provide a number of infrastructure resource services including network transmission, cloud computing, cloud storage, general-purpose big data processing environment, high-performance computing grid, artificial intelligence computing and data services;
[0015] The environment support layer is used to provide environment support for processing massive oil and gas multi-source heterogeneous data bodies;
[0016] The data middle platform layer includes data collection unit, data processing unit and data calculation unit;
[0017] Microservice application layer, used to implement applications in various scenarios based on microservice technology;
[0018] The user layer is used to enable different users to use the system microservice functions and oil and gas data.
[0019] In an optional embodiment, the oil and gas multi-source heterogeneous data governance module includes
[0020] The original database is used to store multi-source heterogeneous data in the oil and gas field;
[0021] The original oil and gas big data resource pool is used to classify and package the stored multi-source heterogeneous data in the oil and gas field;
[0022] Data processing unit, used to standardize the raw data according to the existing data standards and custom standards of the oil and gas industry;
[0023] The data service unit is used to directly extract the data in the original database and open a data service channel to the outside world, so that the system can quickly call the data in the original database according to demand.
[0024] In an optional embodiment, the oil and gas production intelligent decision-making platform business module includes:
[0025] The intelligent decision-making unit for upstream oil and gas production is used to realize integrated intelligent algorithm services from oil field development, intelligent prediction, effect evaluation, parameter optimization, and intelligent decision-making;
[0026] The smart storage, transportation and gathering unit for oil and gas midstream is used for midstream service personnel and managers to grasp the dynamic information of oil and gas and provide guidance for oil and gas storage and transportation;
[0027] The smart operation and sales unit for downstream oil and gas is used to make intelligent forecasts based on downstream oil and gas facility management, market price analysis, market supply and demand, and user conditions, and to achieve smart guidance for oil and gas trade;
[0028] The system's management and operation and maintenance unit is used to control the operating permissions of different users for each module in the system.
[0029] In an optional embodiment, the intelligent algorithm component library includes:
[0030] A basic algorithm library consisting of machine learning and classic simulation algorithms.
[0031] An intelligent algorithm library customized for specific scenarios.
[0032] In an optional embodiment, the functional module containerization packaging and automatic management module includes:
[0033] Distributed storage unit, cache read-write unit, interface authentication unit, unified authentication unit, access control unit, service control unit, device service unit, user service unit and analysis interface unit.
[0034] In an optional embodiment, the upstream oil and gas production intelligent decision-making unit includes: an injection and production parameter intelligent optimization unit, a new well target intelligent decision-making unit, a well logging interpretation intelligent analysis unit, a fracturing effect intelligent evaluation unit, a development effect intelligent regulation unit, an oil and gas production intelligent prediction unit, an oil and gas production capacity intelligent calibration unit, and a reservoir property intelligent prediction unit.
[0035] In an optional embodiment, the midstream intelligent oil and gas storage, transportation and gathering unit includes: an LNG gas storage management unit, an LNG receiving station management unit, an LNG plant management unit and a natural gas pipeline transportation unit.
[0036] In an optional embodiment, the downstream oil and gas intelligent operation and sales unit includes a natural gas facility intelligent management unit, a market price intelligent analysis unit, a market supply and demand intelligent forecasting unit, a natural gas trade intelligent guidance unit, a natural gas user unit and an underground gas storage management unit.
[0037] On the other hand, a production intelligent decision-making method for the oil and gas field is provided, the method comprising:
[0038] Through the oil and gas multi-source heterogeneous data governance module, the original data is uniformly standardized, cleaned, supplemented, associated and integrated according to the existing data standards and custom standards of the oil and gas industry, and then placed in the data lake or data pool;
[0039] Comprehensively sort out upstream, midstream and downstream oil and gas businesses based on expert knowledge, analyze the correlation of data in different scenarios, establish an oil and gas data knowledge graph, aggregate all sorted businesses and build a business architecture;
[0040] Integrate traditional numerical simulation methods and classic machine learning methods, and can customize machine learning models that integrate physical constraints according to different application scenarios, and perform intelligent calculations based on specific data and specific businesses through the intelligent algorithm component library;
[0041] Through the containerization of functional modules and the automation management module, oil and gas multi-source heterogeneous data, intelligent algorithm component library, and intelligent service component library are containerized and packaged, and containers are uniformly scheduled and managed;
[0042] According to the results of intelligent calculation, build specialized models for different scenarios through custom development modules of models in different scenarios;
[0043] Through the business modules of the oil and gas production intelligent decision-making platform, intelligent decision-making, intelligent storage, transportation, gathering and distribution, intelligent operation and sales, as well as system management and maintenance of oil and gas production can be realized.
[0044] The beneficial effects of the above technical solution provided by the embodiment of the present invention are as follows:
[0045] The system provided by the embodiment of the present invention establishes an integrated framework for pipeline aggregation, integration, processing and fusion of multi-source heterogeneous data bodies of oil and gas through the oil and gas multi-source heterogeneous data governance module, thereby realizing integrated governance of data in the oil and gas field with many types and complex structures; realizes intelligent decision-making, intelligent storage and transportation, intelligent operation and sales, and system management and operation and maintenance of oil and gas production through the business module of the oil and gas production intelligent decision-making platform, thereby enhancing the close connection of industry knowledge; integrates multi-scenario production control and prediction technology based on big data and artificial intelligence through the intelligent algorithm component library, and forms an end-to-end code-free custom development mode for the knowledge graph of the oil and gas field and models in different scenarios, providing basic algorithms and intelligent algorithms customized for specific scenarios; through the containerization packaging of functional modules and the automatic management module, the oil and gas multi-source heterogeneous data, the intelligent algorithm component library, and the intelligent service component library are containerized and packaged, and the containers are uniformly scheduled and managed; through the custom development modules of models in different scenarios, specialized models for different scenarios are built, which solves the problem of few business scenario associations in the production of the oil and gas industry, and realizes intelligent decision-making for production in the oil and gas field through the cooperation of the above modules. The system provided by the embodiment of the present invention plays an important role in the digital and intelligent development of the oil and gas industry, and achieves the goal of reducing costs and increasing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 It is a simple block diagram of the production intelligent decision-making system for the oil and gas field;
[0047] Figure 2 This is the overall framework diagram of the production intelligent decision-making system for the oil and gas field;
[0048] Figure 3 Oil and gas multi-source heterogeneous data governance flow chart;
[0049] Figure 4 Schematic diagram of the knowledge graph structure for oil well blockage scenarios;
[0050] Figure 5 This is the business architecture diagram of the production intelligent decision-making system for the oil and gas field;
[0051] Figure 6 This is a schematic diagram of the business modules of the production intelligent decision-making system for the oil and gas field;
[0052] Figure 7 It is a schematic diagram of the artificial intelligence algorithm library, the classic model algorithm library, and the fusion algorithm based on specific scenarios;
[0053] Figure 8 This is a schematic diagram of different scenario modules of the production intelligent decision-making system for the oil and gas field;
[0054] Fig. 9Model building interface diagram for algorithm custom editing module;
[0055] Fig.10 A visualization interface for model results of the algorithm custom editing module. DETAILED DESCRIPTION
[0056] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.
[0057] In the digital age, the oil industry has long been implementing a data business model compiled in the IT era in the field of oilfield exploration and development, called POSC, which has been implemented as a data business model for oilfield data construction and information management systems, and handles the operational relationship between data and business. However, in oilfield exploration and development, the existing IT architecture can no longer meet production needs and faces problems such as a large number of data types, complex structures, loose industry knowledge connections, and few business scenario connections. In view of this, an embodiment of the present invention provides a production intelligent decision-making system and decision-making method for the oil and gas field to solve the above technical problems.
[0058] On the one hand, please also refer to Figure 1 and Figure 2 The embodiment of the present invention provides a production intelligent decision-making system for the oil and gas field, the system comprising:
[0059] The oil and gas multi-source heterogeneous data governance module 101 is used to achieve integrated data governance in the oil and gas field.
[0060] The oil and gas production intelligent decision-making platform business module 102 is used to realize intelligent decision-making, intelligent storage, transportation, gathering and transmission, intelligent operation and sales, and system management and maintenance of oil and gas production.
[0061] The intelligent algorithm component library 103 is used to provide basic algorithms and intelligent algorithms customized for specific scenarios.
[0062] The functional module containerization packaging and automated management module 104 is used to containerize the oil and gas multi-source heterogeneous data, the intelligent algorithm component library 103, and the intelligent service component library, and to perform unified scheduling and management of containers.
[0063] The model customization development module 105 for different scenarios is used to build specialized models for different scenarios.
[0064] The beneficial effects of the above technical solution provided by the embodiment of the present invention are as follows:
[0065] The system provided by the embodiment of the present invention establishes an integrated framework for pipeline aggregation, integration, processing and fusion of oil and gas multi-source heterogeneous data bodies through the oil and gas multi-source heterogeneous data governance module 101, realizing integrated governance of oil and gas field data with multiple data types and complex structures; realizes oil and gas production intelligent decision-making, intelligent storage and transportation, intelligent operation and sales, and system management and maintenance through the oil and gas production intelligent decision-making platform business module 102, and enhances the close connection of industry knowledge; integrates multi-scenario production control and prediction technology based on big data and artificial intelligence through the intelligent algorithm component library 103, and forms a knowledge graph for the oil and gas field. The system provides basic algorithms and intelligent algorithms customized for specific scenarios; through the containerization of functional modules and the automated management module 104, the oil and gas multi-source heterogeneous data, the intelligent algorithm component library 103, and the intelligent service component library are containerized and packaged for unified container scheduling and management; through the custom development module 105 of the model in different scenarios, specialized models for different scenarios are built to solve the problems of few business scenario associations in the production of the oil and gas industry, and the above modules are used to cooperate to realize intelligent decision-making for production in the oil and gas field. The system provided by the embodiment of the present invention plays an important role in the digitalization and intelligent development of the oil and gas industry, and achieves the purpose of reducing costs and increasing efficiency.
[0066] The system provided by the embodiment of the present invention is further explained and described below through optional embodiments.
[0067] In an optional embodiment, the production intelligence decision module includes:
[0068] The equipment layer is used to provide a number of infrastructure resource services including network transmission, cloud computing, cloud storage, general-purpose big data processing environment, high-performance computing grid, artificial intelligence computing and data services.
[0069] The equipment layer includes China Science and Technology Cloud servers, which meet the Windows, Linux and Unix operating system environments, and provide a number of infrastructure resource services including network transmission, cloud computing, cloud storage, general-purpose big data processing environment, high-performance computing grid, artificial intelligence computing and data services.
[0070] The environmental support layer is used to provide environmental support for processing massive oil and gas multi-source heterogeneous data.
[0071] The environmental support layer includes a big data full-stack component management system, which uses Hadoop distributed storage as the basis, Apache Spark and Storm as the hybrid computing framework, and combines data stream processing systems such as Piflow to build an environmental support layer for processing massive oil and gas multi-source heterogeneous data.
[0072] It should be noted that Hadoop is a distributed system infrastructure developed by the Apache Foundation. Users can develop distributed programs without understanding the underlying details of the distribution, and make full use of the power of the cluster for high-speed computing and storage.
[0073] The data middle platform layer includes data collection unit, data processing unit and data calculation unit.
[0074] It should be noted that the data middle platform layer is the core of the production intelligent decision-making system for the oil and gas field, which includes data acquisition unit, data processing unit and data calculation unit.
[0075] Furthermore, the source of data collected by the data acquisition unit is open knowledge data, including monitoring experimental data through oil and gas professional databases (Oracle, Mysql, SQL, etc.), oil and gas numerical simulation software (matlab, CMG, Eclipse, etc.) and oil and gas Internet of Things (Scada, etc.), including the ability to extract structured data such as dynamic production data and seismic inversion data; unstructured data, such as logging curves and digital core images, semi-structured data such as seismic interpretation and logging reports, etc., to form an oil and gas big data resource pool, numerical simulation data, etc.
[0076] The data processing unit mainly performs data cleaning and data fusion on multi-source heterogeneous data bodies of oil and gas, cleans missing and abnormal data using local cleaning, global cleaning and statistical methods, and correlates and fuses the cleaned data by establishing a professional knowledge map for the oil and gas field, forming a core research database, method library, achievement library and expert knowledge base.
[0077] Data computing mainly includes data batch processing computing based on MapReduce in Hadoop, Spark streaming computing, independently developed machine learning algorithm library and oil and gas numerical simulation method library, which can realize batch processing / stream processing and offline / real-time computing of massive oil and gas data.
[0078] The microservice application layer is used to implement applications in various scenarios based on microservice technology.
[0079] The microservice application layer implements application modules in various scenarios based on microservice technology.
[0080] In an optional embodiment, the microservice application layer includes an upstream oil and gas production and development unit, a midstream oil and gas intelligent storage, transportation and gathering unit, a downstream oil and gas intelligent sales unit, and a management unit.
[0081] The upstream oil and gas production and development unit includes application modules suitable for on-site developers, such as reservoir property prediction, oil and gas capacity calibration, oil and gas production prediction, and effect evaluation, to provide targeted guidance on oilfield production and development, well logging, etc. The midstream intelligent storage, transportation, and gathering unit of oil and gas includes LNG gas storage management, LNG receiving station management, LNG plant management, and intelligent operation and maintenance of natural gas pipelines, which provide effective guidance for oil and gas storage, transportation, and gathering for midstream service personnel and management personnel; the downstream intelligent sales unit of oil and gas includes modules such as natural gas infrastructure management, market supply and demand forecasting, and price analysis, which is a professional microservice application for downstream oil and gas sales and market analysts; the management unit is specifically for the system management and operation and maintenance of the oil and gas production intelligent decision-making platform, and provides customized management and authority allocation for different departments and personnel, including user management, role management, system announcements, and tenant management. At the same time, the microservice application layer also has a special API service call interface for connecting to the oil and gas expert system or tenant experience system.
[0082] User layer 5 is aimed at government agencies, researchers and managers, enabling different personnel to use the platform microservice functions and oil and gas data.
[0083] The user layer is used to enable different users to use the system microservice functions and oil and gas data. As an example, Figure 2 As shown, it includes government agencies, researchers, and managers.
[0084] See also Figure 3 In an optional embodiment, the oil and gas multi-source heterogeneous data governance module 101 includes: an original database for storing multi-source heterogeneous data in the oil and gas field.
[0085] The production decision-making platform in the oil and gas field is connected to the original database at the bottom layer, including the oil and gas research database of universities, the professional database of oil companies, the professional database of natural gas companies and the economic evaluation database. The database types include Oracle, Mysql, SQL, etc.
[0086] The original oil and gas big data resource pool is used to classify and package the stored multi-source heterogeneous data in the oil and gas field.
[0087] The original oil and gas big data resource pool classifies and packages the data of the original database into 9 categories, namely exploration, logging, mud logging, production, drilling, operation and maintenance, experimental simulation, economic evaluation and others, forming an oil and gas big data resource pool based on original data and breaking down the barriers between various institutions.
[0088] The data processing unit is used to perform unified and standardized processing on the original data according to the existing data standards and customized standards of the oil and gas industry.
[0089] The data processing unit standardizes the raw data according to the existing data standards and custom standards of the oil and gas industry, and uses the global cleaning algorithm to search for missing and abnormal problems in the data, and completes and removes noise from the data. The cleaned data is evaluated through their respective quality assessment systems. Data that fails the evaluation needs to be cleaned again according to its field and data characteristics until it passes the data assessment, so as to improve the data quality. Then, the cleaned data entities are used as nodes, and the relationships between entities are used as edges, such as Figure 4 As shown in the figure, a knowledge graph for the oil and gas field was constructed through the experts' experience knowledge, and a graph database for the oil and gas field was established.
[0090] The data service unit is used to directly extract data from the original database and open data service channels to the outside world, so that the system can quickly call data from the original database according to needs.
[0091] The data service unit can directly extract data from the original database and open the data service channel to the outside world. Each module in the application layer can quickly call the original data interface according to its own needs. Finally, the application layer includes microservice applications, algorithm calls, visualization services, and external services. It can directly call the core data that has passed the data quality assessment, the graph database built by the oil and gas knowledge graph, and the original oil and gas big data resource pool. The oil and gas multi-source heterogeneous data governance module 101 realizes an integrated data governance system of data collection-cleaning-extraction-fusion. For example, the professional knowledge graph for oil well blockage problems.
[0092] In an optional embodiment, if Figure 5 and Figure 6 As shown, the oil and gas production intelligent decision-making platform business module 102 includes:
[0093] The intelligent decision-making unit for upstream oil and gas production is used to provide integrated intelligent algorithm services from oil field development, intelligent prediction, effect evaluation, parameter optimization to intelligent decision-making.
[0094] In an optional embodiment, the intelligent decision-making unit for upstream oil and gas production includes: an injection and production parameter intelligent optimization unit, a new well target intelligent decision-making unit, a well logging interpretation intelligent analysis unit, a fracturing effect intelligent evaluation unit, a development effect intelligent regulation unit, an oil and gas production intelligent prediction unit, an oil and gas production capacity intelligent calibration unit, and a reservoir property intelligent prediction unit.
[0095] The intelligent decision-making unit for upstream oil and gas production mainly includes eight integrated algorithm modules. Each functional module contains intelligent algorithms and application examples corresponding to different scenarios, forming an integrated intelligent algorithm service system from oilfield development-intelligent prediction-effect evaluation-parameter optimization-intelligent decision-making.
[0096] The intelligent storage, transportation and gathering unit for midstream oil and gas is used for midstream service personnel and managers to grasp the dynamic information of oil and gas and provide guidance for oil and gas storage and transportation.
[0097] In an optional embodiment, the smart storage, transportation and gathering unit for oil and gas midstream includes: an LNG gas storage management unit, an LNG receiving station management unit, an LNG plant management unit and a natural gas pipeline transportation unit. The smart storage, transportation and gathering unit for oil and gas midstream is aimed at midstream service personnel and management personnel, and can grasp the dynamic information of oil and gas from collection, transportation and reception, and provide guidance for oil and gas storage and transportation.
[0098] The downstream oil and gas intelligent operation and sales unit is used to make intelligent forecasts based on downstream oil and gas facility management, market price analysis, market supply and demand, and user conditions, and to achieve intelligent guidance for oil and gas trade.
[0099] In an optional embodiment, the downstream oil and gas intelligent operation and sales unit includes a natural gas facility intelligent management unit, a market price intelligent analysis unit, a market supply and demand intelligent forecasting unit, a natural gas trade intelligent guidance unit, a natural gas user unit and an underground gas storage management unit.
[0100] Through natural gas infrastructure management, the smart operation and sales unit of the downstream oil and gas industry can obtain detailed data on gas stations, refueling stations, power plants and city gas, and use basic data for intelligent market price analysis and market supply and demand forecasting to achieve intelligent guidance for natural gas trade.
[0101] The system's management and operation unit is used to control the operating permissions of different users for each module in the system.
[0102] The system's management and operation and maintenance units mainly include user management, role management, department management, position management, tenant management and data dictionary. They can control the operating permissions of personnel from different departments and positions on various modules of the production intelligent decision-making system for the oil and gas field, thereby allowing upstream developers, downstream sales personnel, system testers and system leasing personnel to operate modules in their respective fields without interfering with each other.
[0103] In an optional embodiment, the intelligent algorithm component library 103 includes:
[0104] A basic algorithm library consisting of machine learning and classic simulation algorithms.
[0105] like Figure 7 As shown. Intelligent algorithm component library 103 is Figure 7The integrated algorithm diagram of physical laws and AI in the oil and gas field is mainly composed of two parts. One part is the basic algorithm library composed of machine learning and classical simulation algorithms. The machine learning library is based on the Python environment, and directly calls support vector machines, random forests, neural networks, naive Bayes and other intelligent algorithms through learning libraries such as sklearn, keras, and TensorFlow; the classical algorithm is based on MATLAB and Python language, and independently developed classical algorithms such as finite element, finite difference and finite volume. At the same time, it can also call already integrated algorithms such as finite element, finite volume, finite difference, Kriging difference, cubic spline interpolation and function fitting. Figure 8 This is a schematic diagram of different scenario modules of the intelligent decision-making system for oil and gas production.
[0106] An intelligent algorithm library customized for specific scenarios.
[0107] The core idea of intelligent algorithms customized for specific scenarios is to use machine learning algorithms (support vector machines, random forests, decision trees, artificial neural networks, Xgboost, K nearest neighbors, naive Bayes, etc.) as the basis, and integrate the control equations, boundary conditions, and initial conditions in the classic simulation algorithm as constraints into the machine learning algorithm in the form of building a new loss function. Based on the development model of the above algorithms, special algorithms such as intelligent prediction of reservoir physical properties, intelligent calibration of oil and gas production capacity, and intelligent evaluation of fracturing effects have been established. In addition, considering the programming ability of personnel in the oil and gas field, an end-to-end code-free operation platform has been established, and the basic algorithms and customized algorithms have been integrated and packaged into structured algorithm modules (H5, PTH, T7, PKL, MAT, etc. formats), which are called by dragging and dropping. The front end triggers the algorithm event by clicking the drag icon, and the back end automatically builds the model according to the front end response, realizing the rapid assembly and pipeline calling mode of the algorithm.
[0108] In an optional embodiment, the function module containerization packaging and automatic management module 104 includes:
[0109] Distributed storage unit, cache read-write unit, interface authentication unit, unified authentication unit, access control unit, service control unit, device service unit, user service unit and analysis interface unit.
[0110] Furthermore, the functional module containerization encapsulation and automated management module 104 adopts the Docker technology under the cloud native architecture to containerize production operation data, intelligent algorithms, intelligent service component libraries, etc., mainly including distributed storage, cache reading and writing, interface authentication, unified authentication, access control, service control, device services, user services, analysis interfaces, etc., to form an independent unit for application deployment and achieve high-level resource isolation; the Kubernetes technology under the cloud native architecture is adopted for unified container scheduling and management, and containers are organized into groups, and load balancing between containers is provided, so that the container scale can be expanded or contracted at any time, and the rolling release function is adopted to promote the application from one environment to another without stopping the application, so as to form a cloud native model management for distributed model output, reading and utilization. With "Function as a Service" as the foothold, through loose coupling, the Dapr framework under the cloud native architecture is adopted to organize and separate interface authentication, website authentication, access control, routing control, intelligent algorithm library, device service, knowledge service, and model service according to business capabilities, so that the code can be updated more easily, and scaled independently from each other, so as to achieve rapid construction and coordinated call of microservice applications.
[0111] In an optional embodiment, the model customization development module 105 for different scenarios is used to build specialized models for different scenarios.
[0112] like Fig. 9 As shown, Fig. 9 Model building interface diagram for algorithm custom editing module; Fig.10 It is a visualization interface for model results of the algorithm custom editing module. After logging into the platform, users can obtain the required data set through local upload, online import and web crawler. The uploaded data will be temporarily saved in the background database and assigned specific label codes. The production intelligent decision-making system for the oil and gas field takes into account the complexity of data and algorithms in various oil and gas industries, and builds a model editor. Import the required data into the editor, select the corresponding data cleaning algorithm to improve data quality, and build specialized models for different scenarios by dragging and dropping based on the packaged algorithm component library. It helps users better build complex autonomous learning models in a graphical way, and can see the output results and application effects of the model through visualization technology, such as Fig.10 As shown. The self-built models can also be packaged in a process-based manner to facilitate one-click calling next time. At the same time, these customized algorithm models have API interfaces that allow other software to call them, and internal data can also be accessed through URLs.
[0113] On the other hand, a production intelligent decision-making method for the oil and gas field is provided, the method comprising:
[0114] Through the oil and gas multi-source heterogeneous data governance module 101, the original data is uniformly standardized, cleaned, supplemented, associated and integrated according to the existing data standards and custom standards of the oil and gas industry, and then placed in the data lake or data pool;
[0115] Comprehensively sort out upstream, midstream and downstream oil and gas businesses based on expert knowledge, analyze the correlation of data in different scenarios, establish an oil and gas data knowledge graph, aggregate all sorted businesses and build a business architecture;
[0116] Integrate traditional numerical simulation methods and classic machine learning methods, and can customize machine learning models that integrate physical constraints according to different application scenarios, and perform intelligent calculations based on specific data and specific businesses through the intelligent algorithm component library;
[0117] Through the containerization of functional modules and the automation management module 104, oil and gas multi-source heterogeneous data, intelligent algorithm component library, and intelligent service component library are containerized and packaged, and containers are uniformly scheduled and managed;
[0118] According to the results of intelligent calculation, build specialized models for different scenarios through custom development modules of models in different scenarios;
[0119] Through the business modules of the oil and gas production intelligent decision-making platform, intelligent decision-making, intelligent storage, transportation, gathering and distribution, intelligent operation and sales, as well as system management and maintenance of oil and gas production can be realized.
[0120] The method provided in the embodiment of the present invention covers various modules of upstream oil and gas production and development, midstream storage and transportation, and downstream production and sales, and processes multi-source heterogeneous data of oil and gas in a streamlined and automated manner, thereby improving the work efficiency of management personnel and service personnel.
[0121] A governance system for multi-source heterogeneous data in oil and gas is proposed, which can break down industry barriers, efficiently integrate databases of different types from different institutions, and perform data standardization, cleaning and integration. Based on expert experience, a knowledge graph in the oil and gas field is established, realizing an integrated data governance system of data collection-cleaning-extraction-fusion, and strengthening the integration of industry knowledge.
[0122] A platform business architecture for the oil and gas sector was proposed, which integrated the upstream oil and gas production module, midstream transportation module, downstream sales module and the system management module within the platform to form a complete one-stop service model, thereby improving the overall decision-making ability of managers.
[0123] An intelligent algorithm component library 103 for the oil and gas field was proposed. It not only integrated the machine learning algorithm and the classical reservoir simulation algorithm into the basic algorithm library, but also deeply combined the two algorithms to form special algorithms based on different scenarios. It constructed a special algorithm component library for the oil and gas field, realized a code-free packaging and calling mode, and improved the modeling efficiency of scientific researchers.
[0124] A containerized packaging and automated management model for each upstream, midstream and downstream module of the oil and gas industry is proposed. Oil and gas data operations, intelligent algorithm calls and special algorithm displays are containerized and independently deployed to achieve rapid construction and coordinated calls of microservice applications, thereby enhancing the scalability and portability of the application layer.
[0125] A custom model development model was proposed, which utilized the platform's internal algorithm library and knowledge graph data for the oil and gas field to form a drag-and-drop process-based model editor, thereby improving the work efficiency of developers.
[0126] The above are preferred embodiments of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.
Claims
1. A production intelligent decision-making system for the oil and gas field, characterized in that: The system comprises: The oil and gas multi-source heterogeneous data management module is used to achieve integrated data management of data collection, cleaning, extraction and fusion; The business module of the oil and gas production intelligent decision-making platform is used to realize intelligent decision-making, intelligent storage, transportation, gathering and transmission, intelligent operation and sales, and system management and maintenance of oil and gas production; Intelligent algorithm component library, which is used to provide basic algorithms and customized intelligent algorithms developed based on specific scenarios; Functional module containerization and automation management modules are used to containerize oil and gas multi-source heterogeneous data, intelligent algorithm component library, and intelligent service component library, and to perform unified container scheduling and management; Model custom development modules for different scenarios, used to build specialized models for different scenarios; The module architecture of the business module of the oil and gas production intelligent decision-making platform includes: The equipment layer is used to provide a number of infrastructure resource services including network transmission, cloud computing, cloud storage, general-purpose big data processing environment, high-performance computing grid, artificial intelligence computing and data services; The environment support layer is used to provide environment support for processing massive oil and gas multi-source heterogeneous data bodies; The data middle platform layer includes data collection unit, data processing unit and data calculation unit; Microservice application layer, used to implement applications in various scenarios based on microservice technology; The user layer is used to enable different users to use the system microservice functions and oil and gas data; The data processing unit performs data cleaning and data fusion on the oil and gas multi-source heterogeneous data body, cleans the missing and abnormal data using local cleaning, global cleaning and statistical methods, and associates and fuses the cleaned data by establishing a professional knowledge map for the oil and gas field to form a core research database, method library, achievement library and expert knowledge base; The data computing unit includes data batch processing computing based on MapReduce in Hadoop, Spark streaming stream processing computing, self-developed machine learning algorithm library and oil and gas numerical simulation method library, realizing batch processing / stream processing and offline / real-time computing of massive oil and gas data; The oil and gas multi-source heterogeneous data governance module includes: The original database is used to store multi-source heterogeneous data in the oil and gas field; The original oil and gas big data resource pool is used to classify and package the stored multi-source heterogeneous data in the oil and gas field; Data processing unit, used to standardize the raw data according to the existing data standards and custom standards of the oil and gas industry; A data service unit, used to directly extract data from the original database and open a data service channel to the outside, so that the system can quickly call the data in the original database according to demand; The original oil and gas big data resource pool classifies and packages the data of the original database into 9 categories, namely exploration, well logging, mud logging, production, drilling, operation and maintenance, experimental simulation, economic evaluation and others, to form an oil and gas big data resource pool based on the original data; The data processing unit performs unified standardization processing on the original data according to the existing data standards and custom standards of the oil and gas industry, and uses a global cleaning algorithm to search for missing and abnormal problems in the data, and completes and removes noise from the data. The cleaned data is evaluated through their respective quality assessment systems. Data that fails the evaluation needs to be cleaned again according to its field and data characteristics until it passes the data evaluation, so as to improve the data quality; then, the cleaned data entities are used as nodes, and the relationships between entities are used as edges. A knowledge graph for the oil and gas field is constructed through the experience and knowledge of experts, and a graph database for the oil and gas field is established; The data service unit directly extracts data from the original database and opens a data service channel to the outside world; The business process of the business module of the oil and gas production intelligent decision-making platform includes: The intelligent decision-making unit for upstream oil and gas production is used to realize integrated intelligent algorithm services from oil field development, intelligent prediction, effect evaluation, parameter optimization, and intelligent decision-making; The smart storage, transportation and gathering unit for midstream oil and gas is used for midstream service personnel and managers to grasp the dynamic information of oil and gas and provide guidance for oil and gas storage and transportation; The smart operation and sales unit for downstream oil and gas is used to make intelligent forecasts based on downstream oil and gas facility management, market price analysis, market supply and demand, and user conditions, and to achieve smart guidance for oil and gas trade; The system's management and operation and maintenance unit is used to control the operation rights of different users for each module in the system; The intelligent decision-making unit for upstream oil and gas production includes: intelligent optimization unit for injection and production parameters, intelligent decision-making unit for new well targets, intelligent analysis unit for well logging interpretation, intelligent evaluation unit for fracturing effect, intelligent regulation unit for development effect, intelligent prediction unit for oil and gas production, intelligent calibration unit for oil and gas production capacity, and intelligent prediction unit for reservoir physical properties; The oil and gas midstream intelligent storage, transportation and gathering unit includes: LNG gas storage management unit, LNG receiving station management unit, LNG plant management unit and natural gas pipeline transportation unit; The oil and gas downstream intelligent operation and sales unit includes a natural gas facility intelligent management unit, a market price intelligent analysis unit, a market supply and demand intelligent prediction unit, a natural gas trade intelligent guidance unit, a natural gas user unit, and an underground gas storage management unit; The intelligent algorithm component library is an integrated algorithm diagram for the physical laws and AI in the oil and gas field. It consists of two parts: one is a basic algorithm library composed of machine learning and classical simulation algorithms, and the other is an intelligent algorithm library customized for specific scenarios. Among them, the core idea of the intelligent algorithm customized for specific scenarios is to use the machine learning algorithm as the basis, and take the control equations, boundary conditions, and initial conditions in the classic simulation algorithm as constraints, and integrate them into the machine learning algorithm in the form of constructing a new loss function. Based on the development model of the above algorithm, special algorithms for intelligent prediction of reservoir properties, intelligent calibration of oil and gas production capacity, and intelligent evaluation of fracturing effects have been established; and an end-to-end code-free operation platform has been established, which integrates and packages the basic algorithms and customized algorithms into structured algorithm modules, which are called in a drag-and-drop manner. The front end triggers the algorithm event by clicking the drag icon, and the back end automatically builds the model according to the front end response, realizing the rapid assembly and pipeline calling mode of the algorithm.
2. The production intelligent decision-making system for the oil and gas field according to claim 1 is characterized in that: The functional module containerization packaging and automatic management module include: Distributed storage unit, cache read-write unit, interface authentication unit, unified authentication unit, access control unit, service control unit, device service unit, user service unit and analysis interface unit.
3. A production intelligent decision-making method for the oil and gas field, characterized in that: The method is carried out by the production intelligent decision-making system for the oil and gas field according to claim 1 or 2, and the method comprises: Through the oil and gas multi-source heterogeneous data governance module, the original data is uniformly standardized, cleaned, supplemented, associated and integrated according to the existing data standards and custom standards of the oil and gas industry, and then placed in the data lake or data pool; Comprehensively sort out upstream, midstream and downstream oil and gas businesses based on expert knowledge, analyze the correlation of data in different scenarios, establish an oil and gas data knowledge graph, aggregate all sorted businesses and build a business architecture; Integrate traditional numerical simulation methods and classic machine learning methods, and customize machine learning models that integrate physical constraints according to different application scenarios, and perform intelligent calculations based on specific data and specific businesses through an intelligent algorithm component library; Through the containerization of functional modules and the automation management module, oil and gas multi-source heterogeneous data, intelligent algorithm component library, and intelligent service component library are containerized and packaged, and containers are uniformly scheduled and managed; According to the results of intelligent calculation, build specialized models for different scenarios through custom development modules of models in different scenarios; Through the business modules of the oil and gas production intelligent decision-making platform, intelligent decision-making, intelligent storage, transportation, gathering and distribution, intelligent operation and sales, as well as system management and maintenance of oil and gas production can be realized.
Citation Information
Patent Citations
Platform construction method of intelligent iron and steel factory
CN108171422A
Intelligent Operation and Maintenance Analysis System Based on Multi-source Heterogeneous Data Fusion, Machine Learning and Customer Service Robot
CN109343995A
