Reverse modeling method of three-mode digital twin AI management and control system for oil field station

Abstract

The invention relates to a modeling method, in particular to a three-mode digital twinborn AI management and control system reverse modeling method for an oil field station. The method comprises the following steps: 1, carrying out full-area scanning work on a station by using laser scanning equipment to collect spatial data in an area range; 2, generating station laser point cloud data based on the space scanning data; 3, importing the point cloud data into 3dsMax, and modeling according to a space picture presented by the point cloud data; 4, repeatedly checking the established model with an actual object, and carrying out local modification and adjustment; 5, data of different measurement and control systems and industrial control systems of different models in the operation room are unified to one data management server; and 6, the established model is put into a development engine, so that the real-time production data is accurately displayed on the corresponding production device, and the display is visual and accurate. Healthy and stable operation of the whole station can be guaranteed, and the rough estimation retrograde motion modeling method can prolong the whole service life of the station by 5-10 years.

Description

Technical Field

[0001] This invention relates to a modeling method, specifically to a reverse modeling method for a three-mode digital twin AI control system for oilfield stations. Background Technology

[0002] Currently, conducting experiments on dynamic and static equipment, above-ground and underground pipelines, instruments, and electrical equipment in oilfield production is extremely complex. Dynamic equipment refers to rotating devices such as motors, pumps, compressors, expanders, and generators. Static equipment refers to stationary equipment such as towers, tanks, furnaces, and various containers. Electrical equipment refers to the station's electrical equipment, such as distribution cabinets, transformers, contactors, and meters. Instrumentation equipment includes instruments, actuators, regulating valves, and transmitters. Pipeline equipment refers to all types of pipelines within the station. The process requires stopping production, changing the object to be tested, and observing data changes over a period. If the data is incorrect, the object needs to be changed again and observed for another period until the requirements are met. This is almost impossible for large production equipment and devices, and it also wastes a lot of human and material resources. The safe, long-term, stable, and healthy operation of a station mainly depends on the normal, safe, and healthy condition of all the aforementioned equipment; none can be missing. Each piece of equipment requires regular inspection, repair, maintenance, and upkeep to ensure its reliable and stable operation. Existing technologies cannot guarantee that the stations can operate stably for a long time, and their lifespan is relatively short. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a reverse modeling method for a three-model digital twin AI control system for oilfield stations. This method establishes a virtual model, a real-world model, and a process mechanism model (referred to as three models) for the station, and then merges these three models to map the real space to the virtual space, thereby reflecting the full (or semi-)lifecycle process of the corresponding physical equipment.

[0004] The technical solution adopted in this invention is: a reverse modeling method for a three-mode digital twin AI control system for oilfield stations, the reverse modeling method including the following steps: Step 1: Use a laser scanner to scan the entire site area and collect spatial data within the area, including indoor and outdoor areas, underground and surface pipelines. The underground pipeline data needs to be completed in conjunction with the design drawings. Step 2: Generate site laser point cloud data from the spatial scanning data in Step 1; Step 3: Import the point cloud data into 3ds Max (or other processing software) and model the space based on the point cloud data. Step 4: Verify the established model against the actual object several times, and make local modifications and adjustments after verification; Step 5: Unify the data from different measurement and control systems and different models of industrial control systems in the control room into a single data management server; Step 6: Place the established model into the development engine so that the real-time production data can be accurately displayed on the corresponding production equipment in a clear and intuitive manner.

[0005] Furthermore, a three-model digital twin AI control system for oilfield stations was built through reverse modeling.

[0006] Furthermore, the three modes of the three-mode digital twin AI control system are virtual model, real-scene model, and mechanism model.

[0007] Furthermore, the virtual model digitally recreates all dynamic and static equipment, above-ground and underground pipelines, instruments, and electrical equipment of the site.

[0008] Furthermore, the reality model is a computer-generated representation of all the actual equipment within the site.

[0009] Furthermore, the mechanism model is a model generated by utilizing the working mechanism of the equipment in the station, the physical properties of the working medium, and the detection parameters of the instruments and meters. It is a reproduction of the working mechanism of the actual production process.

[0010] Furthermore, digital twins previously consisted of independent CAD models, independent SCADA data, and offline simulation software, forming information silos. Digital twins connect these silos, creating a closed-loop system driven by data, supported by models, defined by software, and dominated by intelligence. The physical entity influences the digital model, and the simulation analysis and intelligent decision-making of the digital model, in turn, optimizes the physical entity. This ability to map the virtual and the real, and to interact bidirectionally, is precisely why digital twins are considered a core technology for digital transformation.

[0011] Furthermore, the aforementioned reverse modeling can be performed on the digital twin three-model AI control system for oil and gas processing stations, the digital twin three-model AI control system for oil transfer stations, and the digital twin three-model AI control system for storage depots.

[0012] The beneficial effects of this invention are as follows: This invention provides a reverse modeling method for a three-model digital twin AI control system for oilfield stations. The modeling process, along with the management, control, and integration of the established model with other systems, truly creates a virtual station corresponding to the actual production process of the station. Problems can be identified in advance in the virtual space, allowing for system optimization to save on actual equipment investment, thereby achieving energy conservation, emission reduction, and efficiency improvement. It ensures the healthy and stable operation of the entire station, and a rough estimate suggests that the reverse modeling method can extend the overall service life of the station by 5-10 years. Attached Figure Description

[0013] Figure 1 This is a block diagram of the reverse modeling method for the three-mode digital twin AI control system in Example 1; Figure 2 This is a block diagram of the three-mode digital twin AI control system built using the reverse modeling method in Example 2. Detailed Implementation Example

[0014] The reverse modeling method first uses a laser scanning instrument to scan the existing facility. Then, it uses the point cloud data acquired from the scan to build virtual and real-world models. These models are then fed into a development engine, where corresponding business processing modules are developed based on the specific workflow of the facility. These modules include functions such as real-time data monitoring, remote equipment control, real-time data report generation, and AI alarm analysis and management. The establishment of the mechanistic model is based on a mechanistic simulation environment, real-time production data from the facility, some offline data, parameters of all equipment, and physical property data of the medium. With this complete data, a mechanistic model of the facility is provided. This model can simulate the operation of all dynamic and static equipment, above-ground and underground pipelines, instruments, and electrical equipment within the facility. By adjusting various parameters on the mechanistic model, an ideal operating condition for the facility can be obtained, thereby achieving an optimized operating state for the entire facility. Figure 1 As shown, the specific modeling steps are as follows.

[0015] The reverse modeling method for a three-mode digital twin AI control system used in oilfield stations includes the following steps: Step 1: Use a laser scanner to scan the entire site area and collect spatial data within the area, including indoor and outdoor areas, underground pipelines, and underground pipelines. The underground pipeline data needs to be collected in conjunction with the design drawings.

[0016] Step 2: Generate site laser point cloud data from the spatial scanning data in Step 1.

[0017] Step 3: Import the point cloud data into 3ds Max and model the space based on the point cloud data.

[0018] Step 4: Verify the established model against the actual object several times, and make local modifications and adjustments after verification.

[0019] Step 5: Unify the data from different measurement and control systems and different models of industrial control systems in the control room into a single data management server.

[0020] Step 6: Place the established model into the development engine so that the real-time production data can be accurately displayed on the corresponding production equipment in a clear and intuitive manner.

[0021] The above modeling method is for existing oilfield sites that lack a three-model digital twin AI control system. Due to historical reasons, these sites lack virtual, real-world, and mechanistic models, thus requiring reverse modeling. The virtual model is a digital reconstruction of all static and dynamic equipment, above-ground and underground pipelines, instruments, and electrical equipment within the site. The real-world model is a computer-generated representation of all actual equipment within the site (its actual spatial location, shape, color, and size are faithfully reproduced). The mechanistic model is generated using the working mechanisms of the equipment, the physical properties of the working medium, and the detection parameters of the instruments. Using this three-model digital twin AI control system allows for complete simulation of the real system in a virtual environment, without affecting production. It provides a highly intuitive and comprehensive view of the simulation data, allowing for arbitrary modification of process, control, and electrical parameters to meet experimental needs. This guides the actual production process at the oilfield site, achieving energy conservation, emission reduction, and safe production. Example

[0022] like Figure 2 As shown, the three-model digital twin AI control system, modeled using reverse modeling, can achieve real-time management of the entire (semi-)lifecycle of site equipment (semi-lifecycle refers to situations where management is not implemented from the beginning). The main components of the control system include automatic control, real-time production data display, parameter voice alarms and analysis, real-time comprehensive energy consumption analysis, production data report management, video monitoring within the twin, asset management, real-scene measurement, drawing management, historical data query, historical trends, process flow diagrams, pipeline topology, parameter early warning, mechanism models, emergency plans, AI assistant, fire protection facility prompts, equipment internal structure display, control management, virtual roaming, measurement tools, intelligent question answering, intelligent inspection, target tracking, electronic fences, historical trajectories, grid division, TDCS control, energy consumption management, high-risk handling, escape routes, system management, personnel management, vehicle management, and office automation (OA). Additionally, it can be supplemented with integrated acoustic, vibration, and temperature detection, infrared thermal imaging, personnel positioning systems, flame detection, automatic pump leak detection, automatic lubricating oil detection, and other subsequently added instruments and measurement parameters. All of the above functions are displayed in the form of function bars. Clicking on each function bar will take you to that function.

[0023] Automatic control: This function in the digital twin enables all automatic control functions within the plant, including continuous control, on / off control, batch control, and interlocking control. Control parameters are clearly displayed, and the controlled objects are intuitive and straightforward. TDCS control can be considered for future adoption when conditions are ripe.

[0024] Real-time production data display: This function displays all real-time parameters within the site on the digital twin, accurately corresponding to their locations and corresponding precisely with the actual installed instruments. Subsequent updates or additions of sensors and transmitters can be easily added to the digital twin system.

[0025] Parameter voice alarm and analysis: This function fundamentally changes the previous alarm method of sound and light alarm. Once an alarm occurs on all displayed parameters on the digital twin model, a voice prompt will immediately appear. At the same time, the AI ​​big model will give all possible reasons for the alarm, which will facilitate analysis by operators and technicians.

[0026] Comprehensive energy consumption real-time analysis: This function provides statistical analysis of all energy consumption at the site, turning all energy consumption data into historical data. It can realize daily, weekly, monthly, and yearly energy consumption statistics, as well as comparison of energy consumption in the same period, providing a basis for decision-makers.

[0027] Production data report management: Utilizing all data within the digital twin to generate various production management reports. The data in these reports is accurate and reliable, avoiding potential problems caused by manual data entry and providing a reliable basis for upper-level decision-making.

[0028] Viewing video surveillance in a digital twin: View video surveillance from the corresponding actual location within the digital twin, avoiding the video wall monitoring mode. You can directly click on whichever you want to watch. It is also equipped with AI analysis software to detect the wearing of work clothes, hats, and shoes, as well as making and receiving phone calls, smoking, and other dangerous behaviors. The detected information is recorded for future reference.

[0029] Asset Management: This function enables site asset management within the digital twin, allowing for full (or partial) lifecycle management of every single asset within the twin. Real-time viewing provides a clear and straightforward overview, making visual asset management extremely convenient.

[0030] Real-world measurement: In digital twins, we have virtual models, real-world models, and mechanism models. For real-world models, we can measure length and area, so that inspection, maintenance, and other work can be carried out without going to the site.

[0031] Drawing Management: This function links the design drawings of various devices in the digital twin. Whenever needed, you can open the design drawing of the device and view the detailed information on the drawing. Whether it is replacing equipment, repairing equipment, or dealing with emergencies, you can open the drawing immediately to view it, avoiding a series of tedious operations such as searching for archives and wasting time.

[0032] Historical data query: This function stores all recorded historical data for easy querying at any time, forming historical data curves, which is very convenient for analyzing production equipment and can also reveal possible future trend changes.

[0033] Process flow diagram: In order to facilitate the handover between digital twin and the original planar process flow diagram, and to provide a concise and clear description of the production process, we have equipped the digital twin with both planar process flow diagram and 2.5D process flow diagram.

[0034] Pipeline topology: This function provides the location and direction of surface and underground pipelines, media flow direction, and pipeline parameters. Especially for underground pipelines, the location, direction, and media flow given in the design drawings can help quickly pinpoint the location of any sudden production accidents during long-term operations.

[0035] Parameter early warning: By utilizing stored historical data and employing predictive models to forecast future data trends, we can plan ahead and provide strong data support for decision-makers.

[0036] Mechanism Model: Establish a mechanism model for the entire plant, use the mechanism simulation results to optimize the operation of the equipment, and guide the plant's production to achieve emission reduction, energy saving and consumption reduction in an optimized state.

[0037] Emergency plans: The digital twin enables emergency plans to be readily available and respondable at any time. In the event of a sudden incident, the plan on how to deal with it can be found immediately, avoiding a chaotic and unprepared situation.

[0038] AI Assistant: This function uses the Deep Seek large model to provide reasonable explanations for alarm parameters and alarm causes. On the other hand, it can search the knowledge base to find answers to all your questions about the site.

[0039] Fire safety facility reminder: This function clearly shows the exact location of fire safety facilities on the digital twin, so that fire safety facilities and tools can be accessed immediately in the event of an emergency.

[0040] Equipment internal structure display: This function displays the internal structure of certain equipment on a digital twin, which is beneficial for operators to observe and for training new employees.

[0041] Control Management: All modifications and adjustments to control parameters must be documented and traceable for future reference. Improper modifications to control parameters can cause instability in the control system, leading to interlocking shutdowns of the equipment.

[0042] Virtual roaming: Virtual roaming is a program developed based on the actual situation of the site, which allows a virtual person to walk along a pre-designed path and perspective to showcase the actual scene within the site, and can also add other functions.

[0043] Intelligent Q&A: Intelligent Q&A uses a large AI model to provide a reasonable explanation for all questions within the site. This function helps in analyzing problems and also assists in training new employees.

[0044] Intelligent inspection: Intelligent inspection is a program developed according to specific requirements to realize intelligent inspection of key parts and key parameters. It can be parameter inspection or video inspection. When a problem is found, it will promptly alarm and notify the management personnel.

[0045] Target tracking: Using PTZ cameras and bullet cameras for video surveillance installed in the site, as well as personnel positioning systems, to achieve target tracking of people and vehicles.

[0046] Electronic fence: Electronic fence is a system for managing people and vehicles in different shaped areas under a personnel positioning system. Once certain areas are designated as areas where certain people and vehicles are not allowed to enter, an alarm will be triggered immediately if they enter, and the information will be recorded and stored.

[0047] Historical trajectory: The historical trajectory is realized based on the personnel positioning system, which allows people and vehicles to query their walking trajectories within the station, facilitating subsequent analysis and management.

[0048] Grid partitioning: Grid partitioning is the process of dividing different functional areas on a 3D digital twin model into grids for easier management and observation.

[0049] TDCS control: By using existing instruments and meters and adopting the TDCS control system, true centralized and decentralized control can be achieved, which saves on cable laying, PLC controllers and control cabinets, and DCS systems.

[0050] Energy consumption management: By utilizing existing energy metering instruments for oil, gas, water, and electricity, the energy consumption of the station is calculated and statistically analyzed in real time, enabling daily, weekly, monthly, and yearly management of energy consumption as well as comparisons with the same period, providing a scientific basis for energy conservation and consumption reduction.

[0051] High-risk handling: By utilizing AI big data models and the site's own years of experience in handling high-risk incidents, operators can immediately obtain detailed handling methods when emergencies or high-risk events occur, avoiding confusion and delays in the best handling time.

[0052] Escape routes: In the event of a high-risk incident or sudden danger, the pre-designed escape routes will be immediately visible, facilitating the safe evacuation of employees.

[0053] System Management: This function manages the overall three-model AI digital twin system, including system version, update time, etc.

[0054] Personnel Management: This function is for personnel management in the personnel positioning system, which involves classifying and managing personnel such as employees, visitors, and external construction workers.

[0055] Vehicle management: The system automatically manages vehicles entering and leaving the station, recording detailed information such as license plate number, cargo loaded, driver, and entry and exit times.

[0056] Office Automation (OA): Utilizes all data accessed through the three-model digital twin system to generate various required reports.

[0057] Integrated sound, vibration and temperature detection: This function uses an integrated sound, vibration and temperature sensor to monitor the noise, vibration and temperature of vibrating equipment such as motors, pumps and compressors in real time, and displays the measured parameters and curves on the digital twin to realize monitoring, early warning and alarm.

[0058] Infrared thermal imaging: This function uses infrared thermal imaging technology to achieve non-contact temperature measurement. For regional temperature measurement or in places where it is not suitable to install contact sensors, an infrared thermal imager can be installed to measure the highest temperature, lowest temperature, and average temperature of a region (the size range can be customized) or a fixed point.

[0059] Personnel Positioning System: This function uses personnel positioning base stations and positioning tags to achieve accurate personnel positioning, area delineation, route planning, and other functions within the site.

[0060] Flame detection: This function uses video recognition to identify flames and smoke in videos, thereby enabling real-time monitoring of potential fires.

[0061] Automatic pump leak detection: This function automatically detects leaks in various pumps online. Once a leak is detected, an alarm is immediately triggered, preventing further safety incidents. The leak alarm point is immediately displayed on the digital twin system, providing a clear and reliable view, and simultaneously notifies the operator via voice.

[0062] Automatic lubricating oil detection: This function monitors the quality of lubricating oil in high-speed pumps and large rotating equipment online. If the lubricating oil quality does not meet the requirements, an immediate voice alarm will be triggered, and the reason for the alarm will be given at the same time. The specific alarm location and alarm device will be displayed on the digital twin.

[0063] All other parameters that need to be detected: All other parameters that need to be detected can be added to the digital twin as needed, and alarm information for these parameters can be set.

[0064] The digital twin aims to achieve full (partial) or partial lifecycle management of all dynamic, static, electrical, instrumentation, and pipeline equipment within the entire facility. The digital twin system provides clear management methods and timelines for the maintenance and upkeep of each piece of equipment, along with traceable image storage management from start to finish. For example, each piece of equipment has a corresponding QR code. When maintenance personnel need to perform maintenance, they first scan the code, report the operation, and take a photo. The management software on the digital twin then lowers the health level of that equipment, adjusting the overall health level of the facility accordingly based on the equipment's importance. Maintenance and upkeep photos are taken for future reference. After maintenance and upkeep are completed, the code is scanned again, photos are taken for later review, and the original health level is restored. Similarly, alarm issues and equipment replacements can be handled according to this procedure. This ensures the overall healthy and stable operation of the facility, and a rough estimate suggests that the reverse modeling method can extend the overall lifespan of the facility by 5-10 years.

[0065] The above are merely embodiments of this document and do not limit its scope. The three models referred to herein are a virtual model, a real-world model, and a mechanistic model. These three models can be simplified to any single-model or dual-model, and various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this application should be included within the scope of these claims.

Claims

1. A reverse modeling method for a three-mode digital twin AI control system for oilfield stations, characterized in that: The reverse modeling method includes the following steps: Step 1: Use a laser scanner to scan the entire site and collect spatial data within the area. Step 2: Generate site laser point cloud data from the spatial scanning data in Step 1; Step 3: Import the point cloud data into 3ds Max and model the space based on the spatial image presented by the point cloud data; Step 4: Verify the established model against the actual object several times, and make local modifications and adjustments after verification; Step 5: Unify the data from different measurement and control systems and different models of industrial control systems in the control room into a single data management server; Step 6: Put the established model into the development engine so that the real-time production data can be accurately displayed on the corresponding production equipment.

2. The reverse modeling method for a three-mode digital twin AI control system for oilfield stations according to claim 1, characterized in that: A three-model digital twin AI control system for oilfield stations was built through reverse modeling.

3. The reverse modeling method for a three-mode digital twin AI control system for oilfield stations according to claim 2, characterized in that: The three modes of the three-mode digital twin AI control system are virtual model, real-scene model and mechanism model.

4. The reverse modeling method for a three-mode digital twin AI control system for oilfield stations according to claim 3, characterized in that: The virtual model is a digital reconstruction of all dynamic and static equipment, above-ground and underground pipelines, instruments, and electrical equipment of the site.

5. The reverse modeling method for a three-mode digital twin AI control system for oilfield stations according to claim 3, characterized in that: The real-world model is a computer-generated representation of all the actual equipment within the site.

6. The reverse modeling method for a three-mode digital twin AI control system for oilfield stations according to claim 3, characterized in that: The aforementioned mechanism model is a model generated using the working mechanism of the equipment in the plant, the physical properties of the working medium, and the detection parameters of the instruments and meters. It is a reproduction of the working mechanism of the actual production process.

Images