CN119476097A - A digital twin online experimental method for complex geothermal systems based on deep fusion of multi-source data - Google Patents
A digital twin online experimental method for complex geothermal systems based on deep fusion of multi-source data Download PDFInfo
- Publication number
- CN119476097A CN119476097A CN202411507661.3A CN202411507661A CN119476097A CN 119476097 A CN119476097 A CN 119476097A CN 202411507661 A CN202411507661 A CN 202411507661A CN 119476097 A CN119476097 A CN 119476097A
- Authority
- CN
- China
- Prior art keywords
- geothermal
- data
- digital twin
- online
- geothermal system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Geometry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Computer Graphics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Physiology (AREA)
- Biomedical Technology (AREA)
- Genetics & Genomics (AREA)
- Algebra (AREA)
- Fluid Mechanics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
本发明提供了一种基于多源数据深度融合的复杂地热系统数字孪生在线实验方法，解决了地热系统数字孪生体的精准构建以及虚实交互低时延可视化的问题。首先，搭建典型的地热储多源异构数据并行采集系统，利用大量现场实验测试数据对地热储层参数进行精细反演、高精度辨识和不断修正；然后，建立地热系统完备的物理模型，通过高性能科学计算对实际系统进行在线模拟，并与测试数据相互校准；最后，基于实时采集的测试数据、可靠的机理模型和人工智能方法搭建地热系统数字孪生在线实验平台，实现地热开采过程的精准测量、在线监控以及地热储开采方案的迭代演进。该方法深度融合地热系统多源数据，实现地热实验过程的高精度在线模拟和低时延虚实交互，可以全方位地透明化地热储开采细节。进而，通过开展虚拟在线实验对地热储层多尺度渗流和传热过程进行智能、快速反演，大幅提高全生命周期地热开采的科学性和效能。同时，该方法所提出的多智能体协同优化调控框架具有较强的可扩展性和灵活性，通过系统调控策略的自适应优化和迭代演进，可以自动挖掘出地热井群最优布井设计方案和最优实验方案，在地热储虚拟实验的分布式优化领域具有很强的应用前景。
The present invention provides a digital twin online experimental method for a complex geothermal system based on deep fusion of multi-source data, which solves the problem of accurate construction of digital twins of geothermal systems and low-latency visualization of virtual-real interaction. First, a typical geothermal reservoir multi-source heterogeneous data parallel acquisition system is built, and a large amount of field experimental test data is used to perform fine inversion, high-precision identification and continuous correction of geothermal reservoir parameters; then, a complete physical model of the geothermal system is established, and the actual system is simulated online through high-performance scientific computing, and calibrated with the test data; finally, a geothermal system digital twin online experimental platform is built based on real-time collected test data, reliable mechanism models and artificial intelligence methods to achieve accurate measurement, online monitoring of geothermal mining process and iterative evolution of geothermal storage mining scheme. The method deeply integrates multi-source data of the geothermal system, realizes high-precision online simulation and low-latency virtual-real interaction of the geothermal experimental process, and can make the details of geothermal storage mining transparent in all directions. Furthermore, by conducting virtual online experiments, the multi-scale seepage and heat transfer processes of geothermal reservoirs are intelligently and quickly inverted, greatly improving the scientificity and efficiency of geothermal mining throughout the life cycle. At the same time, the multi-agent collaborative optimization and control framework proposed by this method has strong scalability and flexibility. Through the adaptive optimization and iterative evolution of the system control strategy, it can automatically discover the optimal well layout design scheme and the optimal experimental scheme for the geothermal well group, and has strong application prospects in the field of distributed optimization of geothermal storage virtual experiments.
Description
Technical Field
The invention belongs to the digital twin on-line test and virtual experiment technology in the geothermal field, and particularly relates to a digital twin on-line experiment method of a complex geothermal system based on multi-source data deep fusion.
Background
The strong heterogeneity and multiscale nature of geothermal reservoirs cause the flow within the subsurface deep rock to coexist in multiple modes, the spatial and temporal distribution of these flow modes is very complex, the flow mechanism is rich and diverse, and a typical multiscale physical coupling flow. In addition, geothermal systems are typically in a geological environment with many physical fields of seepage-stress-temperature-chemistry, there are many physical, mechanical and chemical processes, mainly heat transfer processes, fluid flow processes, stress and deformation processes of the medium, and chemical reactions and solute migration processes. These processes always coexist and interact to varying degrees, embodying features of multiphasics, multimediums, multiscale and multimode. The geothermal system is accelerating to advance the development of high efficiency, wisdom, low carbon and toughness, and has higher requirements on the high-efficiency perception, dynamic response, rapid analysis and optimization decision of the experimental testing process of the geothermal reservoir with complex occurrence environment, and how to accurately invert the flowing and heat transfer processes in the geothermal reservoir by using the experimental technology is a challenging work. However, the conventional geothermal system experimental test and investigation technology has a plurality of problems that on one hand, the geothermal system experimental method under the prior technical framework is simpler and coarser, and on the other hand, the conventional experimental test method often selects a typical section for idealized analysis, the authenticity is to be improved, and the experimental test effect is difficult to guarantee due to insufficient information under the hidden geological condition.
As an effective method for communicating a digital virtual world and a real physical world, digital twin technology has been widely used in different industries in recent years. The digital twin technology integrates simulation processes of multiple disciplines, multiple physical quantities, multiple scales and multiple probabilities by fully utilizing data such as physical models, sensor updating, operation histories and the like, so that the real behaviors of physical entities are accurately simulated in a virtual space. The online experimental method based on the digital twin technology can effectively support real characteristic perception, high-fidelity modeling, virtual-real mapping and multi-physical simulation of the geothermal system, brings a brand new technical paradigm for the omnibearing digital intelligent experimental test in the geothermal field, and is becoming a new development direction of intelligent geothermal field construction and digital transformation. However, the current digital twin on-line experimental method is still in an initial stage, the conceptual framework, key theory and technical direction are still unclear, the problems of 'incapacity of measuring', 'incapacity of calculating', and the six problems are permeated in each link of accurate construction and virtual-real collaborative interaction of the geothermal storage digital twin body:
(1) In the aspect of system hardware testing, the sensor arrangement is difficult to be carried out on all measuring points of the system under the influence of factors such as the scale of a geothermal system, design capacity, geological environment monitoring means of a geothermal reservoir, data transmission cost and the like, so that the experimental progress of the system is sensed truly, objectively, comprehensively and accurately. In addition, the underground sensors used for geothermal reservoir geology experiments are various and based on independent matched upper computer software, the diversity of similar sensing monitoring instruments in the geothermal reservoir experiment site causes non-uniform data formats, synchronous acquisition of data is difficult to complete, and finally, data measured by the sensors or other testing methods are often poor in quality and high in noise, the information density of the monitoring data is low, and the real states of all parts of the system cannot be reflected in multiple angles, all directions and multiple attributes.
(2) In the aspect of system software simulation, the geothermal system is characterized by complex dynamics of multiple fields, multiple scales, uncertainty and time variability. Under the existing simulation technology system, almost no global, trans-scale, multidisciplinary and dynamic accurate digital twin body can be constructed, the characteristics, behaviors and performances of a geothermal system are difficult to mirror the characteristics, behaviors and performances of the geothermal system from the physical world to the digital world with high fidelity, on one hand, geothermal experimental test data have contradictions in resolution and scale, all high-resolution data are discontinuous, and continuous data resolution is limited. On the other hand, the traditional geothermal storage geological model has the problems of poor simulation effect and the like in multi-scale seepage simulation, parameter optimization and the like, and the problem of improving the simulation precision of the geothermal digital twin system by collaborative fusion of a physical model and a data driving model is further to be explored. In addition, the digital model has huge calculation scale, needs to consume a great deal of time and calculation resources to solve, even cannot calculate, and is difficult to meet the timeliness requirement of constructing digital twin.
Disclosure of Invention
In order to solve the problems, the invention constructs the geothermal storage digital twin body by fully fusing field experimental tests, physical models and machine learning methods, and can overcome the defects of the prior art. Based on transient response of the precise digital twin body simulated geothermal reservoir in the early experimental test stage, a geothermal system well distribution scheme and an intelligent control strategy are optimized based on the transient response, iterative evolution of geothermal system regulation strategies under different exploitation working conditions is realized, and therefore an optimal experimental scheme is found.
In view of the above, the digital twin on-line experimental method of the complex geothermal system based on multi-source data depth fusion comprises the following basic operation steps;
S1, developing multisource heterogeneous test data parallelism of a geothermal system based on a Labview software platform by adopting signal conversion, real-time communication, multithread management and C/S programming technology, and unifying multiple sensor monitoring data formats by a universal acquisition interface so as to realize online reading, analysis, storage, conversion and transmission of various monitoring data;
S2, based on the inverse problem theory, machine learning and a group intelligent optimization algorithm, deeply fusing partial parameters acquired by the sensors in real time, constructing an intelligent test method for accurately inverting and identifying the global dynamic parameters of geothermal reservoirs, and solving the problem that the parameters of the geothermal reservoirs with various dynamic changes such as porosity and permeability coefficients are difficult to invert and test on line;
S3, adopting a limited sensor measuring point to comprehensively and objectively sense the geothermal system, establishing an optimal layout scheme of the complex measuring point, and further comprehensively testing parameters of 'incomplete measurement' of a digital twin model of the geothermal system;
S4, establishing a multi-source data segmentation, noise reduction and calibration method provided by a multi-sensor, realizing accurate sensing of a measured object from the same dimension at different angles and different dimensions at different angles, further performing accurate testing on parameters of a digital twin model of a geothermal system, and overcoming the problems of poor quality, high noise and low information density of test data;
S5, constructing a virtual-real mapping method of digital twin low-delay visualization and real-time collaborative interaction of a geothermal system;
S6, summarizing the hierarchical coupling characteristics among all subsystems of the geothermal system, defining the functions of all subsystem agents, further designing a reasonable multi-agent system structure, and constructing a multi-agent distributed optimization regulation experiment framework through task interaction among all agents and mutual coordination;
and S7, constructing a digital twin on-line experiment platform of the geothermal system, starting a multi-agent co-evolution work flow to develop a virtual experiment on line, optimizing a geothermal system well distribution scheme and an intelligent control strategy, and realizing iterative evolution of the system regulation strategy under different exploitation working conditions so as to find an optimal experiment scheme.
Further, in the step S1, multiple sensor monitoring data formats are unified by developing multi-source heterogeneous test data of the geothermal system in parallel and a universal acquisition interface, so as to realize online reading, analysis, storage, conversion and transmission of various monitoring data, and the method specifically comprises the following steps:
Step 1-1, aiming at the problem of non-uniform data format caused by sensor diversification, a systematic database is established according to custom protocol specification data, and the complexity of a data interaction interface between a geothermal storage digital twin body and a physical entity is reduced;
and 1-2, organically and cooperatively connecting each sensor measuring point into a whole, intelligently acquiring monitoring data information based on a multi-source sensor data acquisition system, carrying out on-line transmission and processing, integrating self attribute data, sensor data representing the state of the monitoring point, GPS data and the like, and constructing an acquisition terminal database.
Further, in the step S2, an intelligent testing method for accurately inverting and identifying the global dynamic parameters of geothermal reservoirs is constructed, so as to solve the problem that the parameters of the geothermal reservoirs, such as porosity and permeability coefficient, which are dynamically changed, are difficult to invert and test on line, and the method comprises the following steps:
step 2-1, obtaining a sample set of a test variable by adopting uniform sampling or Latin hypercube sampling, and driving a geothermal system seepage positive problem calculation model through the obtained sample set;
2-2, comparing partial parameters acquired by the sensor in real time with thermodynamic characteristics of a system for solving a positive problem model, and estimating optimal model parameters by utilizing intelligent group optimization algorithms such as machine learning and simulated annealing, genetic algorithm, particle swarm algorithm and the like based on various inverse solving criteria such as a least square criterion, a maximum likelihood criterion, a minimum mean square error criterion and the like;
And 2-3, substituting the reverse parameter into a seepage dynamics analysis model to obtain information such as underground water flow, ground temperature field and the like which are difficult to monitor in the complex geothermal reservoir.
Further, in the step S3, a geothermal system is comprehensively and objectively perceived by using limited sensor measuring points, an optimal layout scheme of complex measuring points is established, and further, parameters of "incomplete measurement" of a digital twin model of the geothermal system are comprehensively tested, and the steps are as follows:
Step 3-1, performing significance evaluation on sensor measuring points, screening out measuring points with significant influence on analysis results, and realizing optimal arrangement of sensors;
and 3-2, screening characteristic parameters capable of comprehensively, accurately and reliably representing the multiscale seepage dynamics performance of the geothermal system, improving the sensing information density and reducing the redundancy of the information.
Further, in step S4, aiming at the problems of poor data quality, high noise and low information density of monitored data measured by the geothermal system, a multi-source data segmentation, noise reduction and calibration method provided by a multi-sensor is established, and the parameters of "inaccurate measurement" of a digital twin model of the geothermal system are accurately tested, which specifically comprises the following steps:
And 4-1, reducing deviation caused by poor quality of data from the source by adopting a multi-source data segmentation and noise reduction technology. The experimental data, the fast Fourier transform, the wavelet algorithm and the machine learning depth are fused, and an artificial intelligent twin device is driven to perform reliability time-frequency analysis on the geothermal reservoir measurement data;
And step 4-2, by arranging different types of sensors at the same measuring point and arranging different types of sensors at different measuring points, the information density of the monitoring data is effectively improved, and the measured object is more accurately and finely perceived in different angles and different dimensions at the same dimension and different angles.
Further, in the step S5, a virtual-real mapping method of digital twin low-delay visualization and real-time collaborative interaction of the geothermal system is established, and the detailed steps are as follows:
Step 5-1, obtaining three-dimensional geologic body boundary range, geologic conditions (including aquifer structure, hydrogeologic parameters and the like), stratum lithology change data, deep stratum geothermal resource supply and drainage conditions (including exploitation, lateral outflow and the like) and experimental process shaft injection and production working conditions and the like required by constructing a geothermal reservoir simulation model through on-site testing, reservoir inversion and other methods, further constructing a three-dimensional geothermal storage digital twin model, and describing the distribution characteristics of aquifers and geothermal resources;
5-2, summarizing the dynamics characteristics of multiple fields, multiple scales, uncertainty and time variation of a geothermal system, constructing a trans-scale, high-fidelity, stable and reliable machine learning modeling method for fusing on-line monitoring data and physical constraints, and overcoming the problem that the digital twin experiment of the geothermal system cannot be calculated;
Step 5-3, summarizing the diversity distribution characteristics of three reservoir space types of geothermal reservoirs, including pores, cracks and fractures, and further summarizing the multi-scale flow pattern characteristics of three properties inside the geothermal reservoirs, including darcy flow, mixed flow and pipeline flow, accordingly;
Step 5-4, based on the specific characteristic volume unit scale, directly constructing a trans-scale flow online simulation method of a nested multi-scale physical process in a discrete space, and automatically and efficiently describing multi-scale flow characteristics in a geothermal storage and reservoir space according to the requirement of a thermal storage test in a digital twin online experiment;
step 5-5, combing the difference characteristics of the geothermal storage digital twin model and a real system in characteristic parameters and characteristic dimensions, establishing a method of measuring data and analog data 'measuring and calculating fusion', improving the accuracy of an analog result, carrying out high-fidelity mirroring on the characteristics, behaviors and performances of the geothermal system to a digital space, and improving the problem of digital twin 'inaccurate calculation' of the geothermal system;
And 5-6, establishing a real-time and rapid solving method based on model reduction and machine learning acceleration for multiphase, multi-medium, multi-physical and multi-scale simulation models for actual geothermal system seepage dynamics analysis, advancing a system simulation flow with huge calculation cost into second-level response efficiency, and overcoming the problem of digital twinning of a geothermal system, thereby providing real-time virtual-real mapping results for digital twinning of the geothermal system.
Further, in step S6, a multi-intelligent system structure with reasonable design of hierarchical coupling characteristics among subsystems of the geothermal system is summarized, and then a multi-intelligent body distributed optimization regulation experimental framework is constructed, which comprises the following steps:
step 6-1, respectively designing a central coordination intelligent agent, an underground heat storage/ground energy station equipment group intelligent agent and a single intelligent agent from the top layer to the bottom layer, and defining the functions and targets of the intelligent agents so as to establish a geothermal system hierarchical multi-intelligent agent structure and realize efficient distributed online control;
And 6-2, dividing the underground thermal storage multi-agent group structure into a geothermal reservoir and a well group, dividing the underground thermal storage multi-agent group structure into a pore agent, a crack agent, a fracture zone agent and the like based on the reservoir according to typical geothermal storage characteristics, dividing the ground equipment agent group into a geothermal water injection and production device and a circulating pump, and directly carrying out information communication and control action execution based on a ground/underground single agent to realize system-level optimization regulation and control in a geothermal exploitation process.
Further, in step S7, a digital twin online experimental platform of the geothermal system is built, a multi-agent co-evolution workflow is started to develop a virtual experiment online, and iterative evolution of a system regulation strategy under different exploitation working conditions is realized so as to find an optimal experimental scheme, and the detailed steps are as follows:
And 7-1, taking a digital twin model of the geothermal system as a research object, and determining an objective function, decision variables and constraint conditions of the system regulation strategy optimization problem through theoretical analysis. The objective function is determined as an experimental energy consumption index, an economic index or a multi-objective optimization function according to the optimization direction, and the decision variable is determined as a system regulation and control parameter such as the number of drilling designs, the water injection flow of a shaft and the like, and the constraint condition is the upper limit value and the lower limit value of the decision variable or the equality constraint and the inequality constraint about the decision variable;
Step 7-2, training a geothermal storage multi-scale seepage dynamics process prediction model based on model input parameters (including real-time and historical data or other characteristic variables and the like), and adopting an interval estimation method to realize real-time evaluation and prediction of the reliability degree of a prediction result, so as to provide an accurate geothermal storage seepage field as a data support for experimental scheme optimization;
step 7-3, designing a simulation module of the geothermal subsystem with high fidelity based on the on-line bidirectional virtual-real mapping method of the geothermal system established in the step 5, and starting a collaborative work flow of real-time simulation and operation parameter optimization of the digital twin model through a data interaction interface;
Step 7-4, in the iterative process, the central coordination intelligent agent establishes a collaborative optimization strategy among the intelligent agents by receiving the bottom layer information according to an experimental process optimization target, field priori knowledge and a negotiation mechanism and adopting dynamic planning and heuristic intelligent algorithm to optimize the regulation and control parameters in decision variables;
Step 7-5, transmitting the optimized regulation and control parameters into a geothermal reservoir multi-scale seepage dynamics model for simulation calculation, wherein the result of the simulation calculation is used for calculating an objective function to evaluate the effect of optimizing regulation and control;
Step 7-6, ending the optimization process of the central coordination agent, and transmitting a control signal to the underground heat storage/ground equipment group agent to control the optimization logic of the single agent at the bottom layer of the geothermal system;
And 7-7, carrying out iterative optimization on the decision variable for a plurality of times to realize the design scheme of the geothermal system and the iterative evolution of the geothermal exploitation process regulation strategy under different working conditions.
Compared with the prior art, the digital twin on-line experimental method for the complex geothermal system based on the multi-source data depth fusion has the advantages that:
(1) An optimal layout scheme of complex measuring points is constructed, and the measured object is sensed more accurately and finely from the same dimension in different angles and different dimensions in different angles, so that the sensing information density is greatly improved, the redundancy of information is reduced, and the inversion of detailed information of a geothermal system by using limited sensor measuring points is realized;
(2) The method of establishing 'measuring and calculating fusion' by deeply fusing the real-time collected test data, a reliable mechanism model and an artificial intelligence method carries out high-fidelity mirror image on the characteristics, behaviors and performances of the geothermal system to a digital space, so that the accuracy of the online simulation result of the digital twin experiment is greatly improved;
(3) Establishing a real-time and quick solving method based on model order reduction and machine learning acceleration, converting a system simulation flow with huge calculation cost into second-level response efficiency, and remarkably improving the problem of digital twin 'quick calculation' of a geothermal system;
(4) Through task interaction and mutual coordination among all intelligent agents of the geothermal system, a multi-intelligent-agent co-evolution regulation and control experiment framework is constructed, virtual experiments can be efficiently developed, a geothermal system well arrangement scheme and an intelligent control strategy are optimized on line, iterative evolution of the system regulation and control strategy under different exploitation working conditions is realized, and therefore an optimal experiment scheme is found.
Drawings
FIG. 1 is a diagram of a digital twin on-line experimental method for a complex geothermal system in accordance with an embodiment of the present invention.
Fig. 2 is a flow chart of a digital twin experiment of a sandstone thermal storage multi-scale seepage process with multi-source data fusion according to an embodiment of the method.
FIG. 3 is a machine learning modeling method for fusion of sandstone thermal storage seepage dynamics physics according to an embodiment of the present invention.
FIG. 4 is a diagram of a multiple intelligent architecture of a sandstone geothermal system, according to an embodiment of the present invention.
Fig. 5 is an online virtual-real interaction method of a sandstone thermal storage digital twin body according to an embodiment of the present invention.
FIG. 6 is an intelligent optimization method for sandstone thermal storage well group layout and experimental scheme according to an embodiment of the method of the present invention.
FIG. 7 is a process of collaborative optimization of agents based on dynamic programming and heuristic intelligent algorithms in accordance with the present invention.
Detailed Description
The following describes embodiments of the present invention in detail with reference to the accompanying drawings, and the embodiments and specific operation procedures are given in the present invention on the premise of the technical solution of the present invention, but the scope of the present invention is not limited to the following embodiments.
The present embodiment is described based on a mid-deep sandstone thermal reservoir geothermal exploitation scenario, as shown in fig. 1. The digital twin on-line experimental method of the complex geothermal system based on multi-source data depth fusion is shown in fig. 2, and comprises the following steps of;
S1, developing multisource heterogeneous test data parallelism of a sandstone geothermal system based on a Labview software platform by adopting signal conversion, real-time communication, multithread management and C/S programming technology, unifying multiple sensor monitoring data formats by a universal acquisition interface, and further realizing online reading, analysis, storage, conversion and transmission of various monitoring data, and specifically comprising the following steps:
step 1-1, aiming at the problem of non-uniform data format caused by sensor diversification, a systematic database is established according to custom protocol specification data, and the complexity of a data interaction interface between a sandstone thermal storage digital twin body and a physical entity is reduced;
and 1-2, organically and cooperatively connecting each sensor measuring point into a whole, intelligently acquiring monitoring data information based on a multi-source sensor data acquisition system, carrying out on-line transmission and processing, integrating self attribute data, sensor data representing the state of the monitoring point, GPS data and the like, and constructing an acquisition terminal database.
S2, constructing an intelligent testing method for accurately inverting and identifying global dynamic parameters of geothermal reservoirs based on inverse problem theory, machine learning and intelligent cluster optimization algorithm depth fusion of partial parameters acquired by sensors in real time, solving the problem that the parameters of various dynamic changes such as porosity and permeability coefficients of sandstone thermal reservoirs are difficult to invert and test on line, and comprising the following steps:
Step 2-1, obtaining a sample set of a test variable by adopting uniform sampling or Latin hypercube sampling, and driving a seepage positive problem calculation model of the sandstone geothermal system through the obtained sample set;
2-2, comparing partial parameters acquired by the sensor in real time with thermodynamic characteristics of the sandstone geothermal system solved by the positive problem model, and estimating optimal model parameters by utilizing intelligent optimization algorithms such as machine learning and simulated annealing, genetic algorithm, particle swarm algorithm and the like based on various inverse-solving criteria such as a least square criterion, a maximum likelihood criterion, a minimum mean square error criterion and the like;
And 2-3, substituting the reverse parameter into a seepage dynamics analysis model to obtain information such as underground water flow, ground temperature field and the like which are difficult to monitor in the complex sandstone thermal storage.
S3, adopting a limited sensor measuring point to comprehensively and objectively sense the sandstone geothermal system, establishing an optimal layout scheme of complex measuring points, and further comprehensively testing parameters of a digital twin model 'under test' of the sandstone geothermal system, wherein the steps are as follows:
Step 3-1, performing significance evaluation on sensor measuring points, screening out measuring points with significant influence on analysis results, and realizing optimal arrangement of sensors;
and 3-2, screening characteristic parameters capable of comprehensively, accurately and reliably representing the multiscale seepage dynamics performance of the sandstone geothermal system, improving the sensing information density and reducing the redundancy of the information.
S4, establishing a multi-source data segmentation, noise reduction and calibration method provided by a multi-sensor, realizing accurate sensing of a tested object from the same dimension at different angles and different dimensions at different angles, further accurately testing parameters of a digital twin model of a sandstone geothermal system, and overcoming the problems of poor quality, high noise and low information density of test data, wherein the method specifically comprises the following steps:
And 4-1, reducing deviation caused by poor quality of data from the source by adopting a multi-source data segmentation and noise reduction technology. The experimental data, the fast Fourier transform, the wavelet algorithm and the machine learning depth are fused, and an artificial intelligent twin device is driven to perform reliability time-frequency analysis on sandstone geothermal reservoir measurement data;
And step 4-2, by arranging different types of sensors at the same measuring point and arranging different types of sensors at different measuring points, the information density of the monitoring data is effectively improved, and the measured object is more accurately and finely perceived in different angles and different dimensions at the same dimension and different angles.
S5, constructing a virtual-real mapping method for digital twin low-delay visualization and real-time collaborative interaction of a sandstone geothermal system, wherein the method comprises the following detailed steps:
Step 5-1, obtaining three-dimensional geologic body boundary range, geologic conditions (including aquifer structure, hydrogeologic parameters and the like), stratum lithology change data, deep stratum geothermal resource supply and drainage conditions (including exploitation, lateral outflow and the like) and experimental process shaft injection and production working conditions and other data required by constructing a sandstone thermal reservoir simulation model through on-site testing, reservoir inversion and other methods, further constructing a three-dimensional sandstone thermal reservoir digital twin model, and describing aquifer and geothermal resource distribution characteristics;
Step 5-2, summarizing the multi-field, multi-scale, uncertain and time-varying dynamics characteristics of the sandstone geothermal system, as shown in fig. 3, constructing a trans-scale, high-fidelity, stable and reliable machine learning modeling method with on-line monitoring data and physical constraint fusion, and overcoming the problem that the digital twin experiment of the sandstone geothermal system cannot be calculated;
Step 5-3, summarizing the diversity distribution characteristics of three reservoir space types of the sandstone thermal reservoir, including pores, cracks and fractures, and further summarizing the multi-scale flow mode characteristics of three properties in the sandstone thermal reservoir, including darcy flow, mixed flow and pipeline flow, accordingly;
Step 5-4, based on the specific characteristic volume unit scale, directly constructing a trans-scale flow online simulation method of a nested multi-scale physical process in a discrete space, and automatically and efficiently describing multi-scale flow characteristics in the sandstone thermal storage space according to the requirements of a sandstone thermal storage test in a digital twin online experiment;
Step 5-5, carding the difference characteristics of the sandstone thermal storage digital twin model and a real system in characteristic parameters and characteristic dimensions, establishing a method of measuring data and analog data 'measuring and calculating fusion', improving the accuracy of a simulation result, carrying out high-fidelity mirror image on the characteristics, behaviors and performances of the sandstone geothermal system to a digital space, and solving the problem of digital twin 'inaccurate calculation' of the sandstone geothermal system;
And 5-6, establishing a real-time and rapid solving method based on model order reduction and machine learning acceleration for a multiphase, multi-medium, multi-physical and multi-scale simulation model for the seepage dynamics analysis of an actual sandstone geothermal system, advancing a system simulation flow with huge calculation cost into second-level response efficiency, and overcoming the problem of 'quick' of digital twin, thereby providing a real-time virtual-real mapping result for the digital twin of the sandstone geothermal system.
S6, summarizing the hierarchical coupling characteristics among all subsystems of the sandstone geothermal system, defining the functions of all subsystem agents, further designing a reasonable multi-agent system structure, and constructing a multi-agent distributed optimization regulation experimental framework through task interaction among all agents and mutual coordination, wherein the steps are as follows:
Step 6-1, as shown in fig. 4, respectively designing a central coordination intelligent agent, an underground heat storage/ground energy station equipment group intelligent agent and a single intelligent agent from the top layer to the bottom layer, and defining the functions and targets of the intelligent agents so as to establish a hierarchical multi-intelligent agent structure of the sandstone geothermal system and realize efficient distributed online control;
And 6-2, dividing the underground thermal storage multi-agent group structure into a geothermal reservoir and a well group, dividing the underground thermal storage multi-agent group structure into a pore agent, a crack agent, a fracture zone agent and the like based on the reservoir according to the typical sandstone thermal storage characteristics, dividing the ground equipment agent group into a geothermal water injection and production device and a circulating pump, and directly carrying out information communication and control action execution based on a ground/underground single agent to realize the system-level optimization regulation and control of the geothermal exploitation process.
S7, as shown in FIG. 5, a digital twin on-line experiment platform of the sandstone geothermal system is built, a multi-agent co-evolution workflow is started to develop a virtual experiment on line, a well distribution scheme and an intelligent control strategy of the sandstone geothermal system are optimized, iterative evolution of the system regulation strategy under different exploitation working conditions is realized, and therefore an optimal experiment scheme is found out, and the method comprises the following detailed steps:
And 7-1, as shown in fig. 6, taking a digital twin model of the sandstone geothermal system as a research object, and determining an objective function, decision variables and constraint conditions of the system regulation strategy optimization problem through theoretical analysis. The objective function is determined as an experimental energy consumption index, an economic index or a multi-objective optimization function according to the optimization direction, and the decision variable is determined as a system regulation and control parameter such as the number of drilling designs, the water injection flow of a shaft and the like, and the constraint condition is the upper limit value and the lower limit value of the decision variable or the equality constraint and the inequality constraint about the decision variable;
step 7-2, training a sandstone thermal storage multi-scale seepage dynamics process prediction model based on model input parameters (including real-time and historical data or other characteristic variables and the like), and adopting an interval estimation method to realize real-time evaluation and prediction of the reliability degree of a prediction result, thereby providing an accurate sandstone thermal storage seepage field for experimental scheme optimization as a data support;
Step 7-3, designing a simulation module of the geothermal subsystem with high fidelity based on the online bidirectional virtual-real mapping method of the sandstone geothermal system established in the step5, and starting a collaborative work flow of real-time simulation and operation parameter optimization of the digital twin model through a data interaction interface;
step 7-4, in the iterative process, the central coordination intelligent agent establishes a collaborative optimization strategy among the intelligent agents by receiving the bottom layer information according to an experimental process optimization target, field priori knowledge and a negotiation mechanism and adopting dynamic planning and heuristic intelligent algorithm to optimize the regulation and control parameters in decision variables, as shown in fig. 7;
step 7-5, transmitting the optimized regulation and control parameters into a sandstone thermal reservoir multi-scale seepage dynamics model for simulation calculation, wherein the result of the simulation calculation is used for calculating an objective function to evaluate the effect of optimizing regulation and control;
Step 7-6, ending the optimization process of the central coordination agent, and transmitting a control signal to the underground heat storage/ground equipment group agent to control the optimization logic of the single agent at the bottom layer of the geothermal system;
and 7-7, carrying out repeated iterative optimization on the decision variable to realize the design scheme of the sandstone geothermal system and the iterative evolution of the geothermal exploitation process regulation strategy under different working conditions.
The above embodiment is only a preferred embodiment of the present invention, but it is not intended to limit the present invention. Various changes and modifications may be made by one of ordinary skill in the pertinent art without departing from the spirit and scope of the present invention. Therefore, all the technical schemes obtained by adopting the equivalent substitution or equivalent transformation are within the protection scope of the invention.
Claims (8)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411507661.3A CN119476097A (en) | 2024-10-28 | 2024-10-28 | A digital twin online experimental method for complex geothermal systems based on deep fusion of multi-source data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411507661.3A CN119476097A (en) | 2024-10-28 | 2024-10-28 | A digital twin online experimental method for complex geothermal systems based on deep fusion of multi-source data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN119476097A true CN119476097A (en) | 2025-02-18 |
Family
ID=94586657
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411507661.3A Withdrawn CN119476097A (en) | 2024-10-28 | 2024-10-28 | A digital twin online experimental method for complex geothermal systems based on deep fusion of multi-source data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119476097A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119805961A (en) * | 2025-03-17 | 2025-04-11 | 中国航空工业集团公司西安飞机设计研究所 | Integrated development method for digital twin model of aircraft |
| CN119939686A (en) * | 2025-04-09 | 2025-05-06 | 西北农林科技大学 | A method for constructing a rock mass permeation grouting model with complex fracture networks based on COMSOL |
| CN120068665A (en) * | 2025-04-27 | 2025-05-30 | 北京地航时代科技有限公司 | Geological digital twin model construction method and system in petroleum exploration |
| CN120145704A (en) * | 2025-05-14 | 2025-06-13 | 洛阳云源智能科技有限公司 | Thermal processing control method, system, equipment and storage medium based on digital twin |
| CN120181622A (en) * | 2025-05-09 | 2025-06-20 | 台州市污染防治技术中心有限公司 | A cell evaluation method and system based on noise map |
| CN120296674A (en) * | 2025-04-14 | 2025-07-11 | 中原工学院 | Roadbed and pavement test and data acquisition system under moving load |
| CN120744270A (en) * | 2025-08-21 | 2025-10-03 | 陕西新眉清洁能源有限公司 | Working parameter linkage regulation and control method and system for middle-deep geothermal water source heat pump |
-
2024
- 2024-10-28 CN CN202411507661.3A patent/CN119476097A/en not_active Withdrawn
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119805961A (en) * | 2025-03-17 | 2025-04-11 | 中国航空工业集团公司西安飞机设计研究所 | Integrated development method for digital twin model of aircraft |
| CN119939686A (en) * | 2025-04-09 | 2025-05-06 | 西北农林科技大学 | A method for constructing a rock mass permeation grouting model with complex fracture networks based on COMSOL |
| CN120296674A (en) * | 2025-04-14 | 2025-07-11 | 中原工学院 | Roadbed and pavement test and data acquisition system under moving load |
| CN120068665A (en) * | 2025-04-27 | 2025-05-30 | 北京地航时代科技有限公司 | Geological digital twin model construction method and system in petroleum exploration |
| CN120068665B (en) * | 2025-04-27 | 2025-09-30 | 北京地航时代科技有限公司 | Method and system for constructing geological digital twin model in petroleum exploration |
| CN120181622A (en) * | 2025-05-09 | 2025-06-20 | 台州市污染防治技术中心有限公司 | A cell evaluation method and system based on noise map |
| CN120145704A (en) * | 2025-05-14 | 2025-06-13 | 洛阳云源智能科技有限公司 | Thermal processing control method, system, equipment and storage medium based on digital twin |
| CN120744270A (en) * | 2025-08-21 | 2025-10-03 | 陕西新眉清洁能源有限公司 | Working parameter linkage regulation and control method and system for middle-deep geothermal water source heat pump |
| CN120744270B (en) * | 2025-08-21 | 2025-11-07 | 陕西新眉清洁能源有限公司 | A method and system for linkage control of operating parameters of a medium-deep geothermal source heat pump |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN119476097A (en) | A digital twin online experimental method for complex geothermal systems based on deep fusion of multi-source data | |
| US11715034B2 (en) | Training of machine learning algorithms for generating a reservoir digital twin | |
| CA2640725C (en) | Methods, systems, and computer-readable media for fast updating of oil and gas field production models with physical and proxy simulators | |
| US12125141B2 (en) | Generation of a virtual three-dimensional model of a hydrocarbon reservoir | |
| Wang et al. | Fast history matching and optimization using a novel physics-based data-driven model: an application to a diatomite reservoir | |
| CN120068665B (en) | Method and system for constructing geological digital twin model in petroleum exploration | |
| Liu et al. | Reservoir automatic history matching: Methods, challenges, and future directions | |
| CN115618750B (en) | Underground oil-water seepage agent model based on coupling neural network | |
| Peng et al. | Predicting rate of penetration in ultra-deep wells based on deep learning method | |
| CN103149600A (en) | Automatic wiring method for determining hydrogeological parameters based on optimization control point | |
| CN119849327A (en) | Shale crack modeling platform based on cloud computing | |
| CN118296975A (en) | Inter-well connectivity prediction method based on knowledge interaction graph neural network | |
| Gensheng et al. | Current status and construction scheme of smart geothermal field technology | |
| CN117684948B (en) | A method for evaluating connectivity between gas reservoir well groups | |
| Du et al. | A novel prediction method for coalbed methane production capacity combined extreme gradient boosting with bayesian optimization | |
| Katashov et al. | Digital platform as a tool for efficient reservoir management | |
| CN118211503A (en) | A method for interpreting the production profile of horizontal wells based on ES-MDA and DTS | |
| Jiang et al. | The big data-based reservoir and fluid dynamic simulation in oilfields: Oil and gas flow prediction and injection-production optimization strategies | |
| GB2518172A (en) | Improvements in or relating to optimisation techniques | |
| Zhou et al. | Optimization of the heat recovery performance of enhanced geothermal system based on PSO-GA-BP neural networks and analytic hierarchy process | |
| Yao et al. | Explainable deep learning for dual goals: Predicting total phosphorus concentrations and identifying pollution sources in a tidal river network | |
| Ankui et al. | Dynamic safety information modeling of underground cavern groups in the entire construction process | |
| Li et al. | Graph neural networks and hybrid optimization for water-flooding regulation | |
| CN119989922B (en) | CCS fast simulation method based on multi-feature fusion neural network | |
| Liu et al. | Adaptive real-time prediction for oil production rate considering model parameter uncertainties |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20250218 |