Multi-source monitoring data fusion processing system for digital twin water conservancy

Abstract

The invention relates to the technical field of analogue simulation, and discloses a multi-source monitoring data fusion processing system for digital twin water conservancy. Comprising an area identification module used for establishing a three-dimensional model and identifying a key area from the three-dimensional model; the model conversion module is used for constructing a physical model; the digital twinning module is used for constructing a digital twinning model of the dam; the analogue simulation module is used for simulating a real-time compression value and a future compression value; according to the invention, multi-dimensional local simulation can be carried out from the aspects of structural form, seepage material and dynamic load, so that the real physical interaction condition in the dam can be comprehensively reflected, and the problems of limitation and low efficiency caused by single-dimensional simulation of an integral structure are avoided; it is ensured that the physical model can have dynamic and accurate time-varying characteristics and high dynamic response capacity, and then the dynamic change of the dam under the complex load can be truly simulated.

Description

Technical Field

[0001] This invention relates to the field of simulation technology, and more specifically, to a multi-source monitoring data fusion and processing system for digital twin water conservancy. Background Technology

[0002] Digital twin technology, as a core means to realize dynamic interaction between physical entities and virtual models, has been gradually applied to the safety monitoring of water conservancy projects. By constructing a digital twin model corresponding to the physical dam, it can realize the simulation and status assessment of multiple physical processes such as the structural behavior, seepage state, and dynamic response of the dam, providing important support for safety early warning and operation and maintenance decisions of water conservancy projects.

[0003] Reference patent application CN120671230A discloses a digital twin-enabled adaptive early warning method and system for dams, including constructing a digital twin model of the dam; deploying multiple sensors to collect seepage, deformation, stress, and strain data of the dam, and collecting meteorological and hydrological external environmental data for multi-source data fusion; based on the digital twin model of the dam and the fused multi-source data, using numerical simulation algorithms to simulate the dam's operating status in real time, comparing and analyzing the simulation results with actual monitoring data; establishing a dynamic early warning threshold model to automatically adjust the early warning threshold based on historical dam operating data, real-time monitoring data, and simulation analysis results; and issuing an early warning message when the monitoring data or simulation analysis results exceed the early warning threshold, and feeding the early warning message back to the digital twin model for updating and optimization.

[0004] Existing multi-source monitoring data fusion processing systems typically use a single-dimensional physical field for modeling and simulation when constructing virtual models of dams. This fails to efficiently couple the multi-dimensional physical fields of the dam itself, such as structure, seepage, and dynamics. Consequently, the constructed virtual models have limitations when simulating complex real-world conditions, and are prone to excessive deviations in simulation results. As a result, they cannot realistically simulate the dynamic changes of the dam under complex loads, thus reducing the accuracy of dam simulation results.

[0005] In view of this, the present invention proposes a multi-source monitoring data fusion processing system for digital twin water conservancy to solve the above problems. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a multi-source monitoring data fusion processing system for digital twin water conservancy, applied to a water conservancy monitoring platform, comprising:

[0007] The area identification module is used to collect the basic parameters of the dam, including geometric and non-geometric parameters. It uses BIM technology to build a three-dimensional model corresponding to the basic parameters and identifies key areas from the three-dimensional model.

[0008] The model conversion module is used to construct single-field models of key regions. The single-field models include structural field models, seepage field models, and dynamic field models. The module determines the relationship between two single-field models and couples the single-field models of key regions based on the relationship, thereby converting the three-dimensional model into a physical model.

[0009] The digital twin module is used to determine the monitoring time intervals, collect multi-source monitoring data of the dam at the monitoring time, including environmental meteorological data, hydrological and water flow data, structural response data and geological structure data, and fuse the multi-source monitoring data with the physical model to construct a digital twin model of the dam.

[0010] The simulation module is used to simulate the real-time stress value of the digital twin model using multi-source monitoring data at a single monitoring moment, and to simulate the future stress value of the digital twin model using multi-source monitoring data at multiple monitoring moments, and to formulate continuous safety information, current safety information, or dangerous runaway information.

[0011] Furthermore, when building the 3D model, the geometric and non-geometric parameters of the dam are retrieved from the design drawings and database. A basic model with a contour architecture and an attribute architecture is constructed using BIM technology. Strip-shaped contour units are set on the contour architecture, and ring-shaped attribute units are set on the attribute architecture. The geometric and non-geometric parameters are imported into the contour units and attribute units respectively, thus converting the basic model into a 3D model.

[0012] Furthermore, the method for identifying key regions is as follows:

[0013] Import the 3D model into the finite element simulation software, and mesh the 3D model using a preset unit length as the side length of the mesh to generate a mesh model;

[0014] In the mesh model, mark the meshes covered by the dam body, dam foundation, and gallery respectively, and denot them as dam body mesh, dam foundation mesh, and gallery mesh. The remaining meshes are denoted as meshes to be tested.

[0015] Draw lines along the boundaries of the dam body and dam foundation, the dam foundation and gallery, and the dam body and gallery respectively to generate the first boundary line, the second boundary line and the third boundary line. Measure the distance values ​​from the grid to be tested to the first boundary line, the second boundary line and the third boundary line one by one, and record the grids to be tested with distance values ​​less than the calibration distance threshold as key grids.

[0016] The key grids that are continuously distributed are aggregated into a grid set, and a closed curve is drawn along the outer edge of the key grids in the same grid set. The region located inside the closed curve is recorded as the key region, resulting in C key regions.

[0017] Furthermore, the method for constructing the structural field model is as follows:

[0018] Model and parameter determination: Concrete dams are analyzed using a nonlinear elastic model, while rock-based dams are analyzed using an elastoplastic model. Time-varying characteristics are set using a creep model or a shrinkage model to generate structural sub-models.

[0019] Boundary conditions and load determination: The bottom of the dam foundation is fixed, and normal constraints are applied to both the left and right banks of the dam for fixation. The self-weight, hydrostatic pressure, uplift pressure, sediment pressure and temperature load of the dam are set.

[0020] Solution settings: Combine the Newton-Raphson iteration method with the solver and set the convergence criteria for the solver;

[0021] Time history analysis: Discretizes time into multiple increment steps, updates material properties within each increment step, and causes the structural sub-model to be transformed into a structural field model.

[0022] Furthermore, the method for constructing the seepage field model is as follows:

[0023] Determination of seepage zone and boundary conditions: The area where the dam body, dam foundation and downstream area of ​​the dam are located is denoted as the seepage zone, and known head boundary, known flow boundary, impermeable boundary and free seepage boundary are configured on the seepage zone to generate a seepage sub-model;

[0024] Determination of permeability coefficient: The permeability coefficient of the dam body is found, and the permeability coefficient of the dam foundation is simulated through water pressure test. The saturated permeability coefficient, porosity, unsaturated parameters and shape parameters of the seepage sub-model are set.

[0025] Determination of initial conditions: In a steady seepage field, the initial seepage field is obtained by solving the steady-state seepage equation with the design water level as the boundary.

[0026] Discretization and Solution: The seepage control equations are discretized using the finite element method or the finite difference method, and the seepage sub-model is solved using an iterative method;

[0027] Calculate the permeability: Calculate the permeability using the permeability calculation formula, and then integrate the permeability with the seepage sub-model to transform the seepage sub-model into a seepage field model.

[0028] Furthermore, the method for constructing the dynamic field model is as follows:

[0029] Modal analysis: The natural frequencies and mode shapes of the dam are detected by vibration sensors, and a dynamic sub-model adapted to the natural frequencies and mode shapes is configured.

[0030] Determination of load and damping: Seismic loads were determined using seismic acceleration time histories, and Rayleigh damping was used as the damping model;

[0031] Dynamic time history analysis: The direct integration method is used to analyze dynamic loads, explicit integration is used to analyze wave propagation, and implicit integration is used to analyze structural vibration.

[0032] Model Construction and Fusion: A secondary model containing the dam foundation and part of the ground foundation is established. The infinite ground foundation of the secondary model is simulated by using non-reflective boundaries or viscoelastic boundaries. The secondary model is then fused with the dynamic sub-model to construct the dynamic field model.

[0033] Furthermore, relationships include master-slave relationships, causal relationships, and parallel relationships;

[0034] The relationship between the structural field model and the seepage field model is defined as causal; the relationship between the structural field model and the dynamic field model is defined as parallel; and the relationship between the dynamic field model and the seepage field model is defined as master-slave.

[0035] Furthermore, the method for transforming the physical model is as follows:

[0036] A1: Construct a basic scene with a closed contour line in three-dimensional space, and set up three ring-shaped coupling positions distributed at equal angles within the basic scene;

[0037] A2: Import the structural field model, seepage field model, and dynamic field model in the key area into the three coupling positions of the coupled scenario, and establish a coupling link between two adjacent coupling positions;

[0038] A3: Add causal relationships to the coupling link between the structural field model and the seepage field model, add parallel relationships to the coupling link between the structural field model and the dynamic field model, and add master-slave relationships to the coupling link between the dynamic field model and the seepage field model, generating causal links, parallel links, and master-slave links respectively.

[0039] A4: Repeat steps A2-A3 until causal links, parallel links, and master-slave links are generated in all C key regions, thus converting the 3D model into a physical model.

[0040] Furthermore, the method for determining the monitoring time is as follows:

[0041] Starting from the last time the database was updated, and ending at the current time, retrieve the D dam logs between the starting point and the ending point.

[0042] Using the preset range of dam pressure variation as the standard range, calculate the maximum and minimum duration of the dam pressure variation phenomenon within a standard range in D dam logs one by one, and obtain D peak durations and D valley durations.

[0043] The interval duration is calculated by subtracting the D peak durations from the corresponding D valley durations. The maximum and minimum duration differences are then removed. The remaining D-2 duration differences are summed and averaged to calculate the interval duration.

[0044] Using the current time as the first monitoring time and an interval as the standard, the monitoring times are determined by counting backwards along the timeline and identifying E interval distributions.

[0045] Furthermore, the method for constructing a digital twin model is as follows:

[0046] Time-align environmental meteorological data, hydrological and flow data, structural response data, and geological structure data at the same monitoring time.

[0047] The material elastic modulus, permeability coefficient, and thermal expansion coefficient of the physical model are summarized into a parameter set. Trigger states are configured on the parameter set and initialized to adjustable states.

[0048] By using data assimilation technology to combine multi-source monitoring data with a physical model, the trigger state is switched to a closed state, and the timestamps of the multi-source monitoring data are marked, thus enabling the physical model to be converted into a digital twin model.

[0049] The technical advantages of this invention, a multi-source monitoring data fusion processing system for digital twin water conservancy, are as follows:

[0050] (1): This invention constructs a single-field model of structural field model, seepage field model and dynamic field model in key areas, and constructs a physical model based on the correlation relationship. It can perform multi-dimensional local simulation from the aspects of structural morphology, seepage material and dynamic load, so as to realistically and comprehensively reflect the real physical interaction inside the dam. It avoids the limitations and inefficiencies caused by the simulation of the whole structure and a single dimension, and ensures that the physical model can have dynamic and accurate time-varying characteristics and high dynamic response capability. In this way, it can realistically simulate the dynamic changes of the dam under complex loads and improve the accuracy of dam simulation.

[0051] (2): This invention combines multi-source monitoring data with physical models to construct a digital twin model. With the constraint of the model simulation mechanism, it can simulate and predict the specific pressure situation of the dam at the current and future times. It can play a dual monitoring and early warning role for the pressure safety status of the dam at different time lines. Thus, the monitoring method of physical entities can be transformed into the monitoring method of virtual simulation. In this way, while ensuring the accuracy of the monitoring results, the convenience of dam monitoring operation is also improved. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of a multi-source monitoring data fusion processing system for digital twin water conservancy provided in Embodiment 1 of the present invention. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] Example 1: Please refer to Figure 1 As shown in this embodiment, a multi-source monitoring data fusion processing system for digital twin water conservancy is applied to a water conservancy monitoring platform and includes:

[0055] The area identification module collects the basic parameters of the dam, including geometric and non-geometric parameters. It then uses BIM technology to build a 3D model of the dam, divides the 3D model into meshes, and identifies key areas from the 3D model.

[0056] BIM (Building Information Modeling) technology is a data-driven tool used for engineering design, construction, and management. By integrating the data and information models of buildings, it is possible to quickly transform from physical structures to virtual structures.

[0057] The 3D model is based on BIM technology, which converts the physical structure of a dam into a 3D virtual structure model and provides a standard for subsequent monitoring and analysis of the dam. In order to build a 3D model of a dam, it is necessary to collect the basic parameters of the dam and use the basic parameters as the basis for building the 3D model.

[0058] The basic parameters include geometric parameters and non-geometric parameters;

[0059] Geometric parameters are used to represent the specific structural form, size, location, and other geometric features of a dam, and serve as the basis for establishing the outline framework of a three-dimensional model.

[0060] Geometric parameters include point data, line data, surface data, and volume data;

[0061] For example, point data includes dam axis control points, gallery turning points, gate hinge points, etc.; line data includes dam crest line, dam baseline, overflow surface curve, etc.; surface data includes upstream and downstream surfaces of the dam body, foundation contact surface, gallery inner wall, etc.; volume data includes the overall volume of the dam body, gallery space, gate chamber, etc.

[0062] Non-geometric data is used to represent the non-geometric features of a dam, such as its design, construction, and attributes, and serves as the basis for building the attribute performance of a 3D model.

[0063] Non-geometric parameters include design data, construction data, and geological data;

[0064] For example, design data includes design drawings, design specifications, design parameters, etc.; construction data includes construction records, as-built drawings, material information, construction process records, etc.; and site data includes geological survey reports, foundation treatment records, etc.

[0065] Specifically, when building a 3D model of a dam, the geometric and non-geometric parameters of the dam are retrieved from the design drawings and database. A basic model with an outline architecture and an attribute architecture is constructed using BIM technology. A strip-shaped outline units are set on the outline architecture, and B ring-shaped attribute units are set on the attribute architecture. The geometric and non-geometric parameters are imported into the A outline units and the B attribute units respectively, thus converting the basic model into a 3D model.

[0066] It should be noted that the outline architecture and attribute architecture are used to represent the data import location of the external shape and internal attributes of the 3D model, thus ensuring the accuracy of the 3D model construction.

[0067] In this embodiment, the constructed three-dimensional model can only maintain consistency with the dam in terms of structural form and performance attributes, and can provide a virtual representation of the dam from an overall perspective.

[0068] In order to facilitate the accurate and intuitive display of local details in the 3D model, it is necessary to divide the 3D model of the overall structure into regions, so as to identify key areas in the 3D model that contain key information and can provide analytical basis for the safety and stability of the dam.

[0069] Specifically, the method for identifying key areas is as follows:

[0070] Import the 3D model into the finite element simulation software, and mesh the 3D model using a preset unit length as the side length of the mesh to generate a mesh model. The preset unit length is the pre-set side length of the mesh, which ensures that the size and shape of each mesh are consistent when the 3D model is divided.

[0071] In the mesh model, the meshes covered by the dam body, dam foundation, and gallery are marked and denoted as dam body mesh, dam foundation mesh, and gallery mesh, respectively. The areas corresponding to the dam body mesh, dam foundation mesh, and gallery mesh are denoted as regular areas, and the remaining meshes are denoted as meshes to be tested.

[0072] In the mesh model, lines are drawn along the boundary positions of the dam body and dam foundation, the dam foundation and gallery, and the dam body and gallery, respectively, to generate the first boundary line, the second boundary line, and the third boundary line;

[0073] The distances from each grid to be tested to the first boundary line, the second boundary line, and the third boundary line are measured one by one. Grids to be tested with distances less than the calibration distance threshold are marked as critical grids. The calibration distance threshold is the maximum distance value when a grid is marked as a critical grid, to ensure the accuracy of the critical grid division results.

[0074] Edge detection technology is used to identify the distribution state of critical grids. Critical grids with a continuous distribution are grouped into a grid set, and closed curves are drawn along the outer edges of the critical grids within the same set. Regions within these closed curves are designated as critical regions, resulting in C critical regions. The distribution state indicates whether the critical grids are spatially continuous; specifically, it includes continuous and intermittent distributions. This provides spatial constraints for determining whether critical grids belong to the same critical region.

[0075] In this embodiment, the key area is the area in the three-dimensional model that can intuitively determine whether the actual safety status of the dam is abnormal. The number of key areas is not unique, and the spatial location of the key areas is not specifically limited.

[0076] The model conversion module constructs single-field models for key regions one by one. Each single-field model includes structural field, stress field, temperature field, and seepage field. It determines the correlation between two single-field models and couples the single-field models in the same key region based on the correlation, thereby converting the three-dimensional model into a physical model.

[0077] A single-field model is a virtual model in which there are interrelated physical processes in a key area. This allows the single-field model to comprehensively represent the physical changes of different structures and forms of dams in the key area and to serve as the cornerstone for the subsequent construction of digital twin models.

[0078] Specifically, the single-field model includes the structural field model, the seepage field model, and the dynamic field model.

[0079] The structural field model is used to simulate the stress and strain changes of the dam body and foundation. When constructing the structural field model, it is necessary to comprehensively consider the nonlinear and time-varying characteristics of the materials in the dam, such as creep, shrinkage, and aging, so as to ensure that the constructed structural field model can virtually simulate the structural stress information of the dam.

[0080] Specifically, the method for constructing the structural field model is as follows:

[0081] Model and parameter determination: For concrete dams, a nonlinear elastic model (Duncan-Chang model) is used for preliminary analysis, and for rock-based dams, an elastoplastic model (Drucker-Prager model, Mohr-Coulomb model) is used for preliminary analysis. Time-varying characteristics are set using a creep model or shrinkage model to generate structural sub-models.

[0082] Determination of boundary conditions and loads: The bottom of the dam foundation is fixed, and normal constraints are applied to both the left and right banks of the dam to generate displacement boundaries. The self-weight, hydrostatic pressure, uplift pressure, sediment pressure and temperature load of the dam are set to generate loads.

[0083] Solution settings: The Newton-Raphson iteration method is combined with the solver, and the convergence criteria of the solver are set; the convergence criteria include, but are not limited to, displacement convergence criteria, force convergence criteria, and energy convergence criteria;

[0084] Time history analysis: Discretizes time into multiple increment steps, updates material properties within each increment step, and causes the structural sub-model to be transformed into a structural field model.

[0085] The seepage field model is used to simulate seepage in the dam body and foundation, and to calculate seepage pressure and seepage gradient.

[0086] Specifically, the method for constructing the seepage field model is as follows:

[0087] Determination of seepage zone and boundary conditions: The area where the dam body, dam foundation and downstream area of ​​the dam are located is denoted as the seepage zone, and boundary conditions are configured on the seepage zone. The boundary conditions include known head boundary, known flow boundary, impermeable boundary and free seepage boundary, and a seepage sub-model is generated.

[0088] Determination of permeability coefficient: The permeability coefficient of the dam body is found, and the permeability coefficient of the dam foundation is simulated through water pressure test. Based on the permeability coefficient, the material parameters of the seepage sub-model are set, including saturated permeability coefficient, porosity, unsaturated parameters, and shape parameters.

[0089] Determination of initial conditions: In a steady seepage field, the initial seepage field is obtained by solving the steady-state seepage equation with the design water level as the boundary.

[0090] Discretization and Solution: Discretize the seepage control equations using the finite element method or finite difference method, and solve the seepage sub-model using iterative methods (such as Picard iteration or Newton iteration);

[0091] Calculate the permeability: Calculate the permeability using the permeability calculation formula, and then integrate the permeability with the seepage sub-model to transform the seepage sub-model into a seepage field model.

[0092] The formula for calculating penetration force is: F = -ρ*g*∇h; where F is the penetration force, ρ is the density of water, g is the acceleration due to gravity, and h is the head of water.

[0093] The dynamic field model is a model used to simulate the response under dynamic loads such as earthquakes and water flow pulsations, which can be used to virtually simulate the changes of dams under various stresses;

[0094] Specifically, the method for constructing the dynamic field model is as follows:

[0095] Modal analysis: The natural frequencies and mode shapes of the dam are detected by vibration sensors, and a dynamic sub-model adapted to the natural frequencies and mode shapes is configured.

[0096] Determination of load and damping: Seismic loads are determined using seismic acceleration time histories (natural earthquake records or artificial synthesis), and Rayleigh damping is used as the damping model;

[0097] Dynamic time history analysis: The direct integration method is used to analyze dynamic loads, explicit integration is used to analyze wave propagation, and implicit integration is used to analyze structural vibration.

[0098] Model Construction and Fusion: A secondary model containing the dam foundation and part of the ground foundation is established. The infinite ground foundation of the secondary model is simulated by using non-reflective boundaries or viscoelastic boundaries. The secondary model is then fused with the dynamic sub-model to construct the dynamic field model.

[0099] It should be noted that the constructed structural field model, seepage field model, and dynamic field model can simulate the structural morphology, seepage material, and dynamic load of key areas in the 3D model from multiple dimensions. This allows for a split-type, multi-level simulation monitoring of the stress on the dam structure corresponding to the key areas in the 3D model, avoiding the limitations and inefficiencies caused by the overall and single-dimensional stress monitoring of the 3D model.

[0100] After constructing the single-field model, there is no direct correlation between the structural field model, seepage field model, and dynamic field model. This results in the relative isolation and discreteness of different force manifestations in the key areas of the three-dimensional model, making it impossible to dynamically simulate the dam as a whole. Therefore, it is necessary to match the single-field models in the key areas, and the matching is based on the correlation between the single-field models.

[0101] In this embodiment, the correlation relationship is used to represent the degree of correlation between two single-field models in the same key area during the simulation of a dam.

[0102] Specifically, the relationships include master-slave, causal, and parallel relationships. The master-slave relationship refers to the distinction between the importance of two single-scene models; the causal relationship refers to the correlation between the reasoning content of two single-scene models; and the parallel relationship refers to the parallel logical order of two single-scene models.

[0103] The method for determining the association relationship is as follows:

[0104] Since structural changes in a dam can lead to changes in the seepage pattern, the relationship between the structural field model and the seepage field model is determined to be a causal relationship.

[0105] Since the structural changes of the dam are not directly related to the dynamic load, the relationship between the structural field model and the dynamic field model is determined to be a parallel relationship.

[0106] Since the dynamic load of the dam affects the dynamic changes of seepage, the relationship between the dynamic field model and the seepage field model is determined to be master-slave.

[0107] After determining the relationship between the two single-field models, it is necessary to couple the single-field models with the same relationship in the same key region, and finally convert the 3D model into a physical model.

[0108] In this embodiment, the physical model is based on a three-dimensional model and is a model that can physically represent the static structural form and dynamic change trend of the dam, so that the physical model can be used as a direct object for the digital twin technology transformation of the dam.

[0109] Specifically, the method for converting the physical model is as follows:

[0110] A1: Construct a basic scene with a closed contour line in three-dimensional space, and set up three annular coupling positions distributed at equal angles within the basic scene; the coupling positions are used to provide positional constraints for the coupling operation of the structural field model, the seepage field model and the dynamic field model, so as to ensure that each single field model can remain relatively stable;

[0111] A2: Import the structural field model, seepage field model, and dynamic field model in the key area into the three coupling positions of the coupled scene, and build a coupling link between two adjacent coupling positions; the coupling link is used to provide a transmission channel for coupling interaction and data fusion between two related single field models, thereby ensuring that the single field model can maintain the overall related state;

[0112] A3: Add causal relationships to the coupling link between the structural field model and the seepage field model, add parallel relationships to the coupling link between the structural field model and the dynamic field model, and add master-slave relationships to the coupling link between the dynamic field model and the seepage field model, generating causal links, parallel links, and master-slave links respectively.

[0113] A4: Repeat steps A2-A3 until causal links, parallel links, and master-slave links are generated in all C key regions, thus converting the 3D model into a physical model.

[0114] It should be noted that causal links, parallel links, and master-slave links were generated in each key region of the physical model, enabling each key region of the physical model to perform localized and detailed simulations of the force changes in a specific area of ​​the dam.

[0115] The digital twin module determines the monitoring times at intervals, collects multi-source monitoring data of the dam at the monitoring times, and fuses the multi-source monitoring data with the physical model to construct a digital twin model of the dam.

[0116] The monitoring time is a time limit used to collect multi-source monitoring data on the dam, and it can provide a time basis for the collection of multi-source monitoring data of the dam at different times.

[0117] When determining the monitoring time, two adjacent monitoring times are not continuous on the timeline, but are distributed at intervals, which can achieve the effect of periodic, intermittent and dynamic data collection of the dam.

[0118] When determining the monitoring time, it is necessary to base it on the pressure experienced by the dam within a historical time period that is sufficient for data collection and analysis, and to determine the interval between two adjacent monitoring times by analyzing and comparing the duration of the pressure experienced by the dam.

[0119] Specifically, the method for determining the monitoring time is as follows:

[0120] Starting from the last time the database was updated, and ending at the current time, retrieve the D dam logs between the starting point and the ending point.

[0121] Using the preset range of dam pressure variation as the standard range, calculate the maximum and minimum duration of the dam pressure variation phenomenon within a standard range in D dam logs one by one, and obtain D peak durations and D valley durations.

[0122] The interval duration is calculated by subtracting the D peak durations from the corresponding D valley durations. The maximum and minimum duration differences are then removed. The remaining D-2 duration differences are summed and averaged to calculate the interval duration.

[0123] By taking the current time as the first monitoring time and using an interval as the standard, and counting back E-1 monitoring times along the timeline, the monitoring times distributed at E intervals can be determined.

[0124] Multi-source monitoring data refers to multi-dimensional data that can affect the pressure on a certain part of a dam. This allows multi-source monitoring data to be combined with the physical model of the dam to construct a digital twin model of the dam.

[0125] Specifically, multi-source monitoring data includes environmental meteorological data, hydrological and flow data, structural response data, and geological structure data;

[0126] Environmental meteorological data is used to represent the environment and weather in the area where the dam is located. Environmental meteorological data includes, but is not limited to, rainfall, wind speed, wind direction, temperature, and air pressure.

[0127] Hydrological and flow data are used to represent the flow and conditions of water within a dam. Hydrological and flow data include, but are not limited to, reservoir water level, inflow, outflow, flow velocity, water temperature, and sediment content.

[0128] Structural response data is used to represent the deformation and load changes of the structure within a dam. Structural response data includes, but is not limited to, stress-strain penetration rate, crack length, vibration frequency, static load, etc.

[0129] Geological structure data is used to represent the characteristics of geological strata within a dam. Geological structure data includes, but is not limited to, the properties of foundation soil and rock, groundwater level, and epicenter.

[0130] After collecting multi-source monitoring data, each monitoring moment has corresponding multi-source monitoring data. By combining the multi-source monitoring data at different monitoring moments with the physical model, the physical model of the dam can be dynamically combined with the real-time situation of the dam at different time points, thereby constructing a digital twin model that can perform real-time dynamic simulation of the dam.

[0131] Specifically, the method for constructing a digital twin model is as follows:

[0132] Time-align environmental meteorological data, hydrological and flow data, structural response data, and geological structure data at the same monitoring time.

[0133] All key parameters (material elastic modulus, permeability coefficient, thermal expansion coefficient) in the physical model are compiled into a parameter set. Trigger states are configured on the parameter set and initialized to adjustable states. Initializing the trigger states to adjustable states ensures that the converted digital twin model is not always in a real-time triggered simulation state, thus providing a "safety lock" structure for the simulation of the digital twin model and improving the security of the digital twin model.

[0134] By using data assimilation techniques (such as Kalman filtering, ensemble Kalman filtering, variational methods, etc.), multi-source monitoring data is combined with a physical model, the trigger state is switched to a closed state, and the timestamps of the multi-source monitoring data are marked, thereby prompting the physical model to be converted into a digital twin model.

[0135] It should be noted that the constructed digital twin model can adaptively simulate the real-time multi-dimensional situation of the dam, and can quickly and conveniently simulate the pressure on various areas of the dam through the digital twin model, achieving accurate and convenient simulation results from physical structure to virtual simulation.

[0136] The simulation module configures the model simulation mechanism, simulates the simulation information of the digital twin model, including real-time pressure value and future pressure value, and generates dynamic prompt information;

[0137] Once the digital twin model is obtained, it can be used to simulate the simulation information of the current monitoring time and the next monitoring time. The simulation information can then be used as the pressure values ​​of the dam at the current monitoring time and the next monitoring time, thereby providing a basis for analyzing the safety status of the dam at different times.

[0138] Specifically, the simulation information includes real-time pressure values ​​and future pressure values; the real-time pressure value is the pressure value simulated by the digital twin model at the current monitoring time; the future pressure value is the pressure value simulated by the digital twin model at the next monitoring time.

[0139] In this embodiment, when simulating simulation information using a digital twin model, it is necessary to first configure a model simulation mechanism and then simulate the simulation information under the constraints of the model simulation mechanism.

[0140] Specifically, the model simulation mechanism is as follows: real-time simulation is performed through a single monitoring moment, and future simulation is performed through multiple monitoring moments.

[0141] When simulating real-time pressure values, the corresponding real-time pressure values ​​are simulated using a digital twin model based on multi-source monitoring data at the current monitoring time. When simulating future pressure values, F monitoring times are continuously selected forward along the timeline from the current monitoring time, and the pressure on the dam at the next monitoring time after the current monitoring time is simulated using a digital twin model based on multi-source monitoring data at the F monitoring times. This is recorded as the future pressure value.

[0142] It should be noted that when simulating the pressure on the dam at each monitoring moment, it is necessary to use a digital twin model to simulate the real-time pressure value and the future pressure value, so as to achieve the effect of real-time simulation and advance simulation of the pressure on the dam.

[0143] Dynamic alert information is used to indicate the specific safety risks of the dam under pressure at the current and future times, and serves as the final result obtained after the fusion processing of multi-source monitoring data of the dam.

[0144] Specifically, the dynamic alerts include ongoing safety information, current safety information, and information indicating danger and loss of control.

[0145] The method for creating dynamic prompts is as follows:

[0146] The simulated real-time and future pressure values ​​were compared with the safe pressure threshold. The safe pressure threshold refers to the maximum pressure that a local location of the dam can withstand under normal and safe conditions, and can be used as a numerical basis for judging whether there are dangerous or abnormal phenomena in a local location of the dam.

[0147] When both the real-time pressure value and the future pressure value are less than or equal to the safe pressure threshold, it indicates that the dam remains safe and stable at both the current and future times, and thus continuous safety information is established.

[0148] When the real-time pressure value is less than or equal to the safe pressure threshold, and the future pressure value is greater than the safe pressure threshold, it indicates that the dam is maintaining safety and stability at the current moment, but not at the future moment. In this case, the current safety information is determined.

[0149] When the real-time pressure value exceeds the safe pressure threshold, it indicates that the dam is not maintaining safety and stability at the current moment, and a dangerous loss of control information is generated.

[0150] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-source monitoring data fusion and processing system for digital twin water conservancy, applied to a water conservancy monitoring platform, characterized in that, include: The area identification module is used to collect the basic parameters of the dam, including geometric and non-geometric parameters. It uses BIM technology to build a three-dimensional model corresponding to the basic parameters and identifies key areas from the three-dimensional model. The model conversion module is used to construct single-field models of key regions. The single-field models include structural field models, seepage field models, and dynamic field models. The module determines the relationship between two single-field models and couples the single-field models of key regions based on the relationship, thereby converting the three-dimensional model into a physical model. The digital twin module is used to determine the monitoring time intervals, collect multi-source monitoring data of the dam at the monitoring time, including environmental meteorological data, hydrological and water flow data, structural response data and geological structure data, and fuse the multi-source monitoring data with the physical model to construct a digital twin model of the dam. The simulation module is used to simulate the real-time stress value of the digital twin model using multi-source monitoring data at a single monitoring moment, and to simulate the future stress value of the digital twin model using multi-source monitoring data at multiple monitoring moments, and to formulate continuous safety information, current safety information, or dangerous runaway information.

2. The multi-source monitoring data fusion processing system for digital twin water conservancy as described in claim 1, characterized in that, When building a 3D model, the geometric and non-geometric parameters of the dam are retrieved from the design drawings and database. A basic model with a contour architecture and an attribute architecture is constructed using BIM technology. Strip-shaped contour units are set on the contour architecture, and ring-shaped attribute units are set on the attribute architecture. The geometric and non-geometric parameters are imported into the contour units and attribute units respectively, thus converting the basic model into a 3D model.

3. The multi-source monitoring data fusion processing system for digital twin water conservancy as described in claim 2, characterized in that, The method for identifying key regions is as follows: Import the 3D model into the finite element simulation software, and mesh the 3D model using a preset unit length as the side length of the mesh to generate a mesh model; In the mesh model, mark the meshes covered by the dam body, dam foundation, and gallery respectively, and denot them as dam body mesh, dam foundation mesh, and gallery mesh. The remaining meshes are denoted as meshes to be tested. Draw lines along the boundaries of the dam body and dam foundation, the dam foundation and gallery, and the dam body and gallery respectively to generate the first boundary line, the second boundary line and the third boundary line. Measure the distance values ​​from the grid to be tested to the first boundary line, the second boundary line and the third boundary line one by one, and record the grids to be tested with distance values ​​less than the calibration distance threshold as key grids. The key grids that are continuously distributed are aggregated into a grid set, and a closed curve is drawn along the outer edge of the key grids in the same grid set. The region located inside the closed curve is recorded as the key region, resulting in C key regions.

4. A multi-source monitoring data fusion processing system for digital twin water conservancy as described in claim 3, characterized in that, The method for constructing the structural field model is as follows: Model and parameter determination: Concrete dams are analyzed using a nonlinear elastic model, while rock-based dams are analyzed using an elastoplastic model. Time-varying characteristics are set using a creep model or a shrinkage model to generate structural sub-models. Boundary conditions and load determination: The bottom of the dam foundation is fixed, and normal constraints are applied to both the left and right banks of the dam for fixation. The self-weight, hydrostatic pressure, uplift pressure, sediment pressure and temperature load of the dam are set. Solution settings: Combine the Newton-Raphson iteration method with the solver and set the convergence criteria for the solver; Time history analysis: Discretizes time into multiple increment steps, updates material properties within each increment step, and causes the structural sub-model to be transformed into a structural field model.

5. A multi-source monitoring data fusion processing system for digital twin water conservancy as described in claim 4, characterized in that, The method for constructing the seepage field model is as follows: Determination of seepage zone and boundary conditions: The area where the dam body, dam foundation and downstream area of ​​the dam are located is denoted as the seepage zone, and known head boundary, known flow boundary, impermeable boundary and free seepage boundary are configured on the seepage zone to generate a seepage sub-model; Determination of permeability coefficient: The permeability coefficient of the dam body is found, and the permeability coefficient of the dam foundation is simulated through water pressure test. The saturated permeability coefficient, porosity, unsaturated parameters and shape parameters of the seepage sub-model are set. Determination of initial conditions: In a steady seepage field, the initial seepage field is obtained by solving the steady-state seepage equation with the design water level as the boundary. Discretization and Solution: The seepage control equations are discretized using the finite element method or the finite difference method, and the seepage sub-model is solved using an iterative method; Calculate the permeability: Calculate the permeability using the permeability calculation formula, and then integrate the permeability with the seepage sub-model to transform the seepage sub-model into a seepage field model.

6. A multi-source monitoring data fusion processing system for digital twin water conservancy according to claim 5, characterized in that, The method for constructing the dynamic field model is as follows: Modal analysis: The natural frequencies and mode shapes of the dam are detected by vibration sensors, and a dynamic sub-model adapted to the natural frequencies and mode shapes is configured. Determination of load and damping: Seismic loads were determined using seismic acceleration time histories, and Rayleigh damping was used as the damping model; Dynamic time history analysis: The direct integration method is used to analyze dynamic loads, explicit integration is used to analyze wave propagation, and implicit integration is used to analyze structural vibration. Model Construction and Fusion: A secondary model containing the dam foundation and part of the ground foundation is established. The infinite ground foundation of the secondary model is simulated by using non-reflective boundaries or viscoelastic boundaries. The secondary model is then fused with the dynamic sub-model to construct the dynamic field model.

7. A multi-source monitoring data fusion processing system for digital twin water conservancy as described in claim 6, characterized in that, Relationships include master-slave, causal, and parallel relationships; The relationship between the structural field model and the seepage field model is defined as causal; the relationship between the structural field model and the dynamic field model is defined as parallel; and the relationship between the dynamic field model and the seepage field model is defined as master-slave.

8. A multi-source monitoring data fusion processing system for digital twin water conservancy according to claim 7, characterized in that, The method for converting the physical model is as follows: A1: Construct a basic scene with a closed contour line in three-dimensional space, and set up three ring-shaped coupling positions distributed at equal angles within the basic scene; A2: Import the structural field model, seepage field model, and dynamic field model in the key area into the three coupling positions of the coupled scenario, and establish a coupling link between two adjacent coupling positions; A3: Add causal relationships to the coupling link between the structural field model and the seepage field model, add parallel relationships to the coupling link between the structural field model and the dynamic field model, and add master-slave relationships to the coupling link between the dynamic field model and the seepage field model, generating causal links, parallel links, and master-slave links respectively. A4: Repeat steps A2-A3 until causal links, parallel links, and master-slave links are generated in all C key regions, thus converting the 3D model into a physical model.

9. A multi-source monitoring data fusion processing system for digital twin water conservancy according to claim 8, characterized in that, The method for determining the monitoring time is as follows: Starting from the last time the database was updated, and ending at the current time, retrieve the D dam logs between the starting point and the ending point. Using the preset range of dam pressure variation as the standard range, calculate the maximum and minimum duration of the dam pressure variation phenomenon within a standard range in D dam logs one by one, and obtain D peak durations and D valley durations. The interval duration is calculated by subtracting the D peak durations from the corresponding D valley durations. The maximum and minimum duration differences are then removed. The remaining D-2 duration differences are summed and averaged to calculate the interval duration. Using the current time as the first monitoring time and an interval as the standard, the monitoring times are determined by counting backwards along the timeline and identifying E interval distributions.

10. A multi-source monitoring data fusion processing system for digital twin water conservancy according to claim 9, characterized in that, The method for constructing a digital twin model is as follows: Time-align environmental meteorological data, hydrological and flow data, structural response data, and geological structure data at the same monitoring time. The material elastic modulus, permeability coefficient, and thermal expansion coefficient of the physical model are summarized into a parameter set. Trigger states are configured on the parameter set and initialized to adjustable states. By using data assimilation technology to combine multi-source monitoring data with a physical model, the trigger state is switched to a closed state, and the timestamps of the multi-source monitoring data are marked, thus enabling the physical model to be converted into a digital twin model.

Images