Flood control project virtual simulation and risk rehearsal method based on digital twinning

Abstract

The invention provides a flood control project virtual simulation and risk rehearsal method based on digital twinning, and the method comprises the following steps: obtaining drainage basin multi-source data collected by a space-air-ground integrated monitoring network, the multi-source data comprising a satellite remote sensing image, an unmanned aerial vehicle LiDAR point cloud and ground sensor monitoring data; inputting the multi-source data into a pre-constructed digital twin model, wherein the model comprises a spatio-temporal data fusion layer, a physical process modeling layer, a multi-scale simulation deduction layer and a risk assessment decision-making layer which are connected in sequence; the spatio-temporal data fusion layer is used for performing spatio-temporal registration and feature fusion on input multi-source data, and constructing a total element digital backplane; through the space-air-ground integrated monitoring network and the spatio-temporal data fusion technology, high-precision digital mapping of basin total elements is realized. Compared with a traditional single data source, time-space consistency of flood routing simulation is remarkably improved through multi-source data fusion, and prediction errors of key physical processes such as river channel scouring are greatly reduced.

Description

Technical Field

[0001] The present invention relates to the field of digital technology for water conservancy projects, and in particular to a virtual simulation and risk pre-rehearsal method for flood control projects based on digital twins. Background Art

[0002] As global climate change intensifies and extreme flood events become more frequent, traditional flood control projects face challenges such as insufficient simulation accuracy and limited risk pre-rehearsal capabilities. Existing technologies have the following problems in practical applications:

[0003] 1. Inadequate model accuracy and physical process characterization: Traditional hydrodynamic models oversimplify the simulation of complex river topography (such as bends and floodplains) and the degradation of embankment materials. This makes it difficult to accurately reflect nonlinear physical processes such as sediment movement and piping during flood development, resulting in significant deviations between simulation results and actual operating conditions. For example, when simulating river scour, traditional models often have a prediction error of more than 20% for riverbed evolution, making them inadequate for embankment safety assessments.

[0004] Second, the conflict between efficiency and accuracy in multi-scale simulations: In basin-wide flood simulations, using a uniform grid resolution makes it difficult to balance computational efficiency with the accuracy requirements of critical areas (such as dangerous sections). Traditional fixed-grid solutions can lead to computational distortion in complex terrain due to overly coarse grids, while a finer global grid would exponentially increase the computational workload, making them unable to meet the real-time requirements of emergency drills.

[0005] 3. Lack of Uncertainty Management in Risk Assessment: Existing technologies fail to quantitatively analyze factors such as model parameter uncertainty and data errors. Risk assessment results are often deterministic predictions, lacking probabilistic expression. When faced with exceptional flooding, this fails to provide decision makers with a confidence interval for risk, resulting in insufficient adaptability and robustness in dispatch plans.

[0006] Fourth, weak intelligence and interactivity in decision-making support: Traditional flood control scheduling relies on empirical rules or single-objective optimization, making it difficult to balance multiple objectives such as flood control safety, water resource utilization, and ecological protection. Furthermore, the decision-making process lacks human-machine collaboration, making it impossible to effectively integrate expert knowledge and real-time monitoring data, limiting the scientific nature and timeliness of emergency response. Summary of the Invention

[0007] In view of this, the embodiments of the present invention hope to provide a virtual simulation and risk pre-rehearsal method for flood control projects based on digital twins to solve or alleviate the technical problems existing in the existing technology and at least provide a beneficial option.

[0008] In order to solve the above technical problems, a technical solution adopted in this application is: to provide a virtual simulation and risk pre-rehearsal method for flood control projects based on digital twins, including the following steps: obtaining multi-source data of the watershed collected by the integrated air-space-ground monitoring network, the multi-source data including satellite remote sensing images, UAV LiDAR point clouds and ground sensor monitoring data; inputting the multi-source data into a pre-built digital twin model, the model including a spatiotemporal data fusion layer, a physical process modeling layer, a multi-scale simulation deduction layer and a risk assessment decision layer connected in sequence; the spatiotemporal data fusion layer performs spatiotemporal registration and feature fusion on the input multi-source data to construct a full-factor digital backplane; the physical process modeling layer is based on the The digital baseboard constructs a hydrodynamic model, a levee material degradation model and an ecological response model through a physical information neural network coupling algorithm; the multi-scale simulation and deduction layer performs multi-scale grid division and parallel calculation on the model output by the physical process modeling layer, generates extreme flood scenarios through a generative adversarial network and performs evolutionary rehearsals; the risk assessment and decision-making layer generates a flood control scheduling plan based on the multi-scale simulation and deduction results through a multi-objective reinforcement learning algorithm, and optimizes human-computer collaborative decision-making through a cognitive twin model; according to the flood control scheduling plan output by the risk assessment and decision-making layer, a heat map representing the flood risk distribution and an emergency response plan are generated, and the heat map is annotated with risk confidence intervals based on an uncertainty quantification algorithm.

[0009] As a further preferred embodiment of this technical solution: the acquisition of multi-source watershed data collected by the integrated air-space-ground monitoring network includes: obtaining remote sensing images with a spatial resolution of ≤10m from a satellite group equipped with high-resolution imaging payloads, wherein the remote sensing images contain visible light, near-infrared and short-wave infrared band data for extracting river system distribution, vegetation coverage and land use type; obtaining a point cloud density of ≥50 points / m through an unmanned aerial vehicle platform equipped with a lidar and inertial navigation system 2 The LiDAR point cloud data of embankments and riverbeds are collected simultaneously, and optical images are collected to construct a three-dimensional texture model; ground sensor data deployed in river sections, reservoir dams and embankment projects are collected, including water level gauge data with a sampling frequency of ≥1 time / minute, water velocity profiles monitored by flow meters, pore water pressure of embankments collected by piezometers, and millimeter-level deformation data of GPS displacement monitoring stations; the multi-source data are unified in time and space through the 5G communication network and the Beidou-3 positioning system, where the time labeling accuracy of satellite images is ≤1 second, the time synchronization error between drone point clouds and ground sensor data is ≤50 milliseconds, and the spatial registration error is ≤10 cm, forming a three-dimensional monitoring data set covering the basin's topography, hydrological characteristics and engineering status.

[0010] As a further preferred embodiment of this technical solution: in the digital twin model: the spatiotemporal data fusion layer is used to input satellite remote sensing images, drone LiDAR point clouds, and ground sensor data into the spatiotemporal registration module, and through spatial coordinate conversion and timestamp synchronization, build a full-factor digital baseplate covering the basin's topography, hydrology, and engineering facilities; the physical process modeling layer is used to build a hydrodynamic model based on the digital baseplate to simulate the flood evolution process, establish a levee material degradation model to analyze the impact of long-term flood erosion on engineering structures, and realize multi-physics field coupling modeling through physical information neural networks; the multi-scale simulation and deduction layer is used to adaptively grid the flow field data output by the physical process model, using fine grids in key areas to improve calculation accuracy and coarse grids in other areas to balance calculation efficiency, generate various extreme flood scenarios through generative adversarial networks, and simulate and deduce flood evolution in the entire basin; the risk assessment and decision-making layer is used to input the simulation and deduction results into the multi-objective optimization model to generate a scheduling plan that takes into account flood control safety, water resource utilization, and ecological protection; the scheduling plan is evaluated by human-machine collaboration through the cognitive twin model, and a flood control scheduling plan is output based on historical case analysis.

[0011] As a further preferred embodiment of this technical solution: the spatiotemporal data fusion layer performs spatiotemporal registration and feature fusion on the input multi-source data to construct a full-element digital baseplate, including: converting the geographic coordinate system of the satellite remote sensing image into a unified coordinate system of the watershed, matching it with the spatial coordinates of the UAV LiDAR point cloud, and eliminating the spatial deviation of different data sources through feature point extraction and matching algorithms; adding timestamps to the real-time monitoring data collected by ground sensors, and synchronously calibrating it with the imaging time of the satellite image and the time series of the UAV scanning to ensure the consistency of the multi-source data in the time dimension; fusing the processed satellite images, LiDAR point clouds and sensor data to construct a three-dimensional digital baseplate containing elements such as the watershed topography, river systems, embankment projects, and hydrological monitoring points, thereby realizing full-element digital mapping of the physical watershed.

[0012] As a further preferred embodiment of the present technical solution: the physical process modeling layer is based on the digital baseboard, and constructs a hydrodynamic model, a levee material degradation model and an ecological response model through a physical information neural network coupling algorithm, including: hydrodynamic model construction: inputting data such as terrain elevation and river section in the digital baseboard into the improved Saint-Venant equations, learning the nonlinear mapping of the relationship between water level and flow through the physical information neural network, and constructing a hydrodynamic model that can reflect the complex hydraulic characteristics in the process of flood evolution; levee material degradation model construction: based on the levee structure parameters in the digital baseboard, inputting material composition and construction process information into the quantum field theory driven microstructured model. The system uses a visual structural model to simulate the degradation process of material properties such as concrete carbonization and steel corrosion under flood erosion, and establishes a degradation equation for material strength changing with time; the system constructs an ecological response model: extracts river ecological parameters from the digital base plate, associates hydrodynamic conditions with ecosystem indicators, and constructs an ecological response model through a physical information neural network coupling algorithm to predict the impact of flood processes on the basin ecosystem; multi-model coupling integration: through a physical information neural network coupling algorithm, the hydrodynamic model, embankment material degradation model and ecological response model are integrated to achieve a two-way coupled simulation of the flood evolution process with the embankment structure safety and ecosystem changes.

[0013] As a further preferred embodiment of this technical solution: the multi-scale simulation deduction layer performs multi-scale grid division and parallel calculation on the model output by the physical process modeling layer, generates extreme flood scenarios through generative adversarial networks and performs evolutionary previews, including: multi-scale grid division: for the hydrodynamic model and embankment material degradation model output by the physical process modeling layer, differentiated grid division is performed according to the complexity of the basin terrain and the importance of the project, fine grids are used in key areas of river bends and embankment dangerous projects to capture local hydraulic characteristics, and coarse grids are used in open river sections to improve computing efficiency, forming a grid system that takes into account both accuracy and efficiency; parallel computing framework construction: deploying the multi-scale grid model to distributed The system uses a generative adversarial network to learn the distribution characteristics of historical flood data, input climate change scenarios and basin human activity parameters, and generate extreme flood scenarios that exceed historical records. The system also uses a multi-scale grid model to drive the coupled calculation of the hydrodynamic model and the embankment material degradation model to simulate the evolution of floods in the basin, the safety changes of embankment structures, and the response of the ecosystem, thus realizing dynamic rehearsal of flood risks.

[0014] As a further preferred embodiment of this technical solution: the risk assessment decision-making layer generates flood control scheduling plans based on multi-scale simulation results through a multi-objective reinforcement learning algorithm, and performs human-machine collaborative decision optimization through a cognitive twin model, including: multi-objective optimization model construction: using the inundation range, embankment safety factor, and ecological flow index obtained from the flood evolution preview as constraints, and taking flood control safety guarantee rate, water resource utilization rate, and ecological protection benefit as optimization targets, a multi-objective function system is established; using a multi-objective reinforcement learning algorithm to optimize the decision variables for reservoir group scheduling and flood diversion area activation, and generate a Pareto optimal scheduling plan set with different risk preferences; scheduling plan risk assessment: based on the uncertainty quantification algorithm, analyze the impact of errors in extreme scenario generation, model parameters and other links on the scheduling plan, and calculate the flood risk probability distribution of each plan; by constructing a risk indicator system for flood, engineering and social composite systems system, conduct a comprehensive assessment of the secondary disasters and economic losses that may be caused by the scheduling plan, and mark the risk confidence interval of the plan; cognitive twin model construction: integrate historical flood control cases, expert decision-making logic and watershed management knowledge to build a cognitive twin model based on the knowledge graph; use natural language processing technology to convert the scheduling plan into a semantic network, match it with the historical case library for similarity, extract referenceable decision-making experience, and generate scheme optimization suggestions; human-machine collaborative decision optimization: input the scheduling plan and risk assessment results generated by multi-objective optimization into the cognitive twin model, and present them to decision makers through a visual interactive interface; use the attention mechanism to highlight the key risk points in the plan, support decision makers to adjust the scheduling parameters through natural language interaction, and the model will provide real-time feedback on the risk changes after adjustment, forming a collaborative decision-making closed loop of machine generation, human optimization and machine verification, and finally output a flood control scheduling plan that takes into account both scientificity and operability.

[0015] As a further preferred embodiment of the present technical solution: the flood control scheduling plan outputted by the risk assessment decision-making layer generates a heat map and an emergency response plan representing the flood risk distribution, and the heat map marks the risk confidence interval based on the uncertainty quantification algorithm, including: flood risk index calculation: based on the flood evolution simulation results of the scheduling plan, extract the inundation depth, flow rate, and duration parameters of each region, calculate the flood risk index in combination with the vulnerability curve of the disaster-bearing body, and construct a quantitative indicator system reflecting the regional risk level; uncertainty quantification and visualization: use the Bayesian model averaging method to analyze the propagation path of the multi-source model prediction error, calculate the probability distribution of the inundation range and the peak flow parameters; map the risk index and the uncertainty results into a visual heat map, represent the risk level by a color gradient, mark the risk confidence interval with transparency or contour lines, and present the basin flood Spatial distribution of risks and range of uncertainty; generation of emergency response plans: based on the high-risk areas identified by the heat map, combined with data such as topography, population distribution, and infrastructure layout, differentiated emergency response plans are automatically generated. The plan content includes: personnel transfer route planning, material storage point site selection and deployment plan, important facility protection measures, etc. For areas with a large risk confidence interval, multiple sets of alternative plans are generated simultaneously to deal with prediction uncertainties; plan feasibility verification and optimization: the emergency response plan is input into the digital twin model for simulation verification to evaluate the capacity of the transfer route, the timeliness of material deployment, and the effectiveness of facility protection measures; the plan is iteratively optimized through reinforcement learning algorithms until the emergency response time window requirements are met, and finally a complete emergency disposal plan is formed including risk heat maps, regional response measures and alternative plans.

[0016] In order to solve the above technical problems, another technical solution adopted in this application is: a computer device, which includes a processor and a memory coupled to the processor, wherein program instructions are stored in the memory, and when the program instructions are executed by the processor, the processor executes the steps of the digital twin-based flood control project virtual simulation and risk rehearsal method as described above.

[0017] In order to solve the above technical problems, another technical solution adopted in this application is: a storage medium storing program instructions that can realize the virtual simulation and risk rehearsal method of flood control projects based on digital twins as described above.

[0018] The embodiment of the present invention adopts the above technical solution, which has the following advantages:

[0019] 1. Multi-source data fusion improves simulation accuracy and real-time performance: Through an integrated air-space-ground monitoring network and spatiotemporal data fusion technology, high-precision digital mapping of all river basin elements is achieved. Compared with traditional single data sources, multi-source data fusion significantly improves the spatiotemporal consistency of flood evolution simulations and significantly reduces prediction errors in key physical processes such as river scour.

[0020] Second, the coupling of physical and intelligent algorithms breaks through multi-scale simulation bottlenecks: Adaptive meshing and parallel computing technologies significantly reduce simulation time for the entire river basin while ensuring computational accuracy in key areas. The introduction of physical-informed neural networks and generative adversarial networks enables the model to accurately capture complex physical processes such as piping and icing, and generate highly realistic extreme flood scenarios.

[0021] 3. Full-link uncertainty quantification enhances the credibility of risk assessment: Convert flood risk from deterministic prediction to probabilistic expression, generate risk heat maps with marked confidence intervals, and significantly reduce the prediction error of key parameters.

[0022] 4. Multi-objective optimization and cognitive twinning realize an intelligent decision-making closed loop: Through multi-objective reinforcement learning algorithms, flood control benefits and ecological protection effects can be simultaneously improved in flood control scheduling, breaking through the limitations of traditional single-objective optimization.

[0023] 5. An engineering implementation system ensures the implementation of the technology: The modular deployment architecture supports elastic expansion of the system to meet the real-time computing needs of emergency response; the full-factor digital backplane and closed-loop verification mechanism enable full life cycle management of flood control projects, effectively reducing the full-cycle cost of the project.

[0024] The above summary is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments and features described above, further aspects, embodiments and features of the present invention will be readily apparent by reference to the accompanying drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0026] Figure 1 is a flow chart of the method of the present invention;

[0027] Figure 2 Schematic diagram of the system module of an embodiment of the present invention

[0028] Figure 3A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0030] It should be clear that the following embodiments of the present disclosure are described through specific concrete examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other in the absence of conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present disclosure.

[0031] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on this disclosure, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein can be used to implement this device and / or practice this method.

[0032] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present disclosure. The illustrations only show components related to the present disclosure and are not drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component can be changed at will, and the component layout type may also be more complicated.

[0033] Additionally, in the following description, specific details are provided to provide a thorough understanding of the examples. However, one skilled in the art will appreciate that the aspects described can be practiced without these specific details.

[0034] Figure 1 This is a flow chart of a flood control engineering virtual simulation and risk pre-play method based on digital twins according to an embodiment of the present invention. It should be noted that if there are substantially the same results, the method of this application is not based on Figure 1 The process sequence shown is limited. Figure 1 As shown: A virtual simulation and risk rehearsal method for flood control projects based on digital twins, comprising the following steps: obtaining multi-source data of the watershed collected by the integrated air-space-ground monitoring network, wherein the multi-source data include satellite remote sensing images, UAV LiDAR point clouds and ground sensor monitoring data; inputting the multi-source data into a pre-built digital twin model, wherein the model includes a spatiotemporal data fusion layer, a physical process modeling layer, a multi-scale simulation deduction layer and a risk assessment decision layer connected in sequence; the spatiotemporal data fusion layer performs spatiotemporal registration and feature fusion on the input multi-source data to construct a full-factor digital backplane; the physical process modeling layer is based on the digital backplane, through physical information The neural network coupling algorithm constructs a hydrodynamic model, a levee material degradation model and an ecological response model; the multi-scale simulation and deduction layer performs multi-scale grid division and parallel calculation on the model output by the physical process modeling layer, generates extreme flood scenarios through a generative adversarial network and performs evolutionary rehearsals; the risk assessment and decision-making layer generates a flood control scheduling plan based on the multi-scale simulation and deduction results through a multi-objective reinforcement learning algorithm, and performs human-computer collaborative decision optimization through a cognitive twin model; according to the flood control scheduling plan output by the risk assessment and decision-making layer, a heat map representing the flood risk distribution and an emergency response plan are generated, and the heat map is annotated with risk confidence intervals based on an uncertainty quantification algorithm.

[0035] Specifically: The acquisition of multi-source watershed data collected by the integrated air-space-ground monitoring network includes: obtaining remote sensing images with a spatial resolution of ≤10m from a satellite group equipped with high-resolution imaging payloads. The remote sensing images contain visible light, near-infrared and short-wave infrared band data, which are used to extract river system distribution, vegetation coverage and land use types; obtaining point cloud density ≥50 points / m through an unmanned aerial vehicle platform equipped with a lidar and inertial navigation system 2 The LiDAR point cloud data of embankments and riverbeds are collected simultaneously, and optical images are collected to construct a three-dimensional texture model; ground sensor data deployed in river sections, reservoir dams and embankment projects are collected, including water level gauge data with a sampling frequency of ≥1 time / minute, water velocity profiles monitored by flow meters, pore water pressure of embankments collected by piezometers, and millimeter-level deformation data of GPS displacement monitoring stations; the multi-source data are unified in time and space through the 5G communication network and the Beidou-3 positioning system, where the time labeling accuracy of satellite images is ≤1 second, the time synchronization error between drone point clouds and ground sensor data is ≤50 milliseconds, and the spatial registration error is ≤10 cm, forming a three-dimensional monitoring data set covering the basin's topography, hydrological characteristics and engineering status.

[0036] Specifically: In the digital twin model: the spatiotemporal data fusion layer is used to input satellite remote sensing images, drone LiDAR point clouds and ground sensor data into the spatiotemporal registration module, and through spatial coordinate conversion and timestamp synchronization, build a full-factor digital base plate covering the basin topography, hydrology and engineering facilities; the physical process modeling layer is used to build a hydrodynamic model based on the digital base plate to simulate the flood evolution process, establish a levee material degradation model to analyze the impact of long-term flood erosion on engineering structures, and realize multi-physical field coupling modeling through physical information neural networks; the multi-scale simulation and deduction layer is used to adaptively grid the flow field data output by the physical process model, using fine grids in key areas to improve calculation accuracy and coarse grids in other areas to balance calculation efficiency, generate various extreme flood scenarios through generative adversarial networks, and conduct full-basin flood evolution simulation and deduction; the risk assessment and decision-making layer is used to input the simulation and deduction results into the multi-objective optimization model to generate a scheduling plan that takes into account flood control safety, water resource utilization and ecological protection; the scheduling plan is evaluated by human-machine collaboration through the cognitive twin model, and a flood control scheduling plan is output based on historical case analysis.

[0037] Specifically: the spatiotemporal data fusion layer performs spatiotemporal registration and feature fusion on the input multi-source data to construct a full-factor digital baseplate, including: converting the geographic coordinate system of the satellite remote sensing image into a unified coordinate system for the watershed, matching it with the spatial coordinates of the UAV LiDAR point cloud, and eliminating the spatial deviation of different data sources through feature point extraction and matching algorithms; adding timestamps to the real-time monitoring data collected by ground sensors, and synchronously calibrating it with the imaging time of the satellite image and the time series of the UAV scanning to ensure the consistency of the multi-source data in the time dimension; fusing the processed satellite images, LiDAR point clouds and sensor data to construct a three-dimensional digital baseplate containing elements such as the watershed topography, river systems, embankment projects, and hydrological monitoring points, thereby realizing full-factor digital mapping of the physical watershed.

[0038] Specifically: The physical process modeling layer is based on the digital baseboard and constructs a hydrodynamic model, a levee material degradation model and an ecological response model through a physical information neural network coupling algorithm, including: Hydrodynamic model construction: inputting terrain elevation, river section and other data in the digital baseboard into the improved Saint-Venant equations, learning the nonlinear mapping of the relationship between water level and flow through the physical information neural network, and constructing a hydrodynamic model that can reflect the complex hydraulic characteristics during the evolution of floods; Levee material degradation model construction: Based on the levee structure parameters in the digital baseboard, the material composition and construction process information are input into the microstructure model driven by quantum field theory , simulate the material performance degradation process such as concrete carbonization and steel corrosion under flood erosion, and establish the degradation equation of material strength changing with time; construction of ecological response model: extract river ecological parameters in the digital base plate, associate hydrodynamic conditions with ecosystem indicators, and construct an ecological response model through the physical information neural network coupling algorithm to predict the impact of flood processes on the basin ecosystem; multi-model coupling integration: through the physical information neural network coupling algorithm, integrate the hydrodynamic model, embankment material degradation model and ecological response model to realize the two-way coupling simulation of flood evolution process and embankment structure safety and ecosystem changes.

[0039] Specifically: The multi-scale simulation and deduction layer performs multi-scale grid division and parallel calculation on the model output by the physical process modeling layer, generates extreme flood scenarios through generative adversarial networks and performs evolutionary previews, including: Multi-scale grid division: For the hydrodynamic model and embankment material degradation model output by the physical process modeling layer, differentiated grid division is performed according to the complexity of the basin terrain and the importance of the project. Fine grids are used in key areas of river bends and embankment dangers to capture local hydraulic characteristics, and coarse grids are used in open river sections to improve computing efficiency, forming a grid system that takes into account both accuracy and efficiency; Parallel computing framework construction: Deploy the multi-scale grid model to a distributed computing platform , the splitting and distribution of computing tasks are realized through the message passing interface, and the large-scale flow field calculation is accelerated by the graphics processor cluster, so as to realize the parallel solution of the flood evolution process in the whole basin and shorten the simulation time; extreme scenario generation: using the generative adversarial network to learn the distribution characteristics of historical flood data, input climate change scenarios and basin human activity parameters, and generate extreme flood scenarios that exceed historical records; flood evolution rehearsal: the generated extreme scenarios are input into the multi-scale grid model, driving the hydrodynamic model and the embankment material degradation model to perform coupled calculations, simulating the evolution process of floods in the basin, the safety changes of embankment structures and the response of the ecosystem, and realizing the dynamic rehearsal of flood risks.

[0040] Specifically: The risk assessment decision-making layer generates flood control scheduling plans based on the results of multi-scale simulation and deduction through a multi-objective reinforcement learning algorithm, and performs human-machine collaborative decision optimization through a cognitive twin model, including: Multi-objective optimization model construction: The inundation range, embankment safety factor, and ecological flow index obtained from the flood evolution preview are used as constraints, and the flood control safety guarantee rate, water resource utilization rate, and ecological protection benefits are used as optimization goals to establish a multi-objective function system; The multi-objective reinforcement learning algorithm is used to optimize the decision variables for reservoir group scheduling and flood diversion area activation, and a Pareto optimal scheduling plan set with different risk preferences is generated; Scheduling plan risk assessment: Based on the uncertainty quantification algorithm, the impact of errors in extreme scenario generation, model parameters, etc. on the scheduling plan is analyzed, and the flood risk probability distribution of each plan is calculated; By constructing a flood, engineering and social composite system risk indicator system, the scheduling plan is optimized. A comprehensive assessment is conducted on the secondary disasters and economic losses that may be caused by the plan, and the risk confidence interval of the plan is marked; cognitive twin model construction: historical flood control cases, expert decision-making logic and watershed management knowledge are integrated to build a cognitive twin model based on the knowledge graph; the scheduling plan is converted into a semantic network through natural language processing technology, and similarity matching is performed with the historical case library to extract referenceable decision-making experience and generate optimization suggestions for the plan; human-machine collaborative decision optimization: the scheduling plan and risk assessment results generated by multi-objective optimization are input into the cognitive twin model and presented to the decision maker through a visual interactive interface; the attention mechanism is used to highlight the key risk points in the plan, and the decision maker is supported to adjust the scheduling parameters through natural language interaction. The model provides real-time feedback on the risk changes after adjustment, forming a collaborative decision-making closed loop of machine generation, human optimization and machine verification, and finally outputting a flood control scheduling plan that takes into account both scientificity and operability.

[0041] Specifically: According to the flood control scheduling plan output by the risk assessment decision-making layer, a heat map representing the flood risk distribution and an emergency response plan are generated. The heat map is annotated with risk confidence intervals based on an uncertainty quantification algorithm, including: Flood risk index calculation: Based on the flood evolution simulation results of the scheduling plan, the inundation depth, flow rate, and duration parameters of each region are extracted, and the flood risk index is calculated in combination with the vulnerability curve of the disaster-bearing body to construct a quantitative indicator system reflecting the regional risk level; Uncertainty quantification and visualization: The Bayesian model averaging method is used to analyze the propagation path of the multi-source model prediction error, and the probability distribution of the inundation range and peak flow parameters is calculated; the risk index and uncertainty results are mapped into a visual heat map, the risk level is represented by a color gradient, and the risk confidence interval is annotated with transparency or contour lines to present the spatial distribution of the basin flood risk. Distribution and uncertainty range; Emergency response plan generation: Based on the high-risk areas marked by the heat map, combined with data such as topography, population distribution, and infrastructure layout, differentiated emergency response plans are automatically generated. The plan content includes: personnel transfer route planning, material storage point site selection and deployment plan, important facility protection measures, etc. Among them, for areas with a large risk confidence interval, multiple sets of alternative plans are generated simultaneously to deal with prediction uncertainties; Plan feasibility verification and optimization: The emergency response plan is input into the digital twin model for simulation verification to evaluate the capacity of the transfer route, the timeliness of material deployment, and the effectiveness of facility protection measures; the plan is iteratively optimized through reinforcement learning algorithm until the emergency response time window requirements are met, and finally a complete emergency disposal plan is formed including risk heat map, regional response measures and alternative plans.

[0042] Figure 2 This is a functional module diagram of the digital twin-based flood control engineering virtual simulation and risk pre-rehearsal system according to an embodiment of the present application. Figure 2 As shown in the figure, the digital twin-based flood control project virtual simulation and risk rehearsal system includes:

[0043] 1. Integrated air-space-ground monitoring subsystem: used to obtain multi-source real-time data on the watershed and achieve dynamic perception of the physical watershed, including:

[0044] Space monitoring unit: A high-resolution remote sensing satellite constellation equipped with visible light / infrared imaging payloads can acquire basin-specific remote sensing images with a spatial resolution of ≤10m, which can be used to extract macroscopic features such as river systems and land use.

[0045] Aerial monitoring platform: equipped with a specific LiDAR specific laser radar and inertial navigation UAV, collecting point cloud density ≥ 50 specific points specific / m 2 Specific 3D point cloud data of embankments and riverbeds, and simultaneous acquisition of optical images to construct texture models;

[0046] Ground sensor array: water level gauges, current meters, piezometers, and specific GPS displacement stations distributed in rivers, reservoirs, and embankments, collecting minute-level hydrological data and project status information in real time;

[0047] Data communication module: Based on the specific 5G and Beidou system's time and space synchronization transmission network, it ensures that the time synchronization error of multi-source data is ≤50ms and the spatial registration error is ≤10cm.

[0048] 2. Digital Twin Data Middle Platform: Used to build a full-factor digital backplane and data management, including:

[0049] Spatiotemporal data fusion unit: Through coordinate conversion and timestamp calibration, it fuses satellite images, LiDAR point clouds, and sensor data into a three-dimensional digital baseplate with a unified spatiotemporal reference, covering all elements such as topography, river systems, and embankment projects;

[0050] Multi-source data governance module: cleans, interpolates, and removes outliers from monitoring data, constructs a multidimensional database containing historical floods, engineering parameters, and ecological indicators, and supports second-level spatiotemporal queries;

[0051] Data visualization engine: Renders digital baseplates and real-time data into interactive 3D scenes, supporting dynamic visualization of watershed topography and flood evolution.

[0052] 3. Physical Intelligent Algorithm Coupling Computing Subsystem: used to implement multi-physical process modeling and intelligent simulation, including:

[0053] Physical Process Modeling Unit:

[0054] Hydrodynamic model: Based on the improved Saint-Venant equations and physical information neural network, it simulates the nonlinear hydraulic characteristics of flood evolution;

[0055] Material degradation model: Based on quantum field theory, a microscopic degradation model of embankment materials is constructed to analyze the strength attenuation process under flood erosion;

[0056] Ecological response model: coupling hydrodynamic conditions and ecological indicators to predict the impact of floods on watershed ecosystems;

[0057] Multi-scale simulation engine:

[0058] Adaptive grid module: uses a fine grid of ≤5m in key areas and a coarse grid in other areas, and achieves minute-level simulation of the entire river basin through a parallel computing framework;

[0059] Extreme Scenario Generator: Generates super-standard flood scenarios based on generative adversarial networks, supporting simulation of complex disaster conditions;

[0060] Algorithm integration interface: Realizes real-time interaction among multiple models through the data bus, and supports bidirectional coupling calculations of hydrodynamic-specific-material-specific-ecological processes.

[0061] 4. Intelligent Decision Support Subsystem: used for risk assessment and scheduling plan generation, including:

[0062] Multi-objective optimization module: Based on the reinforcement learning algorithm, it generates a Pareto optimal scheduling solution set with the goals of flood control safety, water resource utilization, and ecological protection;

[0063] Uncertainty Quantification Unit: Uses Bayesian model averaging to calculate risk probability distribution and generate flood risk heat maps with confidence intervals;

[0064] Cognitive twin decision module: Integrates historical cases and expert experience based on knowledge graphs, and optimizes human-machine collaborative solutions through natural language interaction;

[0065] Emergency Plan Generator: Automatically generates emergency response plans such as personnel transfer and material deployment based on risk heat maps, and supports feasibility simulation verification.

[0066] 5. Human-computer interaction and application display subsystem: used for system operation and result presentation, including:

[0067] 3D visualization interactive platform: Based on the Unreal Engine, an immersive simulation environment is built to support decision makers to view the flood evolution process and the effects of the dispatch plan through VR / AR devices;

[0068] Real-time monitoring of specific dashboards: Integrate multi-source monitoring data and simulation results to display basin water conditions, project status, and risk warning information in graphical form;

[0069] Mobile terminal application: A lightweight and specific APP for on-site personnel, providing real-time data query, risk warning reception and emergency instruction push functions.

[0070] The system workflow of the present invention is as follows:

[0071] Data collection and fusion: The air-space-ground monitoring subsystem acquires multi-source data in real time and integrates it into a full-factor digital backplane through the digital twin platform;

[0072] Physically specific and specific intelligent modeling: The algorithm coupling subsystem builds models of hydrodynamics, material degradation, etc. based on the digital baseboard, and generates extreme flood scenarios through multi-scale simulation;

[0073] Risk assessment and decision-making: The intelligent decision-making subsystem performs multi-objective optimization and uncertainty analysis on the simulation results to generate scheduling plans and emergency plans;

[0074] Interaction and application: Present results to decision makers through visualization platforms and mobile terminals, supporting human-machine collaborative optimization and instruction issuance.

[0075] The system of the present invention uses digital twin technology to achieve full-factor mapping and dynamic simulation of the physical watershed, and combines intelligent algorithms to improve the accuracy of flood risk simulation and the scientific nature of decision-making. It can be widely used in scenarios such as watershed flood control planning, engineering safety management and emergency response.

[0076] For other details about the technical solutions for implementing each module in the system of the above embodiment, please refer to the description of the virtual simulation and risk rehearsal method for flood control projects based on digital twins in the above embodiment, which will not be repeated here.

[0077] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similarities between the various embodiments can be referred to in conjunction with each other. For system-related embodiments, since they are generally similar to method-related embodiments, their description is relatively simple. For relevant details, refer to the description of the method-related embodiments.

[0078] An electronic device according to an embodiment of the present disclosure includes a memory and a processor. The memory is used to store non-transitory computer-readable instructions. Specifically, the memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM) and / or cache memory (cache), etc. The non-volatile memory may, for example, include a read-only memory (ROM), a hard disk, a flash memory, etc.

[0079] The processor can be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and can control other components in the electronic device to perform desired functions. In one embodiment of the present disclosure, the processor is used to execute the computer-readable instructions stored in the memory, so that the electronic device executes all or part of the steps of the digital twin-based flood control project virtual simulation and risk pre-rehearsal method described in each embodiment of the present disclosure.

[0080] Those skilled in the art should understand that in order to solve the technical problem of how to obtain a good user experience, this embodiment may also include well-known structures such as a communication bus and an interface, and these well-known structures should also be included in the scope of protection of this disclosure.

[0081] like Figure 3The present invention provides a schematic structural diagram of an electronic device according to an embodiment of the present invention, which is suitable for implementing the electronic device according to an embodiment of the present invention. Figure 3 The electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present disclosure.

[0082] like Figure 3 As shown, the electronic device may include a processor (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) or a program loaded from a storage device into a random access memory (RAM). In the RAM, various programs and data required for the operation of the electronic device are also stored. The processor, ROM, and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0083] Typically, the following devices can be connected to the I / O interface: input devices such as sensors or visual information acquisition devices; output devices such as display screens; storage devices such as tapes and hard disks; and communication devices. The communication device allows the electronic device to communicate with other devices (such as edge computing devices) wirelessly or by wire to exchange data. Figure 3 The electronic device is shown with various devices, but it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed instead.

[0084] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processor, all or part of the steps of the method for virtual simulation and risk pre-rehearsal of flood control projects based on digital twins of the embodiment of the present disclosure are executed.

[0085] For detailed description of this embodiment, please refer to the corresponding description in the aforementioned embodiments, which will not be repeated here.

[0086] According to an embodiment of the present disclosure, a computer-readable storage medium stores non-transitory computer-readable instructions. When executed by a processor, the non-transitory computer-readable instructions execute all or part of the steps of the digital twin-based flood control project virtual simulation and risk rehearsal method described in each embodiment of the present disclosure.

[0087] The above-mentioned computer-readable storage media include, but are not limited to, optical storage media (e.g., CD-ROMs and DVDs), magneto-optical storage media (e.g., MOs), magnetic storage media (e.g., magnetic tapes or mobile hard disks), media with built-in rewritable non-volatile memory (e.g., memory cards), and media with built-in ROM (e.g., ROM cartridges).

[0088] For detailed description of this embodiment, please refer to the corresponding description in the aforementioned embodiments, which will not be repeated here.

[0089] The basic principles of the present disclosure have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, and effects mentioned in this disclosure are merely illustrative and not restrictive, and should not be construed as necessarily possessed by each embodiment of the present disclosure. Furthermore, the specific details disclosed above are provided for illustrative purposes and to facilitate understanding, rather than as limitations. These details do not limit the present disclosure to necessarily being implemented using these specific details.

[0090] In the present disclosure, relational terms such as first and second, etc. are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. The block diagrams of the devices, devices, equipment, and systems involved in the present disclosure are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, devices, equipment, and systems can be connected, arranged, or configured in any manner. Words such as "including," "comprising," "having," and the like are open-ended words, meaning "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or" and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0091] Additionally, as used herein, "or" used in a list of items beginning with "at least one" indicates a separate list, so that, for example, a list of "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not mean that the example described is preferred or better than other examples.

[0092] It should also be noted that in the system and method of the present disclosure, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present disclosure.

[0093] Various changes, substitutions, and modifications may be made to the technology described herein without departing from the teachings defined by the appended claims. Moreover, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, compositions of things, means, methods, and actions described above. Currently existing or later developed processes, machines, manufactures, compositions of things, means, methods, or actions that perform substantially the same function or achieve substantially the same results as the corresponding aspects described herein may be utilized. Accordingly, the appended claims include within their scope such processes, machines, manufactures, compositions of things, means, methods, or actions.

[0094] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0095] The above description has been provided for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present disclosure to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A virtual simulation and risk rehearsal method for flood control projects based on digital twins, characterized by: The following steps are involved: Acquire multi-source watershed data collected by an integrated air-ground-space monitoring network, including satellite remote sensing images, UAV LiDAR point clouds, and ground sensor monitoring data; Inputting the multi-source data into a pre-built digital twin model, the model comprising a spatiotemporal data fusion layer, a physical process modeling layer, a multi-scale simulation deduction layer, and a risk assessment decision layer connected in sequence; The spatiotemporal data fusion layer performs spatiotemporal registration and feature fusion on the input multi-source data to construct a full-factor digital baseboard; The physical process modeling layer is based on the digital baseboard and constructs a hydrodynamic model, a levee material degradation model and an ecological response model through a physical information neural network coupling algorithm; The multi-scale simulation and deduction layer performs multi-scale grid division and parallel calculation on the model output by the physical process modeling layer, generates extreme flood scenarios through a generative adversarial network, and performs evolutionary previews; The risk assessment decision-making layer generates flood control scheduling plans based on multi-scale simulation results through a multi-objective reinforcement learning algorithm, and optimizes human-machine collaborative decision-making through a cognitive twin model; According to the flood control scheduling plan output by the risk assessment decision layer, a heat map representing the flood risk distribution and an emergency response plan are generated, and the heat map is marked with risk confidence intervals based on an uncertainty quantification algorithm.

2. The method for virtual simulation and risk rehearsal of flood control projects based on digital twins according to claim 1 is characterized by: The acquisition of multi-source watershed data collected by the integrated air-space-ground-integrated monitoring network includes: Acquire remote sensing images from satellite constellations equipped with high-resolution imaging payloads. These images contain visible, near-infrared, and short-wave infrared band data, which are used to extract river system distribution, vegetation coverage, and land use types. Through the UAV platform equipped with laser radar and inertial navigation system, the point cloud density is ≥50 points / m 2 The levee and riverbed LiDAR point cloud data are collected simultaneously with optical images to construct a 3D texture model; Collect data from ground sensors deployed at river sections, reservoir dams, and embankment projects, including water level gauge data with a sampling frequency of ≥ 1 time / minute, water velocity profiles monitored by current meters, pore water pressure on embankments collected by piezometers, and millimeter-level deformation data from GPS displacement monitoring stations; The multi-source data are unified in time and space through the 5G communication network and the BeiDou-3 positioning system, forming a three-dimensional monitoring data set covering the basin's topography, hydrological characteristics and engineering status.

3. The method for virtual simulation and risk rehearsal of flood control projects based on digital twins according to claim 1 is characterized by: In the digital twin model: Spatiotemporal data fusion layer: This layer is used to input satellite remote sensing images, UAV LiDAR point clouds, and ground sensor data into the spatiotemporal registration module. Through spatial coordinate conversion and time stamp synchronization, it constructs a full-factor digital baseplate covering the basin's topography, hydrology, and engineering facilities. Physical process modeling layer: Based on the digital baseplate, a hydrodynamic model is constructed to simulate the flood evolution process, a levee material degradation model is established to analyze the impact of long-term flood erosion on engineering structures, and multi-physics field coupling modeling is achieved through physical information neural networks; Multi-scale simulation layer: This layer is used to adaptively mesh the flow field data output by the physical process model. Fine meshes are used in key areas to improve computational accuracy, while coarse meshes are used in other areas to balance computational efficiency. Generative adversarial networks are used to generate a variety of extreme flood scenarios and simulate flood evolution across the entire river basin. Risk assessment decision-making layer: used to input simulation results into the multi-objective optimization model to generate a scheduling plan that takes into account flood control safety, water resource utilization, and ecological protection; The scheduling plan is evaluated collaboratively by humans and machines through the cognitive twin model, and flood control scheduling plans are output based on historical case analysis.

4. The method for virtual simulation and risk rehearsal of flood control projects based on digital twins according to claim 3 is characterized by: The spatiotemporal data fusion layer performs spatiotemporal registration and feature fusion on the input multi-source data to construct a full-factor digital baseboard, including: The geographic coordinate system of satellite remote sensing images is converted into a unified basin coordinate system, and then matched with the spatial coordinates of the UAV LiDAR point cloud. The spatial deviation of different data sources is eliminated through feature point extraction and matching algorithms. Add timestamps to real-time monitoring data collected by ground sensors and synchronize them with the imaging time of satellite images and the time series of drone scans to ensure consistency of multi-source data in the time dimension; By integrating and processing satellite images, LiDAR point clouds and sensor data, we construct a three-dimensional digital base plate that includes elements such as the basin topography, river systems, embankment projects, and hydrological monitoring points, realizing a full-factor digital mapping of the physical basin.

5. The method for virtual simulation and risk rehearsal of flood control projects based on digital twins according to claim 1 is characterized by: The physical process modeling layer is based on the digital baseboard and uses a physical information neural network coupling algorithm to construct a hydrodynamic model, a levee material degradation model, and an ecological response model, including: Hydrodynamic model construction: The terrain elevation, river cross-section, and other data from the digital baseboard are input into the modified Saint-Venant equations. A physical information neural network is used to learn the nonlinear mapping of the relationship between water level and flow, thereby constructing a hydrodynamic model that can reflect the complex hydraulic characteristics of the flood process. Construction of a levee material degradation model: Based on the levee structural parameters in the digital baseplate, material composition and construction process information are input into a quantum field theory-driven microstructure model. This model simulates material degradation processes such as concrete carbonization and steel corrosion under flood erosion, and establishes a degradation equation for material strength over time. Ecological response model construction: Extract river ecological parameters from the digital baseplate, correlate hydrodynamic conditions with ecosystem indicators, and construct an ecological response model using a physical information neural network coupling algorithm to predict the impact of flood processes on the basin ecosystem; Multi-model coupling integration: Through the physical information neural network coupling algorithm, the hydrodynamic model, embankment material degradation model and ecological response model are integrated to achieve a two-way coupling simulation of the flood evolution process, embankment structure safety and ecosystem changes.

6. The method for virtual simulation and risk rehearsal of flood control projects based on digital twins according to claim 1 is characterized by: The multi-scale simulation layer performs multi-scale grid division and parallel calculation on the model output by the physical process modeling layer, generates extreme flood scenarios through a generative adversarial network, and performs evolutionary previews, including: Multi-scale gridding: For the hydrodynamic model and embankment material degradation model output by the physical process modeling layer, differentiated gridding is performed based on the complexity of the basin terrain and the importance of the project. Fine grids are used in river bends and critical areas of embankment projects to capture local hydraulic characteristics, while coarse grids are used in open river sections to improve computational efficiency, forming a grid system that balances accuracy and efficiency. Parallel computing framework construction: Deploy multi-scale grid models to a distributed computing platform, split and distribute computing tasks through a message passing interface, and leverage GPU clusters to accelerate large-scale flow field calculations. This allows for parallel solution of flood evolution processes across the entire river basin, shortening simulation time. Extreme Scenario Generation: Generative adversarial networks are used to learn the distribution characteristics of historical flood data. By inputting climate change scenarios and basin human activity parameters, extreme flood scenarios that exceed historical records are generated. Flood evolution rehearsal: The generated extreme scenarios are input into the multi-scale grid model to drive the coupled calculation of the hydrodynamic model and the embankment material degradation model, simulating the evolution of floods in the basin, the safety changes of embankment structures and the response of the ecosystem, and realizing the dynamic rehearsal of flood risks.

7. The method for virtual simulation and risk rehearsal of flood control projects based on digital twins according to claim 1 is characterized by: The risk assessment decision layer generates flood control scheduling plans based on multi-scale simulation results through a multi-objective reinforcement learning algorithm and optimizes human-machine collaborative decision-making through a cognitive twin model, including: Multi-objective optimization model construction: Using the inundation range, levee safety factor, and ecological flow index obtained from flood evolution simulations as constraints, and taking flood control safety guarantee rate, water resource utilization rate, and ecological protection benefits as optimization objectives, a multi-objective function system is established. A multi-objective reinforcement learning algorithm is used to optimize the decision variables for reservoir group operation and flood diversion area activation, generating a Pareto optimal operation plan set with different risk preferences. Scheduling Scheme Risk Assessment: Based on uncertainty quantification algorithms, this study analyzes the impact of errors in extreme scenario generation and model parameters on the scheduling scheme, and calculates the flood risk probability distribution of each scheme. By constructing a risk indicator system for a composite flood, engineering, and social system, this study comprehensively assesses the secondary disasters and economic losses that may be caused by the scheduling scheme, and annotates the risk confidence intervals of the schemes. Cognitive twin model construction: Integrate historical flood control cases, expert decision-making logic, and watershed management knowledge to build a cognitive twin model based on a knowledge graph. Use natural language processing technology to transform scheduling plans into semantic networks, perform similarity matching with historical case libraries, extract referenceable decision-making experience, and generate solution optimization suggestions. Human-machine collaborative decision-making optimization: The scheduling plan and risk assessment results generated by multi-objective optimization are input into the cognitive twin model and presented to decision makers through a visual interactive interface; the attention mechanism is used to highlight key risk points in the plan, supporting decision makers to adjust scheduling parameters through natural language interaction. The model provides real-time feedback on risk changes after adjustment, forming a collaborative decision-making closed loop of machine generation, human optimization, and machine verification, and ultimately outputting a flood control scheduling plan that is both scientific and operational.

8. The method for virtual simulation and risk rehearsal of flood control projects based on digital twins according to claim 1 is characterized by: The flood control scheduling plan output by the risk assessment decision layer generates a heat map representing the flood risk distribution and an emergency response plan, wherein the heat map is annotated with risk confidence intervals based on an uncertainty quantification algorithm, including: Calculation of flood risk indicators: Based on the flood evolution simulation results of the scheduling plan, the inundation depth, flow velocity, and duration parameters of each region are extracted. Combined with the vulnerability curve of the hazard-bearing body, the flood risk index is calculated to construct a quantitative indicator system reflecting the regional risk level; Uncertainty Quantification and Visualization: Bayesian model averaging is used to analyze the propagation paths of multi-source model prediction errors and calculate the probability distribution of inundation range and peak flow parameters. The risk index and uncertainty results are mapped into a visual heat map, with color gradients representing risk levels and transparency or contour lines marking risk confidence intervals, presenting the spatial distribution of flood risk in the basin and the range of uncertainty. Emergency response plan generation: Based on high-risk areas identified by heat maps, combined with data such as topography, population distribution, and infrastructure layout, differentiated emergency response plans are automatically generated. These plans include personnel transfer route planning, material storage point site selection and deployment plans, and key facility protection measures. For areas with large risk confidence intervals, multiple alternative plans are simultaneously generated to address forecast uncertainties. Feasibility verification and optimization of the plan: The emergency response plan is input into the digital twin model for simulation verification to evaluate the traffic capacity of the transfer route, the timeliness of material allocation, and the effectiveness of facility protection measures; the plan is iteratively optimized through reinforcement learning algorithms until the emergency response time window requirements are met, and finally a complete emergency disposal plan is formed that includes risk heat maps, regional response measures and alternative plans.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the digital twin-based flood control project virtual simulation and risk pre-rehearsal method described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, which are used to enable a computer to execute the digital twin-based flood control project virtual simulation and risk pre-rehearsal method described in any one of claims 1 to 8.

Images