Urban flood process simulation and dispatching control method and system

Abstract

The application discloses a simulation and dispatching control method and system for urban flood process. The application belongs to the technical field of urban water conservancy system control. The method comprises the following steps: collecting and constructing a specific type of basic data set, respectively constructing a hydrological runoff model, a one-dimensional water power model of a drainage pipe network, a river network mixed water power model and an engineering dispatching model; performing three-dimensional dynamic coupling to form a water resource full-space dynamic model; using at least one historical flood event in historical flood data to calibrate parameters and verify accuracy of the water resource full-space dynamic model; after verification, using the water resource full-space dynamic model to build a digital simulation platform to simulate flood and determine a dispatching scheme. The technical scheme can realize multi-scenario dynamic simulation, automatic comparison and selection after generation of a dispatching scheme, enhance system adaptability and response timeliness under extreme weather, and provide an efficient and accurate flood control and drainage scheme generation mechanism for urban area flood control and drainage.

Description

Technical Field

[0001] This application belongs to the field of urban water conservancy system control technology, specifically relating to a simulation and scheduling control method and system for urban flooding processes. Background Technology

[0002] Current urban flood control mainly relies on the separate management of water conservancy projects and municipal drainage systems, and simulation and scheduling mostly adopt a single-process analysis approach.

[0003] Currently, scheduling relies heavily on manual experience, lacking integrated scheduling that incorporates multiple projects such as dams, reservoirs, and drainage pumping stations. This results in poor coordination between water conservancy and municipal systems, hindering efficient scheduling and contingency planning. Furthermore, current methods suffer from low computational efficiency, reliance on external software for some technologies, difficulty in achieving data collaboration and coordinated control, poor adaptability to extreme weather scenarios, and delayed responses. In summary, existing technologies cannot achieve high-precision simulation of the entire urban flooding process or cross-system collaborative intelligent scheduling, making it difficult to support safe and efficient flood control and drainage decision-making in highly urbanized areas. Summary of the Invention

[0004] This application provides a simulation and scheduling control method and system for urban flooding processes, aiming to solve the technical problems in existing urban flood control, such as the separation of simulation and scheduling, poor coordination among multiple projects, reliance on experience-based decision-making, and low computational efficiency, and to achieve high-precision coupled simulation and intelligent and efficient coordinated scheduling of the entire urban flooding process.

[0005] In a first aspect, embodiments of this application provide a method for simulating and controlling urban flooding processes, the method comprising:

[0006] Collect and construct a specific type of basic dataset, which includes at least one of the following: topographic data, land use data, meteorological data, water system engineering data, drainage network data, pumping station characteristic data, historical flood data, and integrated monitoring data; Based on the aforementioned basic dataset, a hydrological runoff generation and confluence model, a one-dimensional hydraulic model of drainage pipe network, a mixed hydrodynamic model of river network, and an engineering scheduling model are constructed respectively. The hydrological runoff model, the one-dimensional hydraulic model of the drainage network, and the mixed hydrodynamic model of the river network are dynamically coupled in three dimensions to form a dynamic model of water resources in the whole space. The parameters of the water resources full-space dynamic model are calibrated and the accuracy is verified using at least one historical flood event from historical flood data. After successful verification, a digital simulation platform is built using the aforementioned full-space dynamic model of water resources to simulate floods and determine scheduling schemes.

[0007] Furthermore, the terrain data includes at least one of the following: DEM digital elevation model, slope, aspect, and depth and area of ​​depressions; The land use data includes at least one of the following: impervious surface ratio, green area, and water coverage ratio; The meteorological data includes at least one of hourly rainfall, hourly temperature, hourly wind speed, hourly humidity, and long-term annual rainfall statistics; The water system engineering data includes at least one of the following: river channel cross-sectional dimensions, riverbed roughness, dam location, opening threshold, and reservoir capacity curve. The drainage network data includes at least one of the following: pipe section length, diameter, material, slope, manhole coordinates, and storm drain grate distribution. The pump station characteristic data includes at least one of the following: single pump flow rate, head, power curve, start / stop threshold, and joint control logic; The historical flood data includes at least one of the following: historical rainstorm events, corresponding water levels, flow rates, inundation areas, and water depth time series. The integrated monitoring data includes at least one of radar rainfall, satellite rainfall, urban flooding video, and real-time water level station data.

[0008] Furthermore, constructing the hydrological runoff model includes the following sub-steps: The underlying surface type, soil moisture content, and terrain slope are determined based on the aforementioned basic dataset. An improved Xin'anjiang model was used to construct the runoff structure, and Horton's formula was used to calculate the surface runoff. In the runoff calculation, correction factors for impermeable surface expansion and terrain slope are introduced; A nonlinear reservoir model and the Muskingen method were used to construct the runoff structure and calculate the surface runoff process. Based on the distribution of rainfall stations, the spatial distribution of areal rainfall was calculated using the Thiessen polygon method, and the model was constructed.

[0009] Furthermore, a one-dimensional hydraulic model of the drainage network is constructed, including: Establish the pipe segment topology and node relationships based on the drainage network data; Establish one-dimensional unsteady flow control equations based on the Saint-Venant equations or the dynamic wave method; The one-dimensional finite volume method is used to spatially discretize the pipe network segments; A dynamic correction model for the resistance coefficient is introduced to reflect the impact of pipe section siltation and aging on flow capacity; The study area is divided into several hydrological units, with each hydrological unit corresponding to at least one inspection well or storm drain grate, to achieve refined calculation of the inflow into the pipe network and obtain a one-dimensional hydraulic model of the drainage pipe network.

[0010] Furthermore, the hybrid hydrodynamic model of the river network is constructed, including: Based on the aforementioned water system engineering data, the scope of key waterways and flood-prone low-lying areas was extracted; A model was constructed for the main waterway using the one-dimensional Saint-Venant equations, and the Preissmann four-point implicit difference was used for the solution. A two-dimensional shallow water equation model was constructed for flood-prone low-lying areas, and the model was discretized using the finite volume method and adaptive mesh refinement was applied. By generalizing sluices, dams, and pumping stations as hydraulic nodes and using the weir flow formula to simulate the flow capacity of the project, a mixed hydrodynamic model of the river network is obtained.

[0011] Furthermore, the project scheduling model is constructed, including: A scheduling knowledge base is constructed based on flood control and drainage rules, gate and pump operation rules, and pipeline network scheduling rules; The scheduling logic was written as an executable script and embedded into the hydrodynamic model calculation process; We construct a deep reinforcement learning optimization structure with the goals of minimizing the flooded area, maximizing drainage efficiency, and minimizing energy consumption. Based on the real-time output of water level and flow status from the model, scheduling instructions are generated to obtain the engineering scheduling model.

[0012] Furthermore, the three-dimensional dynamic coupling forms a full-space dynamic model of water resources, including: Two-way water exchange between the surface and the pipe network is achieved through inspection wells and storm drains. When the surface water level is lower than the node water level, the runoff flows into the pipe network, and vice versa. Using the river water level at the outlet as the dynamic boundary, the rolling coupling calculation between the pipeline network and the river is performed according to the preset time step; The engineering scheduling model reads the calculation results of the hydrodynamic model in real time and outputs control commands to update the boundary conditions, forming a closed-loop coupling.

[0013] Furthermore, the step of using at least one historical flood event from historical flood data to calibrate and verify the accuracy of the full-space dynamic model of water resources includes: At least two independent historical rainstorm and flood events were selected, and a hybrid optimization algorithm consisting of genetic algorithm and particle swarm optimization algorithm was used to calibrate the parameters of roughness coefficient, confluence coefficient and drag coefficient. Select at least one historical rainstorm and flood event that was not included in the rate determination for verification; The accuracy verification is passed when the river flood meets the preset maximum water level error and preset flow error, and the waterlogging meets the preset inundation range error and preset water depth error.

[0014] Furthermore, a digital simulation platform is built using the aforementioned full-space dynamic model of water resources to simulate floods and determine scheduling schemes, including: Constructing a 3D visualization scenario of urban flooding based on digital twins; By accessing real-time rainfall monitoring data, hydrological monitoring data, and engineering operation monitoring data, the dynamic model of water resources in the whole space is driven to perform dynamic simulation. Automatically generate multi-scenario scheduling schemes and perform quantitative evaluations, and use the analytic hierarchy process to determine the optimal scheduling scheme; Output the optimal scheduling scheme.

[0015] Secondly, embodiments of this application provide a simulation and scheduling control device for urban flooding processes, the device being configured as follows: Collect and construct a specific type of basic dataset, which includes at least one of the following: topographic data, land use data, meteorological data, water system engineering data, drainage network data, pumping station characteristic data, historical flood data, and integrated monitoring data; Based on the aforementioned basic dataset, a hydrological runoff generation and confluence model, a one-dimensional hydraulic model of drainage pipe network, a mixed hydrodynamic model of river network, and an engineering scheduling model are constructed respectively. The hydrological runoff model, the one-dimensional hydraulic model of the drainage network, and the mixed hydrodynamic model of the river network are dynamically coupled in three dimensions to form a dynamic model of water resources in the whole space. The parameters of the water resources full-space dynamic model are calibrated and the accuracy is verified using at least one historical flood event from historical flood data. After successful verification, a digital simulation platform is built using the aforementioned full-space dynamic model of water resources to simulate floods and determine scheduling schemes.

[0016] Thirdly, embodiments of this application provide a simulation and scheduling control system for urban flooding processes. The simulation and scheduling control system for urban flooding processes includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the steps of the method described in the first aspect.

[0017] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0018] Fifthly, embodiments of this application provide a chip, the chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the method as described in the first aspect.

[0019] The technical solution provided in this application can realize multi-model coupled simulation and intelligent scheduling in urban flood scenarios, effectively solving the shortcomings of existing technologies such as single simulation dimension, insufficient scheduling coordination, and reliance on experience-based decision-making, and significantly improving the accuracy and computational efficiency of flood simulation. Simultaneously, this solution can realize multi-scenario dynamic simulation and automatic comparison and selection after scheduling scheme generation, enhancing the system's adaptability and response timeliness under extreme weather conditions, and providing efficient and accurate flood control and drainage scheme generation for urban areas. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the simulation and scheduling control method for urban flooding processes provided in Embodiment 1 of this application; Figure 2 This is a schematic diagram of the structure of the simulation and scheduling control system for urban flooding processes provided in Embodiment 2 of this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but may also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subroutine, etc.

[0022] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0023] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0024] The following description, in conjunction with the accompanying drawings, details the urban flooding process simulation and scheduling control method and system provided in this application through specific embodiments and application scenarios.

[0025] Example 1 Figure 1 This is a flowchart illustrating the simulation and scheduling control method for urban flooding processes provided in Embodiment 1 of this application. Figure 1 As shown, the specific steps include the following: S11, collect and build a basic dataset of a specific type.

[0026] The foundational dataset can be a collection of multi-source, heterogeneous data used to support flood process simulation, model building, parameter calibration, and scheduling decisions. Its completeness, accuracy, and timeliness directly determine the reliability of the model's calculation results and the usability of the decisions. This dataset includes at least one of the following: topographic data, land use data, meteorological data, water system engineering data, drainage network data, pumping station characteristic data, historical flood data, and integrated monitoring data.

[0027] This solution can be achieved through on-site surveys, industry data compilation, real-time sensor data acquisition, remote sensing inversion, and underground pipeline detection. It is completed through standardized formatting, spatiotemporal alignment, quality verification, removal of abnormal data, and completion of missing data.

[0028] S12, Based on the aforementioned basic dataset, construct the hydrological runoff generation and confluence model, the one-dimensional hydraulic model of the drainage network, the mixed hydrodynamic model of the river network, and the engineering scheduling model, respectively.

[0029] Among them, the hydrological runoff generation and confluence model can be used to simulate the complete hydrological process of surface runoff generated by different underlying surface types in urban areas under rainfall conditions, as well as the runoff confluence, collection, and conduction along the terrain.

[0030] A one-dimensional hydraulic model of a drainage network can be used to simulate the internal water flow state, pressure distribution, full flow and non-full flow conversion, backflow, return flow, and actual drainage capacity of a municipal drainage network.

[0031] The hybrid hydrodynamic model of the river network can be used to simultaneously simulate the evolution of floods in the main river channels, changes in water level and flow, river regulation and storage, as well as surface runoff, flooding range expansion, and spatial and temporal distribution of water depth in flood-prone low-lying areas.

[0032] The engineering scheduling model can be used to automatically generate collaborative control strategies and operation instructions for water conservancy and municipal engineering projects such as dams, reservoirs, and drainage pumping stations based on real-time water conditions, operating conditions, and preset scheduling rules.

[0033] Specifically, the model structure can be built, initial parameters can be assigned, algorithm programs can be loaded and computational logic can be configured based on the basic dataset, forming a multi-module model system that can operate independently and collaborate with each other.

[0034] S13, the hydrological runoff model, the one-dimensional hydraulic model of the drainage network, and the mixed hydrodynamic model of the river network are dynamically coupled in three dimensions to form a dynamic model of water resources in the whole space.

[0035] Three-dimensional dynamic coupling can realize real-time bidirectional transmission and computational step coordination of key hydraulic elements such as water volume, water level, flow rate, and flow velocity in three physical spatial dimensions: surface runoff, pipeline transmission, and river flood discharge.

[0036] The resulting dynamic water resources model can be an integrated, coupled computational model that unifies rainfall, runoff generation and collection, pipeline drainage, river flood control, and engineering regulation within the same spatiotemporal framework. Specifically, it can achieve dynamic updates of boundary conditions, interactive transfer of hydraulic elements, and synchronization of computational time sequences through an interactive interface.

[0037] S14, At least one historical flood event from the historical flood data is used to calibrate the parameters and verify the accuracy of the full-space dynamic model of water resources.

[0038] Among them, parameter calibration refers to the model calibration process that uses historical measured flood data to adjust key parameters of the model through iterative optimization, so that the simulation results gradually approach the measured data.

[0039] Accuracy verification refers to the evaluation process of using independent historical flood events that were not calibrated to test the model's generalization ability, computational stability, and reliability of results.

[0040] This solution can be driven by historical flood data and is achieved through error calculation, parameter optimization, and accuracy index determination.

[0041] S15. After verification, a digital simulation platform is built using the aforementioned full-space dynamic model of water resources to conduct flood simulation and determine the scheduling scheme.

[0042] A digital simulation platform can be an integrated flood control and drainage decision support platform that integrates model calculation, visualization rendering, scenario simulation, scheme comparison and selection and risk warning functions.

[0043] Specifically, model integration, interface rendering, real-time data access, and decision result output can be achieved through development and interface encapsulation. Real-time simulation and inundation prediction of flood processes can be carried out for current or future rainfall scenarios. Based on simulation calculation results, multiple scheduling schemes can be quantitatively evaluated and the optimal scheduling strategy can be selected.

[0044] The technical solution provided in this embodiment achieves integrated simulation and intelligent decision-making for the entire process of urban flooding, from rainfall, runoff generation and drainage, to river flood discharge and engineering scheduling, through multi-source data acquisition and integration, independent construction of multi-module models, dynamic coupling in three-dimensional space, calibration and verification of historical data, and the construction of a digital platform. It effectively solves the problems of fragmented simulation and decentralized scheduling in traditional technologies, significantly improves the accuracy of flood simulation, computational efficiency, and scientific nature of scheduling decisions, and provides stable and reliable technical support for flood control and drainage in highly urbanized areas.

[0045] In one embodiment, optionally, the terrain data includes at least one of DEM, slope, aspect, and depth and area of ​​depressions.

[0046] A DEM (Digital Elevation Model) is a dataset that uses limited terrain elevation data to digitally represent the ground topography. It can accurately reflect the regional topographic relief and elevation distribution, providing basic topographic data for runoff path and inundation analysis.

[0047] Slope is a quantitative indicator of the steepness or gentleness of a land surface unit, which directly affects the surface runoff velocity, runoff path, and conditions for water accumulation.

[0048] Slope aspect refers to the orientation of a terrain slope, used to determine runoff direction, confluence trend, and differences in solar evaporation, which further affects changes in the underlying surface moisture content.

[0049] The depth and area of ​​low-lying points are used to accurately characterize the scale of water accumulation space, inundation potential, and distribution characteristics of waterlogging risk in pot-bottom depression terrain.

[0050] The impermeable surface ratio refers to the proportion of urban hardened surface to the total area of ​​the region, which directly determines the rainfall runoff capacity, runoff coefficient, and infiltration trend.

[0051] Green space area is used to characterize the surface permeability, water storage, water retention and rainwater absorption capacity of a region, and is an important factor affecting the rate of runoff generation.

[0052] The water coverage ratio is used to reflect the spatial distribution of natural rivers, lakes, and water storage bodies, and to demonstrate the region's natural water storage capacity.

[0053] Hourly rainfall is the core input data that drives hydrological and hydrodynamic models, directly determining the runoff and flood process.

[0054] Temperature, wind speed, and humidity are used to calculate the evaporation capacity of a watershed, changes in the moisture content of the underlying surface, and changes in the dryness and wetness of the soil.

[0055] Long-term annual rainfall statistics are used to conduct rainstorm frequency analysis, design rainstorm calculations, and multi-scenario rainfall simulations, providing a basis for risk assessment.

[0056] The cross-sectional dimensions of a river channel determine its water-carrying area and flood capacity, and are key geometric parameters for river hydrodynamic calculations.

[0057] Riverbed roughness reflects the resistance of the riverbed surface and directly affects the river flow velocity, water level, and flood progression speed.

[0058] The location of the dam and the opening threshold are key parameters for engineering scheduling and control, which determine the scheduling trigger conditions and the control range.

[0059] The reservoir capacity curve is used to characterize the water level-storage capacity relationship of water storage structures such as reservoirs and lakes, reflecting their regulation and storage capacity and scheduling potential.

[0060] The length, diameter, material, and slope of the pipe section together determine the drainage capacity, flow resistance, water conveyance efficiency, and pressure change characteristics of the pipe network.

[0061] The coordinates of inspection wells and the distribution of storm drains are key nodes for surface runoff to enter the pipe network, and also the core interface for the exchange of water between the surface and the pipe network.

[0062] The pump station characteristic data may include at least one of the following: single pump flow rate, head, power curve, start / stop threshold, and joint control logic.

[0063] Single pump flow rate and head are the core performance indicators of a pumping station's drainage capacity, determining the actual drainage efficiency of the pumping station at different water levels.

[0064] The power curve is used to reflect the energy consumption level and operating efficiency of the pumping station under different operating conditions, providing a basis for energy-saving scheduling.

[0065] Start-stop thresholds and joint control logic are the rule basis for automatic operation, coordinated scheduling and drainage of pumping stations, and determine the scheduling mode of pumping station groups.

[0066] Historical rainfall events are used to drive model calibration and validation, recreating the actual flood occurrence process.

[0067] Water level, flow rate, inundation range, and water depth time series are the measured data used for model calibration to determine the reliability of the simulation results.

[0068] Radar and satellite-based rainfall monitoring can achieve large-scale, high spatiotemporal resolution rainfall monitoring, compensating for data errors caused by insufficient ground stations.

[0069] Urban flooding videos and real-time water level station data are used for real-time simulation verification, dynamic early warning, and comparison with actual flooding conditions.

[0070] In this embodiment, by comprehensively, meticulously, and clearly classifying and defining the basic dataset, all key data types supporting flood simulation are clearly defined one by one. This makes the data collection, processing, management, and storage processes more standardized, complete, and orderly. It can comprehensively cover all core information such as topography, underlying surface, meteorology, water conservancy projects, municipal pipe networks, pumping stations, historical floods, and real-time monitoring, significantly improving the accuracy, completeness, and standardization of model input data. It avoids problems such as incomplete model construction, large deviations in simulation results, and unreliable scheduling decisions caused by missing, ambiguous, unclear, or incomplete data elements from the source, laying a solid and reliable data foundation for subsequent high-precision model calculations, coupled simulations, and intelligent scheduling.

[0071] In one embodiment, optionally, constructing the hydrological runoff model includes the following sub-steps: The underlying surface type, soil moisture content, and terrain slope are determined based on the aforementioned basic dataset. An improved Xin'anjiang model was used to construct the runoff structure, and Horton's formula was used to calculate the surface runoff. In the runoff calculation, correction factors for impermeable surface expansion and terrain slope are introduced; A nonlinear reservoir model and the Muskingen method were used to construct the runoff structure and calculate the surface runoff process. Based on the distribution of rainfall stations, the spatial distribution of areal rainfall was calculated using the Thiessen polygon method, and the model was constructed.

[0072] The underlying surface type is the classification and identification result of the permeable and impermeable properties of the surface. Different types correspond to different runoff generation and infiltration characteristics, which are the prerequisites for runoff generation calculation.

[0073] Soil moisture content reflects the dryness or wetness of the soil before rainfall, and directly determines the initial infiltration capacity and the critical conditions for runoff generation. The higher the moisture content, the more runoff is generated.

[0074] The slope of the terrain is used to determine the confluence speed, confluence direction and water retention characteristics. The smaller the slope, the slower the confluence and the easier it is for water to accumulate.

[0075] The determination process is completed by the model processing unit through parsing, classifying, spatial matching, and parameter extraction of the basic dataset.

[0076] An improved Xin'anjiang model was used to construct the runoff structure, and Horton's formula was used to calculate the surface runoff.

[0077] The improved Xin'anjiang model is an optimized runoff calculation model for urban high impermeable surfaces, complex underlying surfaces, and pot-shaped depressions, which can more accurately reflect the runoff patterns in urban areas.

[0078] Horton's formula is a classic infiltration model that describes the decrease in soil infiltration capacity over time. It can accurately calculate the infiltration volume and excess infiltration flow during rainfall.

[0079] The impermeable surface expansion correction coefficient can be used to compensate for the increased flow rate caused by the increase in hardened ground due to rapid urbanization, making the model more consistent with the real urban environment.

[0080] The terrain slope correction coefficient can be used to correct the special hydrological characteristics of the pot-bottom depression terrain, such as concentrated flow, slow flow velocity, and easy water accumulation, thereby improving the calculation accuracy of local areas.

[0081] Nonlinear reservoir models can be used to simulate surface water storage, flood retention, slow flow, and water volume regulation processes, and are suitable for the confluence lag characteristics of complex urban terrain.

[0082] The Muskingan method is the most widely used and most stable flow calculation method in engineering, capable of stably simulating the runoff collection, conduction and evolution process.

[0083] The Thiessen polygon method is a classic spatial interpolation method that transforms discrete station rainfall data into continuous surface rainfall distribution, effectively solving the calculation bias caused by spatial unevenness of rainfall.

[0084] In this embodiment, a refined hydrological runoff generation and confluence model is constructed by combining the characteristics of the urban underlying surface with the special topography of the Guodiwa depression. The improved Xin'anjiang model and Horton infiltration formula are used to accurately describe the runoff generation pattern. Multiple correction coefficients are introduced to improve the adaptability to topography and urbanization. At the same time, a nonlinear reservoir model and Muskingen method are combined to achieve stable runoff calculation. Then, the Thiessen polygon method is used to complete the accurate interpolation of areal rainfall. This can comprehensively improve the accuracy, stability and applicability of urban area runoff simulation, provide high-precision and high-reliability runoff input for subsequent pipeline drainage and river flood control simulation, and improve the simulation quality of the overall model from the source.

[0085] In one embodiment, optionally, constructing a one-dimensional hydraulic model of the drainage network includes: Establish pipe segment topology and node relationships based on the drainage network data; Establish one-dimensional unsteady flow control equations based on the Saint-Venant equations or the dynamic wave method; The one-dimensional finite volume method is used to spatially discretize the pipe network segments; A dynamic correction model for the resistance coefficient is introduced to reflect the impact of pipe section siltation and aging on flow capacity; The study area is divided into several hydrological units, with each hydrological unit corresponding to at least one inspection well or storm drain grate, to achieve refined calculation of the inflow into the pipe network and obtain a one-dimensional hydraulic model of the drainage pipe network.

[0086] Pipeline topology and node relationships provide a complete description of the spatial connection structure, water flow transmission path, upstream and downstream relationships, connectivity and confluence relationships of the entire drainage network, and form the structural basis for hydraulic calculations of the network.

[0087] Specifically, this can be achieved by parsing, spatial matching, node numbering, connectivity checking, and topology reconstruction of pipeline network data.

[0088] The Saint-Venant equations, consisting of the continuity equation and the momentum equation, are the fundamental governing equations for describing unsteady flow in pipes and open channels. They can simulate various flow regimes such as pressure flow, open channel flow, and transitional flow, and have high computational accuracy.

[0089] The dynamic wave method is a simplified calculation method used when computing resources are limited or the model size is large, which improves computational efficiency while ensuring a certain level of accuracy.

[0090] The one-dimensional finite volume method is the numerical method with the best conservation and stability in fluid numerical calculations. It is suitable for simulating complex and intermittent flows and can guarantee the water balance in pipeline network calculations.

[0091] Spatial discretization involves dividing a continuous pipeline network into several computational units of finite length, enabling computers to perform numerical solutions.

[0092] The dynamic correction model for the resistance coefficient can dynamically adjust the hydraulic resistance coefficient based on the pipeline's operating years, degree of siltation, blockage status, and real-time monitoring data, thus restoring the true state of the pipeline's reduced drainage capacity due to aging and siltation.

[0093] This approach can embed the model process through dynamic parameter updates, resistance term correction, and real-time iterative calculations.

[0094] A hydrological unit is the smallest computational unit that precisely connects surface runoff generation and inflow with pipeline inflow, enabling inflow allocation and flow calculation on a plot-by-plot and node-by-node basis.

[0095] Refined calculations can accurately simulate detailed processes such as rainwater inflow, pipe network overload, overflow, backflow, and flooding, significantly improving the simulation accuracy of urban flooding points.

[0096] In this embodiment, a complete one-dimensional hydraulic model of the drainage network is constructed to fully restore the network topology, water flow state, pressure changes, and drainage capacity attenuation. At the same time, by combining hydrological unit division, the model achieves accurate connection between surface runoff and network inflow. This model can realistically simulate actual engineering phenomena such as network full flow, backlog, overflow, and river backflow, significantly improving the precision, realism, and reliability of municipal drainage system simulation. It provides accurate and reliable hydraulic calculation support for urban flooding evolution analysis, flood-prone area identification, pump station scheduling optimization, and drainage capacity assessment.

[0097] In one embodiment, optionally, constructing the river network hybrid hydrodynamic model includes: Based on the aforementioned water system engineering data, the scope of key waterways and flood-prone low-lying areas was extracted; A model of the main waterway is constructed using the one-dimensional Saint-Venant equations and solved using the Preissmann four-point implicit difference method. A two-dimensional shallow water equation model was constructed for flood-prone low-lying areas, and the model was discretized using the finite volume method and subjected to adaptive mesh refinement. By generalizing sluices, dams, and pumping stations as hydraulic nodes and using weir flow formulas to simulate the flow capacity of the projects, a mixed hydrodynamic model of the river network is obtained.

[0098] Main waterways are the core waterways that undertake the main functions of flood control, water conveyance, and water storage in the region, and determine the overall flood control pattern of the region.

[0099] Low-lying areas prone to flooding are key risk areas characterized by low-lying terrain, concentrated runoff, easy flooding, and a high likelihood of prolonged water accumulation.

[0100] Specifically, this can be accomplished through data filtering, spatial identification, terrain analysis, and boundary delineation.

[0101] The one-dimensional Saint-Venant equations can accurately simulate river water level, flow rate, flood propagation, river regulation and flood evolution.

[0102] The Preissmann four-point implicit difference is a classic numerical scheme for river hydrodynamic calculations, which is the most stable, adaptable, and capable of large step size calculations, making it suitable for long-distance river network calculations.

[0103] The model is constructed and solved by the river hydrodynamic module through equation establishment, grid discretization, iterative calculation and result output.

[0104] Two-dimensional shallow water equations are complete governing equations that describe surface runoff, flood expansion, water depth distribution, and velocity field, and can realistically simulate the flood inundation process.

[0105] The finite volume method ensures the conservation of water volume and momentum, making it suitable for simulating complex terrain and complex flows.

[0106] Adaptive mesh refinement automatically refines the mesh in areas with high risk, such as the edge of floods, large water depth gradients, and high risks, while simplifying the mesh in areas with stable water flow, thus balancing computational accuracy and efficiency.

[0107] The construction, discretization, and encryption are achieved by the two-dimensional hydrodynamic module through mesh generation, parameter configuration, and numerical calculation.

[0108] A hydraulic node is a simplified and generalized form of a hydraulic structure that maintains the hydraulic connection between upstream and downstream structures and provides regulation functions.

[0109] The weir flow formula is a classic calculation formula for simulating the water carrying capacity of dams, weirs, culverts, etc., and can accurately reflect the flow rate under different opening degrees and water levels.

[0110] Specifically, this can be achieved through structural generalization, formula invocation, flow calculation, and model coupling.

[0111] In this embodiment, by constructing a hybrid hydrodynamic model, rapid calculation of river flood discharge and detailed simulation of surface inundation can be achieved simultaneously, balancing computational efficiency and simulation accuracy. It can realistically reflect the complete physical processes such as river flood discharge, water level backwater, flood overflow, and water accumulation expansion in low-lying areas. Combined with the generalization of hydraulic structures, it can accurately simulate the impact of engineering scheduling on flood evolution, providing core hydrodynamic calculation support for the simulation of the entire flood process, risk analysis, and evaluation of scheduling effects.

[0112] In one embodiment, optionally, constructing the project scheduling model includes: A scheduling knowledge base is constructed based on flood control and drainage rules, gate and pump operation rules, and pipeline network scheduling rules; The scheduling logic was written as an executable script and embedded into the hydrodynamic model calculation process; We construct a deep reinforcement learning optimization structure with the goals of minimizing the flooded area, maximizing drainage efficiency, and minimizing energy consumption. Based on the real-time output of water level and flow status from the model, scheduling instructions are generated to obtain the engineering scheduling model.

[0113] The scheduling knowledge base is a structured database that stores scheduling thresholds, control logic, constraints, operational limitations, priority strategies, and safety boundaries. It is the core basis for intelligent scheduling. It can be implemented through rule organization, parameter processing, logical structuring, and knowledge base loading.

[0114] Executable scripts are programmed scheduling instructions that can be directly read, called, and run by a computer, supporting real-time calling, dynamic modification, and batch import.

[0115] The writing and embedding are achieved by the scheduling development unit through code implementation, interface docking, process integration and real-time invocation.

[0116] Deep reinforcement learning is an artificial intelligence method that continuously optimizes scheduling strategies based on real-time simulation feedback, and can achieve multi-objective collaborative optimization.

[0117] The optimization targets take into account flood control safety, drainage effectiveness, and operational economy, and are fully in line with the actual dispatching needs of the city.

[0118] Dispatch instructions include the degree of opening and closing of sluice gates and dams, the number of pump stations to be started and stopped, the duration of operation, and the linkage control strategy and timing arrangement.

[0119] In this embodiment, by constructing an engineering scheduling model that integrates a rule knowledge base and deep reinforcement learning, unified and collaborative scheduling of water conservancy projects and municipal pumping stations can be achieved, replacing the traditional experience-based decision-making model. Under the premise of ensuring flood control and drainage safety, the scheduling strategy is automatically optimized to achieve multi-objective balance, effectively improve the coordination of cross-departmental and cross-system scheduling, reduce inundation losses, improve drainage efficiency, and reduce operating energy consumption, providing core algorithm and model support for intelligent, precise, and collaborative scheduling decisions for urban flooding.

[0120] In one embodiment, optionally, the three-dimensional dynamic coupling to form a full-space dynamic model of water resources includes: Two-way water exchange between the surface and the pipe network is achieved through inspection wells and storm drains. When the surface water level is lower than the node water level, the runoff flows into the pipe network, and vice versa. Using the river water level at the outlet as the dynamic boundary, the rolling coupling calculation between the pipeline network and the river is performed according to the preset time step; The engineering scheduling model reads the calculation results of the hydrodynamic model in real time and outputs control commands to update the boundary conditions, forming a closed-loop coupling.

[0121] Two-way water exchange can simultaneously simulate surface runoff into pipes and overload overflow in the pipe network, realistically recreating the formation and development process of urban flooding.

[0122] The implementation process is achieved by the coupled interface unit through hydraulic balance calculation, water level comparison and judgment, flow distribution and interactive transmission.

[0123] Dynamic boundaries can reflect in real time the impact of river water level changes on the backflow and flooding of pipeline outlets, avoiding the huge errors caused by unidirectional coupling.

[0124] Rolling coupling computation ensures the consistency of the model in time and space, achieving precise connection.

[0125] It is achieved by the coupled computing module through boundary updates, variable passing, iterative computation, and time synchronization.

[0126] Closed-loop coupling refers to achieving dynamic circulation and truly reflecting the real-time impact of engineering regulation on the flood process.

[0127] In this embodiment, by realizing the full-process three-dimensional dynamic closed-loop coupling between the surface, pipeline network, river channel, and engineering scheduling, the complete physical process of urban flooding from generation, transmission, overflow to regulation and receding can be realistically reproduced. This effectively solves key technical problems such as unidirectional coupling, fragmented simulation, and inability to reflect backflow in traditional models, significantly improving the realism, coherence, and accuracy of the simulation. It provides a reliable integrated model foundation for dynamic simulation of the entire flood process, pre-simulation of scheduling schemes, risk assessment, and decision optimization.

[0128] In one embodiment, optionally, the step of using at least one historical flood event from historical flood data to calibrate and verify the accuracy of the water resources full-space dynamic model includes: At least two independent historical rainstorm and flood events were selected, and a hybrid optimization algorithm consisting of genetic algorithm and particle swarm optimization algorithm was used to calibrate the parameters of roughness coefficient, confluence coefficient and drag coefficient. Select at least one historical rainstorm and flood event that was not included in the rate determination for verification; The accuracy verification is passed when the river flood meets the preset maximum water level error and preset flow error, and the waterlogging meets the preset inundation range error and preset water depth error.

[0129] The hybrid optimization algorithm balances global optimization capability with local convergence speed, resulting in high calibration efficiency and stable and reliable results.

[0130] Parameter calibration ensures that the model output closely approximates the measured data, thereby improving the model fit.

[0131] Independent verification events can objectively test the model's generalization ability, stability, and reliability.

[0132] The verification process is achieved by comparing the error levels between simulated and measured values. Multi-dimensional error indicators ensure that the model meets engineering application requirements in both river flood control and urban waterlogging simulations. This is accomplished through error statistics, indicator comparison, and compliance assessment.

[0133] In this embodiment, model parameter calibration and independent accuracy verification are carried out based on multiple historical flood events. An efficient hybrid optimization algorithm is adopted to improve the parameter fitting effect. Strict verification standards are set from multiple dimensions such as river level, flow, flood volume, and waterlogging inundation range and water depth. This ensures that the model has high computational accuracy, good stability and strong generalization ability under different rainfall scenarios and different terrain conditions, ensuring that the model output results are reliable and usable, and meeting the requirements of actual engineering scheduling and emergency decision-making.

[0134] In one embodiment, optionally, a digital simulation platform is built using the aforementioned full-space dynamic model of water resources to perform flood simulation and determine scheduling schemes, including: Constructing a 3D visualization scenario of urban flooding based on digital twins; By accessing real-time rainfall monitoring data, hydrological monitoring data, and engineering operation monitoring data, the dynamic model of water resources in the whole space is driven to perform dynamic simulation. Automatically generate multi-scenario scheduling schemes and perform quantitative evaluations, and use the analytic hierarchy process to determine the optimal scheduling scheme; Output the optimal scheduling scheme.

[0135] Digital twin scenarios can faithfully recreate terrain, waterways, pipelines, engineering structures, buildings, roads, and other geographical features, providing an intuitive and visual representation of flood events. Scene construction is achieved by the platform's rendering module through 3D modeling, data loading, scene rendering, and effect optimization. Real-time data-driven processes ensure that simulation results are synchronized with real-world water conditions and operational realities, improving the timeliness of decision-making. This driving process is implemented by the data access module through real-time acquisition, transmission, analysis, and model input.

[0136] Multi-scenario solutions cover different rainfall, operating conditions, and dispatch strategies, providing comprehensive comparisons for decision-making. The Analytic Hierarchy Process (AHP) effectively avoids biases caused by subjective decision-making through objective weighting and scientific comparison. Generation, evaluation, and determination are achieved by the platform's decision-making module through scenario setting, simulation calculations, indicator evaluation, and optimization.

[0137] The output includes solution text, control instructions, visualization charts, risk levels, and early warning information. The output process is handled by the platform's output module through display, storage, push, and printing.

[0138] In this embodiment, a digital simulation platform is built based on a digital twin and a verified coupled model. This platform enables real-time dynamic simulation of flood processes, rapid pre-simulation of multi-scenario dispatching schemes, quantitative comparison and automatic output of optimal strategies. It also features 3D visualization, real-time data-driven operation, and intelligent decision-making capabilities. This provides flood control dispatchers with intuitive, efficient, and reliable decision support, significantly improving emergency response speed and the scientific nature of dispatching decisions, and achieving a complete implementation from model calculation to practical application.

[0139] To enable those skilled in the art to better understand this solution, this application also provides a preferred embodiment.

[0140] Step 1: Basic data collection and preparation; The system should clearly define the geographical scope, water system distribution, key water conservancy projects, municipal drainage projects, and key drainage areas (such as the core flood-prone area of ​​Guodiwa and the old city); it should systematically sort out the scheduling rules and operation logic of urban flood control, drainage and drainage pipe networks, clarify the core parameters such as the water level thresholds, linkage conditions, time constraints, and operation restrictions for the opening and closing of sluice gates and dams and the start and stop of pumping stations, and establish a list of scheduling rules.

[0141] The following diverse basic data were comprehensively collected: topographic and geomorphological data (DEM data with an accuracy of no less than 1:10000, with a focus on refining the topographic accuracy of the Guodiwa area, with a resolution of no less than 5m); land use and underlying surface type data (the latest data from 2024, clearly defining the distribution of impervious surfaces, green spaces, cultivated land, and building land, and calculating the proportion of impervious surfaces); long-term hydrological and meteorological data (rainfall, evaporation, temperature, wind speed, etc., from 1980 to the present, covering different rainfall levels, and using the Thiessen polygon method to calculate the urban area's areal rainfall); and water conservancy engineering parameters (dam dimensions, opening and closing methods). Data includes: reservoir capacity, design water level, flood control limit water level, etc.; municipal drainage network data (pipe diameter, pipe length, slope, manhole location, storm drain distribution, network operating years, etc.); pump station characteristic data (design flow rate, head, operating efficiency, start / stop threshold, energy consumption parameters, etc.); historical flood monitoring data; integrated "sky-ground-hydraulic" monitoring and sensing data (high-resolution remote sensing images, UAV monitoring data, underground pipe network detection data, real-time water level and flow monitoring data, etc.); and economic and social data (population distribution, urban built-up area, transportation network distribution, location of important public facilities, etc.).

[0142] A comprehensive quality inspection of the collected basic data is conducted, including checks on completeness, reliability, consistency, and timeliness. For missing data, an interpolation method based on an LSTM neural network is used to supplement it, ensuring data continuity. For abnormal data, the 3σ criterion is used for identification, and corrections are made by combining historical data from the same period and data from adjacent monitoring stations to remove invalid data. For data of different formats (vector, raster, text) and scales, standardized processing algorithms are used to achieve unified adaptation and establish data coding standards. At the same time, a multi-source data fusion governance platform is established, using domestic databases (such as DM Database) to achieve unified management, sharing, and real-time updates of cross-departmental, cross-level, and cross-business data, providing high-quality and highly reliable data support for subsequent model construction and ensuring data security and independent controllability.

[0143] Step 2: Construct a multi-module mathematical model; Based on the depression topography and complex water system characteristics of a certain urban area, a multi-module mathematical model covering "rainfall-surface runoff generation-pipeline drainage-river flood discharge-engineering regulation" is constructed. Each module is independent yet interconnected. The specific construction method is as follows: (1) Hydrological runoff generation and collection model: The improved Xin'anjiang model is adopted. Combined with the topographic features of the Guodiwa area in a certain urban area, the topographic slope correction coefficient (value 0.85-1.15, dynamically adjusted according to the slope of different areas of Guodiwa), the dynamic change factor of the permeability of the underlying surface (dynamically updated in combination with soil moisture content and rainfall intensity), and the correction term for the expansion of impermeable surface due to urbanization are innovatively introduced. The model fully considers the characteristics of concentrated runoff paths and difficult water receding in the Guodiwa area, as well as the impact of the expansion of impermeable surface on runoff generation and collection during the urbanization process, and accurately simulates the surface runoff generation and collection process under different rainfall intensities and different underlying surface conditions. The runoff generation calculation uses the Horton formula, taking into account the dynamic changes and spatial heterogeneity of the underlying soil moisture content, and corrects the soil infiltration parameters. The runoff collection calculation uses a nonlinear reservoir model combined with the Muskingen method, taking into account topographic slope and drainage path, to simulate the collection and transmission process of surface runoff, focusing on optimizing the runoff collection parameters in the core area of ​​the Guodiwa depression, and improving the accuracy of runoff generation and collection simulation under complex underlying surface conditions. At the same time, a rainfall spatial interpolation module is introduced, and the Thiessen polygon method is used to calculate the urban area's areal rainfall, solving the impact of uneven rainfall spatial distribution on runoff generation and collection simulation, and ensuring the accuracy of runoff generation calculation in different areas.

[0144] (2) One-dimensional unsteady flow hydraulic model of drainage network: using Saint-Venant equations or dynamic wave method, the pipe network segments are discretized by one-dimensional finite volume elements (the discretization length does not exceed 100m, and the discretization length of the core area of ​​Guodiwa does not exceed 80m) to simulate the water flow state, drainage capacity and exchange process with the surface of the pressure flow, open channel flow and transition flow in the pipe network. The model incorporates a dynamic correction model for pipeline resistance coefficients. Based on the pipeline's operational years, siltation level, and real-time monitoring data (such as pipeline flow and water level data), machine learning algorithms are used to adjust the resistance coefficients in real time, accurately reflecting the impact of pipeline blockage, aging, and siltation on drainage capacity. Considering the dense distribution and numerous nodes of the pipeline network in a certain urban area, CPU-GPU heterogeneous parallel computing technology is employed, combined with a pipeline node block parallel algorithm, dividing the pipeline network into multiple independent computing units to achieve parallel computing and significantly improve model computational efficiency. Simultaneously, the catchment areas of urban stormwater wells are divided into independent urban hydrological units. Each unit possesses independent hydrological and hydraulic attributes and consists of a regional hydrological model, a regional catchment unit, and stormwater wells. Rainfall is calculated and generated by the regional hydrological model, flows into the regional catchment unit, and then into surrounding pipelines or river networks through stormwater wells, achieving refined simulation of pipeline inflow and accurately capturing community-level flooding and backflow processes.

[0145] (3) River network hydrodynamic model: It is divided into one-dimensional and two-dimensional models, forming a hybrid simulation system of "one-dimensional backbone river network + two-dimensional key area". Among them, the backbone river network adopts a one-dimensional hydrodynamic model, which is constructed based on the Saint-Venant equations and solved using the Preissmann four-point implicit difference scheme to simulate the evolution of river floods, water level changes and flow transmission processes. It focuses on the flood characteristics of backbone rivers, changes in river cross sections and the influence of hydraulic structures. The river cross sections are preferentially based on measured large cross section data, and optimal generalization is carried out in combination with the model construction objectives and data feasibility to ensure a balance between simulation accuracy and computational efficiency. For key flood control protection areas and flood-prone low-lying areas (such as the core area of ​​the Guodiwa depression in urban areas), a two-dimensional shallow water equation is used to construct a hydrodynamic model. This model is discretized using two-dimensional finite volume elements (with an average discretization length not exceeding 10m, and not exceeding 8m in the core flood-prone areas) to accurately simulate flood overflow, inundation range, and water depth distribution. GPU-accelerated computation combined with adaptive mesh refinement technology is employed to automatically refine the mesh at flood overflow boundaries and in areas with drastic water depth changes, while simplifying the mesh in areas with stable water flow, achieving rapid and high-precision simulation of the flood evolution process. Hydraulic structures (culverts, sluices, pumping stations) are generalized as interconnected model elements, and their flow rates are simulated using hydrodynamic methods. Nodes are set upstream and downstream of the structures to connect with the generalized river channel. Water levels and flow rates between nodes are calculated using weir flow formulas and the structure operation modes. The operation modes are generalized using logical control conditions, supporting separate control modes for sluices and pumps as well as joint control modes, accurately simulating the hydraulic effects of coordinated sluice and pump operation.

[0146] (4) Engineering scheduling simulation module: Integrates the scheduling rules of water conservancy projects such as Nansi Lake and Yunlong Lake Reservoir and municipal drainage pumping stations, and constructs a standardized scheduling rule knowledge base. The water level threshold control, linkage logic, operation constraints, time constraints, and energy consumption constraints of dam opening and closing and pumping station start and stop are written into computer-recognizable Python scripts or rule sets, which are seamlessly embedded into the model calculation process. It supports flexible modification, expansion and batch import of scheduling rules to adapt to different scheduling scenarios. At the same time, deep reinforcement learning algorithms (such as DQN algorithm) are introduced to construct a multi-agent scheduling model. With the goal of "minimizing the flood range, maximizing drainage efficiency, ensuring urban safety and reducing engineering energy consumption", combined with the Pareto optimality criterion, the scheduling strategy is autonomously optimized and dynamically adjusted. The model can adaptively adjust the degree of dam opening and closing and the combination of pumping station operation according to real-time water and rainfall data and model simulation results. It supports the adaptive generation of scheduling schemes under multiple scenarios (different rainfall levels, engineering failures) to solve the problem of insufficient coordination in cross-water conservancy-municipal system scheduling.

[0147] Step 3: Achieve dynamic coupling of multiple models A self-assembly-driven multi-model coupling integration and intelligent invocation technology framework is adopted. Based on a domestic coupling interface protocol, key coupling interface modules are developed to achieve dynamic coupling between various models, forming a complete flood simulation model system. This ensures real-time data interaction and collaborative computation among modules. The specific coupling methods are as follows: (1) Coupling between surface and pipe network: Through nodes such as inspection wells and storm drains, a two-way dynamic exchange between surface runoff and pipe network inflow is realized. When the surface water level is lower than the water level of the pipe network node, the surface runoff flows into the pipe network through the node, and the inflow is calculated based on the node inflow capacity formula; when the surface water level is higher than the water level of the pipe network node or the pipe network drainage capacity is insufficient, the water flow in the pipe network overflows to the surface through the node, and the overflow flow is calculated based on the hydraulic balance equation to ensure the authenticity and accuracy of the coupling process and accurately simulate the surface waterlogging process caused by pipe network overflow.

[0148] (2) Coupling between the pipeline network and the river: Real-time dynamic connection between pipeline outflow and river water level is achieved at the pipeline outlet, with a focus on the impact of river water level on pipeline outlet backflow and backflow. The water level of the corresponding two-dimensional finite volume unit of the river is used as the boundary condition of the outlet, and the outlet flow rate is used as the source term of the river unit. A rolling coupling calculation method is adopted (coupling step length not exceeding 10 min) to realize the dynamic interaction between the pipeline network and the river flow. After each step of calculation is completed, the pipeline outflow rate is fed back to the river model, and the river water level is fed back to the pipeline model to update the boundary conditions, avoid simulation errors caused by unidirectional coupling, and accurately simulate problems such as poor pipeline drainage and backflow caused by river backflow.

[0149] (3) Coupling of engineering scheduling and hydrodynamic model: The engineering scheduling simulation module is seamlessly connected with the pipeline network and river network hydrodynamic model. A real-time data interaction mechanism is adopted. The engineering scheduling module obtains data such as water level, flow rate and inundation range calculated by the model in real time. Based on the embedded scheduling rule knowledge base and deep reinforcement learning optimization strategy, it outputs control commands for dam opening and closing and pump station start and stop (such as the degree of dam opening and closing and the number of pump stations in operation), which are fed back to the hydrodynamic model to adjust the model calculation parameters. This achieves closed-loop coupling of "simulation-scheduling-feedback-resimulation", which truly reflects the impact of engineering regulation on the flood process and ensures the feasibility and effectiveness of the scheduling plan.

[0150] Step 4: Model parameter calibration and validation Two to three typical historical rainstorm and flood events with complete monitoring records were selected in a certain urban area (covering different rainfall levels and flood scenarios, such as the rainstorm events in 2018 and 2021). A hybrid optimization method combining genetic algorithm and particle swarm optimization algorithm was adopted, and simulated annealing algorithm was used to improve the optimization efficiency. The model parameters (including roughness coefficient, confluence coefficient, pipeline resistance coefficient, pump station operating parameters, underlying surface permeability coefficient, terrain slope correction coefficient, etc.) were adjusted to make the simulated water level, flow rate, inundation range, water depth and other data consistent with the measured data, and the model parameters were calibrated.

[0151] Select 1-2 additional historical torrential rain and flood events that were not included in the calibration (such as the 2023 torrential rain event) to validate the calibrated model and evaluate its reliability and generalization ability. The validation criteria strictly adhere to the following requirements: (1) River floods: The difference between the calculated highest water level and the measured highest water level at the station within the river section shall not exceed 20 cm; the relative error between the measured and calculated maximum flow (the difference between the calculated flow and the measured flow / the measured flow) shall not exceed 10%; the relative error between the maximum 1-day, 3-day and 7-day flood volume (the difference between the calculated flood volume and the measured flood volume / the measured flood volume) shall not exceed 5%; the phase difference between the calculated water level process and the flow process and the measured water level process and the flow process shall not exceed 1 h.

[0152] (2) Urban flooding: The simulation error of the flooded area is no more than 15%; the simulation error of the water depth is no more than 10cm; the simulation error of the flood receding time is no more than 2h.

[0153] If the verification results do not meet the above criteria, return to step 4 to readjust the parameters until the model accuracy meets the standards. Simultaneously, establish a parameter sensitivity analysis module, using the Morris screening method to identify core parameters that significantly affect the simulation results, improving the relevance and efficiency of parameter calibration, and ensuring the stability and reliability of the model under different scenarios.

[0154] Step 5: Build a digital scheduling simulation and comparison platform A digital flood scenario for a certain urban area was constructed based on digital twin technology. A domestically developed visualization engine (such as SuperMap) was used for scene rendering, recreating geographical elements such as the topography of the Guodiwa depression, water system distribution, water conservancy and municipal engineering projects, and urban buildings. A digital scheduling, simulation, and comparison platform was built, integrating functions such as real-time monitoring data access, model calculation, result visualization, scheme comparison, decision output, and risk warning. The specific implementation is as follows: (1) Real-time data access: Access meteorological department rainfall forecast data (short-term, medium-term and long-term), hydrological department water situation monitoring data, water conservancy and municipal engineering operation data, "sky-ground-water engineering" integrated monitoring data, etc., adopt domestic data transmission protocol to realize real-time data update and synchronous transmission, and provide data support for dynamic model driving; at the same time, it has a data anomaly alarm function, and automatically issues an alarm prompt when the monitoring data exceeds the normal range.

[0155] (2) Dynamic simulation and visualization: The UNet-VIT water depth prediction network architecture is adopted and coupled with the model to realize the dynamic simulation and visualization of the flood process under different scenarios (different rainfall levels, engineering failures, scheduling schemes, extreme weather), intuitively showing the flood evolution, inundation range, water depth distribution and engineering operation status; it supports multi-view viewing (overhead, eye level, local zoom), and supports the playback of historical flood processes, which is convenient for staff to analyze flood patterns.

[0156] (3) Comparison and optimization of scheduling schemes: Multiple scheduling schemes are preset (such as different dam opening and closing sequences, pump station operation combinations, emergency response measures, etc.). The flood control effect, drainage efficiency, engineering energy consumption and economic and social impact (such as population and facility losses in the flooded area) of each scheme are quantitatively evaluated through model simulation. A multi-objective decision-making model combining the analytic hierarchy process and the entropy weight method is adopted to eliminate subjective weight bias and output the optimal scheduling scheme. The scheme supports custom editing and simulation of scheduling schemes, providing flexible decision support for staff.

[0157] (4) Offline Emergency Support and Risk Warning: The platform has offline operation capabilities, pre-stores basic data and model parameters, and can operate independently in extreme scenarios such as network interruption and power failure, providing all-weather flood control decision support; it also supports the export and printing of dispatch plans, providing technical guidance for on-site emergency response. The platform has real-time dynamic risk analysis capabilities. Based on simulation results combined with population, traffic, and important facility distribution data, it adopts a risk assessment model to automatically identify high-risk areas (such as densely populated areas and low-lying areas), and issues warning prompts in different levels (general, relatively severe, serious, and extremely serious). It supports the push of warning information through multiple channels such as SMS, APP, and audible and visual alarms, improving the efficiency of emergency response.

[0158] In addition, the platform fully meets domestic standards, with core algorithms, software, and databases all developed using domestic technologies, avoiding dependence on foreign technologies and improving system security, autonomy, and data security. The platform has good scalability and compatibility, and can add functional modules and expand the simulation range according to subsequent needs, adapting to the long-term development needs of smart water conservancy construction.

[0159] The innovations of this invention are mainly reflected in the following aspects: 1. By innovatively combining the unique topographic features of the Guodiwa depression, the hydrological runoff generation and confluence model of the Xin'anjiang River was optimized and improved. The topographic slope correction coefficient, the dynamic change factor of the permeability of the underlying surface, and the correction term for the expansion of impermeable surfaces due to urbanization were introduced. Combined with the Thiessen polygon method, the areal rainfall was accurately calculated. This solved the problem that the existing model did not fully consider the runoff generation and confluence characteristics of the Guodiwa depression, the uneven spatial distribution of rainfall, and the impact of urbanization. It significantly improved the runoff generation and confluence simulation accuracy in highly urbanized and complex topographic areas.

[0160] 2. A hybrid hydrodynamic simulation system of "one-dimensional backbone river network + two-dimensional key area" is constructed. By combining adaptive grid densification technology and CPU-GPU heterogeneous parallel computing, the high accuracy and efficiency of flood simulation are achieved. At the same time, urban catchment areas are divided into independent urban hydrological units to simulate the inflow of pipe networks and the process of urban flooding at the community level in a refined manner. This solves the problem that traditional models are difficult to balance between simulation accuracy and computational efficiency and lack detailed simulation.

[0161] 3. It achieves deep closed-loop coupling of the entire process from "surface-pipeline network-river channel-engineering", and solves the connection problem between pipeline network and surface, and pipeline network and river channel by adopting a two-way dynamic coupling method, focusing on the impact of river backflow and backflow; at the same time, it introduces deep reinforcement learning into the engineering scheduling module to build a multi-agent multi-objective scheduling model, realizes collaborative scheduling optimization across water conservancy and municipal systems, and makes up for the shortcomings of insufficient scheduling coordination and decision-making dependence on experience in existing technologies.

[0162] 4. Construct an integrated domestic technology system encompassing "data-model-platform-decision-making". Core algorithms, software, and databases are all developed using domestic technologies. Combined with multi-source data fusion governance and offline emergency support functions, the system's security, autonomy, and adaptability to extreme scenarios are enhanced. At the same time, it integrates real-time risk analysis and multi-channel early warning functions, enabling flood control to shift from "passive response" to "proactive early warning and precise scheduling".

[0163] 5. It innovatively introduces a parameter sensitivity analysis module and a multi-objective decision-making model (analytic hierarchy process + entropy weight method) to improve the pertinence of parameter calibration and the scientific nature of scheduling scheme comparison, avoid the influence of subjective factors, provide quantitative support for precise scheduling decisions, and has strong practicality and scalability.

[0164] Compared with the prior art, the present invention has the following advantages: 1. High simulation accuracy: Targeting the unique topography of the Guodiwa area and the complex water system of "one city, three regions" in a certain urban area, the hydrological runoff and hydrodynamic models were optimized. Multiple correction factors and refined simulation techniques were introduced to achieve bidirectional dynamic coupling throughout the entire process. Combined with a hybrid optimization algorithm for parameter calibration and verification, the simulation results showed a high degree of agreement with the measured data. It can accurately reflect the formation, collection, transmission and receding process of urban floods. In particular, it can accurately capture special phenomena such as concentrated runoff, difficult receding water, backflow of pipelines, and river backwater in the Guodiwa area, providing reliable technical support for scheduling decisions.

[0165] 2. Strong scheduling coordination: By integrating the engineering scheduling rules of water conservancy and municipal systems, a multi-agent, multi-objective scheduling model is constructed, realizing the coordinated scheduling simulation and intelligent optimization of various types of projects such as reservoirs, dams, drainage networks, and pumping stations. This solves the problems of disconnection and insufficient coordination between water conservancy and municipal systems in existing technologies, improves the scientificity and efficiency of multi-departmental collaborative scheduling decisions, and can achieve a balance of multiple objectives such as flood control, drainage, urban safety, and engineering energy consumption.

[0166] 3. High computational efficiency: Employing technologies such as CPU-GPU heterogeneous parallel computing, adaptive grid encryption, and node block parallelism, combined with deep learning-assisted optimization, the model's computational efficiency is significantly improved. It can realize real-time simulation of flood processes and rapid comparison of scheduling schemes (single-scenario simulation time does not exceed 30 minutes), meeting the requirements of dynamic and precise flood scheduling in highly urbanized areas; at the same time, it has offline operation capabilities to adapt to scheduling needs under extreme scenarios.

[0167] 4. Good compatibility with domestic technology: The core algorithms, software and database of the model are all developed using domestic technology, fully meeting the domestic standards, avoiding dependence on foreign technology, improving the system's security, autonomy and data security, and meeting the requirements of the national smart water conservancy construction for domestic development. It can effectively ensure the independent controllability of flood control technology.

[0168] 5. High practicality and promotional value: This invention is not only applicable to a certain urban area, but its core technology can be transferred to other highly urbanized areas with depressions and complex water systems (such as surrounding cities like Huaibei and Zaozhuang), providing standardized and replicable technical solutions for flood control in similar areas; the platform integrates functions such as visualization simulation, risk warning, and emergency support, and is easy to operate. It can be directly applied to daily scheduling and emergency response for urban flood control and drought relief, and has broad application prospects. It can significantly improve the intelligence level and overall effectiveness of urban flood defense in my country, and ensure the safe operation of cities.

[0169] 6. Comprehensive consideration of details: It fully considers details such as uneven spatial distribution of rainfall, siltation and aging of pipelines, gate pump control, and parameter sensitivity, and introduces a number of targeted technologies to make up for the lack of detailed simulation in existing models, further improving the practicality and accuracy of the model, and can better support actual scheduling work.

[0170] Example 2 like Figure 2 As shown, this application embodiment also provides a simulation and scheduling control system 200 for urban flooding processes, including a processor 201, a memory 202, and a program or instruction stored in the memory 202 that can run on the processor 201. When the program or instruction is executed by the processor 201, it implements the various processes of the above-described simulation and scheduling control method embodiment for urban flooding processes and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0171] Example 3 This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described urban flooding process simulation and scheduling control method embodiments and achieve the same technical effects. To avoid repetition, they will not be described again here.

[0172] The processor mentioned above is the processor in the urban flooding process simulation and scheduling control system described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0173] Example 4 This application also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above-described embodiments of the simulation and scheduling control method for urban flooding processes, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0174] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0175] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element. Furthermore, it should be noted that the scope of the methods and systems in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0176] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0177] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above, which are merely illustrative and not restrictive. Those skilled in the art, under the guidance of this application, can make many modifications without departing from the spirit and scope of the claims, all of which fall within the protection scope of this application.

[0178] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of this application, the scope of which is determined by the scope of the claims.

Claims

1. A method for simulating and controlling urban flooding processes, characterized in that, The method includes: Collect and construct a specific type of basic dataset, which includes at least one of the following: topographic data, land use data, meteorological data, water system engineering data, drainage network data, pumping station characteristic data, historical flood data, and integrated monitoring data; Based on the aforementioned basic dataset, a hydrological runoff generation and confluence model, a one-dimensional hydraulic model of drainage pipe network, a mixed hydrodynamic model of river network, and an engineering scheduling model are constructed respectively. The hydrological runoff model, the one-dimensional hydraulic model of the drainage network, and the mixed hydrodynamic model of the river network are dynamically coupled in three dimensions to form a dynamic model of water resources in the whole space. The parameters of the water resources full-space dynamic model are calibrated and the accuracy is verified using at least one historical flood event from historical flood data. After successful verification, a digital simulation platform is built using the aforementioned full-space dynamic model of water resources to simulate floods and determine scheduling schemes.

2. The method according to claim 1, characterized in that, The terrain data includes at least one of the following: DEM digital elevation model, slope, aspect, and depth and area of ​​depressions; The land use data includes at least one of the following: impervious surface ratio, green area, and water coverage ratio; The meteorological data includes at least one of hourly rainfall, hourly temperature, hourly wind speed, hourly humidity, and long-term annual rainfall statistics; The water system engineering data includes at least one of the following: river channel cross-sectional dimensions, riverbed roughness, dam location, opening threshold, and reservoir capacity curve. The drainage network data includes at least one of the following: pipe section length, diameter, material, slope, manhole coordinates, and storm drain grate distribution. The pump station characteristic data includes at least one of the following: single pump flow rate, head, power curve, start / stop threshold, and joint control logic; The historical flood data includes at least one of the following: historical rainstorm events, corresponding water levels, flow rates, inundation areas, and water depth time series. The integrated monitoring data includes at least one of radar rainfall, satellite rainfall, urban flooding video, and real-time water level station data.

3. The method according to claim 1, characterized in that, Constructing the hydrological runoff model includes the following sub-steps: The underlying surface type, soil moisture content, and terrain slope are determined based on the aforementioned basic dataset. An improved Xin'anjiang model was used to construct the runoff structure, and Horton's formula was used to calculate the surface runoff. In the runoff calculation, correction factors for impermeable surface expansion and terrain slope are introduced; A nonlinear reservoir model and the Muskingen method were used to construct the runoff structure and calculate the surface runoff process. Based on the distribution of rainfall stations, the spatial distribution of areal rainfall was calculated using the Thiessen polygon method, and the model was constructed.

4. The method according to claim 1, characterized in that, Constructing a one-dimensional hydraulic model of the drainage network includes: Establish pipe segment topology and node relationships based on the drainage network data; Establish one-dimensional unsteady flow control equations based on the Saint-Venant equations or the dynamic wave method; The one-dimensional finite volume method is used to spatially discretize the pipe network segments; A dynamic correction model for the resistance coefficient is introduced to reflect the impact of pipe section siltation and aging on flow capacity; The study area is divided into several hydrological units, with each hydrological unit corresponding to at least one inspection well or storm drain grate, to achieve refined calculation of the inflow into the pipe network and obtain a one-dimensional hydraulic model of the drainage pipe network.

5. The method according to claim 1, characterized in that, Constructing the hybrid hydrodynamic model of the river network includes: Based on the aforementioned water system engineering data, the scope of key waterways and flood-prone low-lying areas was extracted; A model of the main waterway is constructed using the one-dimensional Saint-Venant equations and solved using the Preissmann four-point implicit difference method. A two-dimensional shallow water equation model was constructed for flood-prone low-lying areas, and the model was discretized using the finite volume method and subjected to adaptive mesh refinement. By generalizing sluices, dams, and pumping stations as hydraulic nodes and using weir flow formulas to simulate the flow capacity of the projects, a mixed hydrodynamic model of the river network is obtained.

6. The method according to claim 1, characterized in that, Constructing the project scheduling model includes: A scheduling knowledge base is constructed based on flood control and drainage rules, gate and pump operation rules, and pipeline network scheduling rules; The scheduling logic was written as an executable script and embedded into the hydrodynamic model calculation process; We construct a deep reinforcement learning optimization structure with the goals of minimizing the flooded area, maximizing drainage efficiency, and minimizing energy consumption. Based on the real-time output of water level and flow status from the model, scheduling instructions are generated to obtain the engineering scheduling model.

7. The method according to claim 1, characterized in that, The three-dimensional dynamic coupling forms a full-space dynamic model of water resources, including: Two-way water exchange between the surface and the pipe network is achieved through inspection wells and storm drains. When the surface water level is lower than the node water level, the runoff flows into the pipe network, and vice versa. Using the river water level at the outlet as the dynamic boundary, the rolling coupling calculation between the pipeline network and the river is performed according to the preset time step; The engineering scheduling model reads the calculation results of the hydrodynamic model in real time and outputs control commands to update the boundary conditions, forming a closed-loop coupling.

8. The method according to claim 1, characterized in that, The step of using at least one historical flood event from historical flood data to calibrate and verify the accuracy of the full-space dynamic model of water resources includes: At least two independent historical rainstorm and flood events were selected, and a hybrid optimization algorithm consisting of genetic algorithm and particle swarm optimization algorithm was used to calibrate the parameters of roughness coefficient, confluence coefficient and drag coefficient. Select at least one historical rainstorm and flood event that was not included in the rate determination for verification; The accuracy verification is passed when the river flood meets the preset maximum water level error and preset flow error, and the waterlogging meets the preset inundation range error and preset water depth error.

9. The method according to claim 1, characterized in that, A digital simulation platform is built using the aforementioned full-space dynamic model of water resources to simulate floods and determine scheduling schemes, including: Constructing a 3D visualization scenario of urban flooding based on digital twins; By accessing real-time rainfall monitoring data, hydrological monitoring data, and engineering operation monitoring data, the dynamic model of water resources in the whole space is driven to perform dynamic simulation. Automatically generate multi-scenario scheduling schemes and perform quantitative evaluations, and use the analytic hierarchy process to determine the optimal scheduling scheme; Output the optimal scheduling scheme.

10. A simulation and control system for urban flooding processes, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the simulation and scheduling control method for urban flooding processes as described in any one of claims 1-9.

Images