Risk level assessment and early warning method for rural waterlogging areas based on multi-source data

Abstract

The present invention relates to the field of data assessment and processing technology, and in particular to a method for risk level assessment and early warning of rural waterlogged areas based on multi-source data. The method comprises the following steps: obtaining geographical information of rural waterlogged areas, hydrological and flood data, and flood control and drainage engineering data, and fusing the multi-source data to obtain multi-source data on rural floods; and calculating river flood overflows on the multi-source data on rural floods to obtain river flood overflow data. The present invention uses data processing technology, pattern recognition technology, and deep learning technology to fuse and analyze rural multi-source feature data to achieve risk level assessment of rural waterlogged areas; and constructs a real-time flood forecasting and early warning model to quickly predict the scope of rural flooding and plan risk avoidance routes when heavy rain occurs or the meteorological department issues a heavy rain warning.

Description

Technical Field

[0001] The present invention relates to the technical field of data evaluation and processing, and in particular to a method for risk level evaluation and early warning of rural waterlogged areas based on multi-source data. Background Art

[0002] Early rural waterlogging risk assessments relied primarily on empirical judgment and simple statistical analysis, lacking systematic theories and methods. With the advancement of science and technology, hydrological models and geographic information systems (GIS) have been gradually introduced, improving the scientific nature of assessments. With the continuous advancement of remote sensing, weather forecasting, and computer technology, rural waterlogging risk assessment technology has rapidly developed. For example, remote sensing technology is used to monitor rainfall and water level changes, combined with weather forecast models to predict short- and medium-term waterlogging conditions. Furthermore, the application of GIS technology enables risk assessment results to be more intuitively displayed on maps, facilitating intuitive analysis and decision-making by decision-makers. However, existing rural waterlogging risk assessment technologies fail to incorporate multi-source characteristic data from rural areas, resulting in low accuracy in assessing risk levels in rural waterlogging areas. Furthermore, rapid prediction of the extent of rural flooding is difficult when heavy rain occurs or when meteorological authorities issue heavy rain warnings. Summary of the Invention

[0003] Based on this, it is necessary to provide a method for assessing and warning the risk level of rural waterlogged areas based on multi-source data to solve at least one of the above technical problems.

[0004] To achieve the above objectives, a method for risk assessment and early warning of waterlogging areas in rural areas based on multi-source data is provided, the method comprising the following steps:

[0005] Step S1: Obtain rural waterlogged area geographic information, hydrological and flood data, and flood control and drainage engineering data, and perform multi-source data fusion to obtain rural flood multi-source data;

[0006] Step S2: Calculate river flood overflows on the rural flood multi-source data to obtain river flood overflow data; perform a two-dimensional hydrological and hydrodynamic quantitative assessment of waterlogging risk points on the rural flood multi-source data based on the river flood overflow data to obtain local waterlogging risk data; and construct a flood overflow and waterlogging risk model based on the river flood overflow data and the local waterlogging risk data.

[0007] Step S3: simulating the rural flood inundation process using a flood overtopping and waterlogging risk model to obtain flood inundation simulation results; dividing the flood inundation simulation results into watershed waterlogging areas to obtain watershed waterlogging area data; detecting rainstorm flood risk areas on the watershed waterlogging area data, and conducting a rural waterlogging risk level assessment to obtain waterlogging risk level data;

[0008] Step S4: Determine the rural flood risk warning index based on the waterlogging risk level data and generate a rural flood warning index system; construct a real-time flood forecast and warning model through the rural flood warning index system, and predict the rural flood risk area based on the real-time flood forecast and warning model, plan the risk avoidance and transfer route, and obtain a rural flood warning planning report.

[0009] The present invention obtains multi-source rural flood data by acquiring geographic information, hydrological and flood data, and flood control and drainage engineering data from rural waterlogging areas and performing multi-source data fusion. This multi-source data fusion method effectively integrates data from different sources, improving data integrity and accuracy. This provides a solid data foundation for subsequent flood overtopping calculations and waterlogging risk assessments, ensuring the reliability and effectiveness of model construction. River flood overtopping calculations are performed on the multi-source rural flood data to obtain river flood overtopping data. Based on the river flood overtopping data, a local two-dimensional hydrological and hydrodynamic quantitative assessment of waterlogging risk points is performed on the multi-source rural flood data to obtain local waterlogging risk data. This process accurately quantifies waterlogging risk points and provides detailed data support for the construction of flood overtopping and waterlogging risk models. This quantitative assessment allows for more accurate identification and assessment of waterlogging risk areas, improving the accuracy of risk warnings. Using the flood overtopping and waterlogging risk model, rural flood inundation processes are simulated to obtain flood inundation simulation results. The flood inundation simulation results are then divided into watershed waterlogging zones to obtain watershed waterlogging zone data. This simulation process dynamically displays the scope and process of flood inundation, providing a visual basis for detecting rainstorm flood risk areas and assessing rural waterlogging risk levels. By zoning watershed flood zones, risk areas can be more clearly defined, providing clear regional delineation for subsequent risk management and early warning. Rainstorm flood risk zones are detected using watershed flood zone data, and rural waterlogging risk levels are assessed to generate waterlogging risk level data. Based on this waterlogging risk level data, rural flood risk warning indicators are determined, generating a rural flood warning indicator system. This assessment and indicator system construction process systematically assesses waterlogging risk levels and provides a scientific basis for the construction of a real-time flood forecasting and early warning model. With a clear risk level classification and early warning indicator system, more effective flood risk early warning and management can be achieved. A real-time flood forecasting and early warning model is constructed using the rural flood warning indicator system. Based on this real-time flood forecasting and early warning model, rural flood risk areas are predicted and evacuation routes are planned, resulting in a rural flood warning planning report. This process enables real-time prediction of flood risk areas, providing scientific route planning for evacuation. Therefore, the present invention uses data processing technology, pattern recognition technology and deep learning technology to fuse and analyze rural multi-source characteristic data to realize the risk level assessment of rural waterlogged areas; construct a real-time flood forecast and warning model to quickly predict the scope of rural flooding and plan risk avoidance routes when heavy rain occurs or the meteorological department issues a heavy rain warning.

[0010] Preferably, step S1 includes the following steps:

[0011] Step S11: determining the rural flood area to be monitored;

[0012] Step S12: Collecting geographic information of the rural flood-prone areas to be monitored to obtain rural geographic information data;

[0013] Step S13: Perform hydrological and flood monitoring on the rural flood-prone area to be monitored to obtain hydrological and flood monitoring data;

[0014] Step S14: Obtaining flood control and drainage engineering data in the rural flood-prone area to be monitored, and performing structural digital conversion to generate flood control and drainage engineering data;

[0015] Step S15: pre-processing the rural geographic information data, hydrological and flood monitoring data, and flood drainage engineering data to obtain rural flood processing data, wherein the data pre-processing includes data cleaning, data format standardization, and feature extraction and conversion;

[0016] Step S16: performing data mapping and alignment on the rural flood processing data, and merging multi-source data to obtain rural flood multi-source data.

[0017] The present invention clarifies the monitoring scope by determining the rural flood area to be monitored, provides a clear target area for subsequent data collection and processing, and ensures the pertinence and effectiveness of data collection. Geographic information collection of the rural flood area to be monitored can accurately obtain the geographical features in the area, improving the geographical adaptability of model construction. Hydrological and flood monitoring of the rural flood area to be monitored can timely grasp the changes in water conditions in the area, provide real-time data support for flood overflow calculation and waterlogging risk assessment, and enhance the timeliness of risk warning. Flood control and drainage engineering data of the rural flood area to be monitored is obtained, and structural digital conversion is performed to convert the engineering data into a digital form, which is convenient for integration with geographic information and hydrological data, thereby improving the operability and analysis efficiency of the data. Data preprocessing of rural geographic information data, hydrological and flood monitoring data, and flood drainage engineering data can effectively remove noise and redundant information in the data, unify the data format, extract key features, and improve the quality and availability of the data. By mapping and aligning rural flood processing data and merging multi-source data, data from different sources can be organically integrated to form complete rural flood multi-source data, providing a comprehensive data foundation for subsequent flood inundation simulation and risk assessment, and ensuring the integrity and accuracy of model construction.

[0018] Preferably, step S2 includes the following steps:

[0019] Step S21: determining river hydrological zoning for the rural flood multi-source data, and performing river flood overflow calculation to obtain river flood overflow data;

[0020] Step S22: performing a two-dimensional hydrological and hydrodynamic quantitative assessment of local waterlogging risk points on the rural flood multi-source data based on the river flood overflow data to obtain local waterlogging risk data;

[0021] Step S23: constructing flood overflow and waterlogging risk based on the river flood overflow data and the local waterlogging risk data to obtain a flood overflow and waterlogging risk pre-model; training the flood overflow and waterlogging risk pre-model based on the river flood overflow data and the local waterlogging risk data to obtain a flood overflow and waterlogging risk training model;

[0022] Step S24: performing model cross-validation evaluation on the flood overflow and waterlogging risk training model to obtain training model evaluation data; adjusting model parameters of the flood overflow and waterlogging risk training model using the training model evaluation data to obtain a flood overflow and waterlogging risk model.

[0023] The present invention determines river hydrological zoning for multi-source rural flood data and performs river flood overflow calculations, which can accurately divide river hydrological areas, provide a clear zoning basis for flood overflow calculations, and improve the accuracy and reliability of flood overflow data. Based on river flood overflow data, a local two-dimensional hydrological and hydrodynamic quantitative assessment of waterlogging risk points is performed on multi-source rural flood data, which can accurately identify and quantify waterlogging risk points, provide detailed data support for subsequent risk model construction, and enhance the scientific nature and accuracy of risk assessment. Flood overflow and waterlogging risk are constructed based on river flood overflow data and local waterlogging risk data, and a flood overflow and waterlogging risk pre-model is trained based on river flood overflow data and local waterlogging risk data. This can effectively construct and train a flood overflow and waterlogging risk model, improve the adaptability and predictive ability of the model, and provide a reliable model foundation for flood inundation simulation and risk assessment. A model cross-validation evaluation was conducted on the flood overflow and urban waterlogging risk training model, and the model parameters of the flood overflow and urban waterlogging risk training model were adjusted based on the training model evaluation data. Through cross-validation and parameter adjustment, the model performance can be effectively evaluated and optimized, the stability and accuracy of the model are improved, and the reliability and effectiveness of the model in practical applications are ensured.

[0024] Preferably, step S21 includes the following steps:

[0025] Step S211: extracting topographic and geomorphic features from multi-source rural flood data to obtain rural topographic and geomorphic features; determining river system distribution based on the rural topographic and geomorphic features to generate river system distribution data;

[0026] Step S212: performing hydrological similarity identification on the river system distribution data to obtain hydrological similarity; partitioning the river system distribution data into river hydrologically similar regions based on the hydrological similarity to generate river hydrological partitioning data;

[0027] Step S213: Extracting mountain rainfall characteristics from the multi-source rural flood data to obtain mountain rainfall characteristics; determining reservoir locations based on river hydrological zoning data, and calculating reservoir water storage at the reservoir locations based on the mountain rainfall characteristics to generate reservoir rainfall-water storage relationship data;

[0028] Step S214: Calculate the river network flow direction, flow velocity, and water level change based on the river hydrological division data according to the reservoir rainfall-water storage relationship data to obtain river network hydrodynamic data;

[0029] Step S215: Calculating the surface water infiltration intensity, propagation parameters, and turbulent viscosity coefficient based on the river network hydrodynamic data and the rural topographic features to obtain surface water dynamic data;

[0030] Step S216: Calculate the forces acting on river floods and surface waterlogging based on the river network hydrodynamic data and the surface hydrodynamic data to obtain river flood overflow data.

[0031] The present invention extracts topographic features from multi-source rural flood data and determines river system distribution based on these features. This allows for accurate extraction of topographic features and determination of river system distribution, providing fundamental data support for subsequent hydrological zoning and flood overtopping calculations, and improving the geographic adaptability and data accuracy of model construction. Hydrological similarity identification is performed on river system distribution data, and based on this hydrological similarity, the river system distribution data is partitioned into similar hydrologically similar regions. Through hydrological similarity identification and partitioning, hydrologically similar regions can be effectively divided, providing a clear zoning basis for flood overtopping calculations and improving the accuracy and reliability of flood overtopping data. Rainfall features in mountainous areas are extracted from multi-source rural flood data, and reservoir locations are determined based on river hydrological zoning data. This allows for accurate extraction of rainfall features in mountainous areas and reservoir storage calculations, providing detailed rainfall and storage data support for subsequent river network hydrodynamic calculations and enhancing the model's hydrological adaptability and predictive capabilities. Based on reservoir rainfall-storage relationship data, river hydrological zoning data is used to calculate river network flow direction, flow velocity, and water level changes, accurately simulating the hydrodynamic changes in the river network. Based on river network hydrodynamic data, surface water infiltration intensity, propagation parameters, and turbulent viscosity coefficients are calculated based on rural topographic features, enabling accurate calculation of surface hydrodynamic parameters and enhancing the model's waterlogging risk assessment capabilities. Calculation of the forces acting on river floods and surface waterlogging based on river network and surface hydrodynamic data allows for accurate assessment of the forces acting on river floods and surface waterlogging, providing a scientific basis for the generation of flood overtopping data, improving its accuracy and reliability and laying a solid foundation for subsequent risk assessment and early warning.

[0032] Preferably, step S22 includes the following steps:

[0033] Step S221: Marking the flood overflow area of ​​the river flood overflow data to generate the flood overflow area; extracting the terrain elevation, slope, and distance from the river channel of the overflow area from the rural flood multi-source data based on the flood overflow area to obtain spatial feature data of the overflow area;

[0034] Step S222: quantifying the waterlogging risk index based on the spatial characteristic data of the overflow area, and calculating the waterlogging risk index based on the terrain elevation, slope, and distance from the river in the overflow area to obtain a waterlogging risk quantification index;

[0035] Step S223: measuring local two-dimensional hydrological and hydrodynamic parameters for the waterlogging risk quantification index, and calculating the water depth and flow velocity hydrodynamics of the local area to obtain local waterlogging hydrodynamic parameters;

[0036] Step S224: Evaluate the waterlogging risk change characteristics of local waterlogging hydrodynamic parameters to obtain local waterlogging risk data.

[0037] The present invention marks flood overflow areas based on river flood overtopping data and extracts the terrain elevation, slope, and distance from the river channel from multi-source rural flood data based on the flood overflow areas. This allows for accurate marking of flood overflow areas and extraction of relevant spatial feature data, providing detailed spatial information for subsequent waterlogging risk assessments and improving the geographic accuracy of risk assessments. Waterlogging risk indicators are quantified based on the spatial feature data of the overflow areas and, combined with the terrain elevation, slope, and distance from the river channel, the waterlogging risk indicators are calculated, converting the spatial feature data into specific waterlogging risk indicators. Local two-dimensional hydrological and hydrodynamic parameters are measured for the quantitative waterlogging risk indicators, and the water depth and flow velocity hydrodynamics of the local area are calculated. This allows for accurate determination of hydrodynamic parameters such as water depth and flow velocity in the local area, providing detailed data support for subsequent assessments of waterlogging risk change characteristics and improving the hydrodynamic simulation accuracy of waterlogging risk assessments. Waterlogging risk change characteristics are assessed based on local waterlogging hydrodynamic parameters, enabling dynamic monitoring of changing trends in waterlogging risk and enhancing the timeliness and accuracy of risk warnings.

[0038] Preferably, step S3 includes the following steps:

[0039] Step S31: setting initial conditions for the flood simulation process based on multi-source rural flood data, including setting initial rainfall intensity, initial water level, and topographic features, to generate simulation initial condition data;

[0040] Step S32: Based on the simulation initial condition data, the rural flood multi-source data is input into the flood overtopping and waterlogging risk model to simulate the rural flood inundation process and obtain the flood inundation simulation results;

[0041] Step S33: determining the flood depth based on the flood simulation result to obtain flood depth simulation data; identifying the flood area boundary based on the flood depth simulation data to obtain flood area boundary data;

[0042] Step S34: performing watershed mapping on the submerged area boundary data to obtain submerged watershed mapping data; performing waterlogging area division on the submerged watershed division data to obtain watershed waterlogging area data;

[0043] Step S35: Detect the rainstorm flood risk area on the watershed waterlogging area data, and evaluate the rural waterlogging risk level to obtain waterlogging risk level data.

[0044] The present invention sets initial conditions for flood simulations based on multi-source rural flood data, including initial rainfall intensity, initial water level, and topographic features. This allows for precise initial conditions for flood simulation, providing accurate starting parameters for subsequent simulations and improving the reliability and accuracy of simulation results. Based on the initial simulation conditions, the multi-source rural flood data is input into a flood overtopping and waterlogging risk model to simulate the rural flood process. This fully considers various influencing factors, improving the comprehensiveness and accuracy of flood simulations and providing detailed data support for subsequent risk assessments. The flood simulation results are then used to determine inundation depth, generating simulated inundation depth data. The inundation area boundaries are then identified based on the simulated inundation depth data, accurately determining the inundation depth and area boundaries. Watershed mapping of the inundation area boundary data allows for clear delineation of risk areas, providing clear regional divisions for subsequent risk management and early warning, and enhancing the geographic adaptability and management efficiency of risk assessments. Rainstorm flood risk areas are detected based on watershed flooding area data, and rural waterlogging risk levels are assessed, enabling dynamic monitoring and assessment of waterlogging risks.

[0045] Preferably, step S35 includes the following steps:

[0046] Step S351: performing multi-dimensional feature correlation analysis on the flood depth, flood range, and flood propagation speed of the watershed waterlogged area data to obtain multi-dimensional flood correlation data;

[0047] Step S352: locating the rainstorm range of the multi-dimensional flood-related data to obtain rainstorm range data; identifying the rainstorm risk area of ​​the multi-dimensional flood-related data based on the rainstorm range data to generate the rainstorm risk area;

[0048] Step S353: performing a dynamic simulation of the rainstorm and flood risk area, and analyzing the changes of the rainstorm and flood risk area at different time points to obtain dynamic change data of the risk area;

[0049] Step S354: Calculate the risk assessment index for the dynamic change data of the risk area to generate risk assessment index data; classify the risk assessment index data into waterlogging risk levels to obtain waterlogging risk level data.

[0050] The present invention performs multi-dimensional feature correlation analysis on the flood depth, flood range and flood propagation speed of the watershed waterlogging area data, and can comprehensively analyze the multi-dimensional features of flood events. The rainstorm range is located for the multi-dimensional flood-related data, and the rainstorm flood risk area is identified for the multi-dimensional flood-related data based on the rainstorm range data. The rainstorm flood risk area can be accurately defined, providing a clear regional division for subsequent risk assessment, and enhancing the geographical accuracy of the risk assessment. The risk area is dynamically simulated for the rainstorm flood risk area, and the changes in the rainstorm flood risk area at different time points are analyzed. The changing trend of the risk area can be monitored in real time, providing dynamic data support for subsequent risk assessment, and enhancing the timeliness and dynamism of the risk assessment. The risk assessment index is calculated for the dynamic change data of the risk area, and the waterlogging risk level can be scientifically assessed through the calculation of risk assessment indexes and the division of waterlogging risk levels.

[0051] Preferably, step S4 includes the following steps:

[0052] Step S41: Arrange the waterlogging risk level data in reverse order of the risk level gradient to obtain risk level gradient data; map low-, medium-, and high-risk rural waterlogging areas based on the risk level gradient data to obtain waterlogging risk mapping areas;

[0053] Step S42: Preliminary selection of flood risk warning indicators for the waterlogging risk mapping area to obtain preliminary selected warning indicator data; performing correlation analysis on the preliminary selected warning indicator data to obtain warning indicator correlation data; performing sensitivity analysis on the preliminary selected warning indicator data to obtain warning indicator sensitivity data;

[0054] Step S43: screening flood risk warning indicators from the preliminary warning indicator data based on the warning indicator correlation data and the warning indicator sensitivity data, and constructing a warning indicator system to generate a rural flood warning indicator system;

[0055] Step S44: constructing a flood warning model through the rural flood warning indicator system to generate a real-time flood forecast warning model;

[0056] Step S45: Predict rural flood risk areas based on the real-time flood forecast and warning model to obtain rural flood risk area prediction data; plan risk avoidance transfer routes based on the rural flood risk area prediction data to generate a rural flood warning planning report.

[0057] The present invention arranges the waterlogging risk level data in reverse order of the risk level gradient, and maps low-, medium-, and high-risk rural waterlogging areas based on the risk level gradient data. It can systematically arrange risk levels and perform regional mapping, providing clear regional divisions for the subsequent preliminary selection of warning indicators, thereby enhancing the geographical adaptability and management efficiency of risk warnings. Flood risk warning indicators are preliminarily selected for the waterlogging risk mapping areas, and correlation analysis is performed on the preliminarily selected warning indicator data. Through the preliminaries, correlation analysis, and sensitivity analysis, effective warning indicators can be scientifically screened out, providing scientific data support for the subsequent construction of a warning indicator system, thereby improving the scientific nature and reliability of the warning indicator system. Flood risk warning indicators are screened for the preliminarily selected warning indicator data based on the warning indicator correlation data and warning indicator sensitivity data, and a warning indicator system is constructed. Through screening and system construction, warning indicators can be effectively integrated to form a complete warning indicator system, providing a solid foundation for the subsequent construction of a flood warning model, thereby enhancing the scientific nature and operability of the warning model. By building a flood warning model based on the rural flood warning indicator system, a real-time forecast and warning model can be constructed based on the warning indicator system, improving the model's real-time performance and predictive capabilities. This provides scientific model support for subsequent risk area prediction and risk avoidance and transfer route planning, enhancing the timeliness and accuracy of risk warnings. Based on the real-time flood forecast and warning model, rural flood risk areas are predicted, and through risk area prediction and risk avoidance and transfer route planning, flood risk areas can be scientifically predicted and clear risk avoidance and transfer routes can be provided.

[0058] Preferably, step S44 includes the following steps:

[0059] Step S441: determining meteorological conditions on rural flood multi-source data to obtain meteorological condition performance data; performing flood warning demand analysis on the meteorological condition performance data to generate flood warning demand data;

[0060] Step S442: constructing a flood warning framework based on the rural flood warning indicator system to obtain flood warning framework information; calibrating the flood warning framework information based on the flood warning demand data to obtain a warning calibration framework;

[0061] Step S443: constructing a flood warning model for the warning calibration framework to generate an initial flood warning model; performing model warning verification on the initial flood warning model to obtain warning performance verification data; optimizing response parameters of the initial flood warning model based on the warning performance verification data to obtain a flood warning response model;

[0062] Step S444: Acquire real-time meteorological monitoring data and real-time water level monitoring data to obtain real-time monitoring data; input the real-time monitoring data into the flood warning response model, and conduct real-time flood forecast and warning training to generate a real-time flood forecast and warning model.

[0063] The present invention determines meteorological conditions based on multi-source rural flood data and analyzes flood warning requirements based on the meteorological condition performance data. This allows for accurate determination of meteorological conditions and analysis of flood warning requirements, providing clear guidance for the subsequent construction of a warning framework and improving the pertinence and adaptability of the warning system. A flood warning framework is constructed based on a rural flood warning indicator system and calibrated based on flood warning requirement data, ensuring that the warning framework aligns with actual needs and enhancing the framework's practicality and reliability. A flood warning model is constructed within the warning calibration framework to generate an initial flood warning model. The initial flood warning model is then validated to obtain warning performance verification data. Based on the performance verification data, the response parameters of the initial flood warning model are optimized to generate a flood warning response model. Through model construction and parameter optimization, the performance and responsiveness of the warning model can be improved. Real-time meteorological and water level monitoring data are acquired and input into the flood warning response model. Real-time flood forecast and warning training is then performed to generate a real-time flood forecast and warning model. Through the acquisition of real-time monitoring data and model training, real-time forecasting and early warning of flood events can be achieved, which improves the real-time and dynamic nature of the early warning system, provides scientific decision-making support for the management and early warning of rural flood disasters, and enhances the practicality and operability of risk warning.

[0064] Preferably, step S45 includes the following steps:

[0065] Step S451: Based on the real-time flood forecasting and warning model, the real-time monitoring data is used to predict rural flood scenarios to obtain rural flood scenario prediction data; the rural flood scenario prediction data is used to delineate flooding ranges to obtain flooding range delineation data;

[0066] Step S452: performing spatiotemporal analysis on the flooding range delineation data, analyzing the spatial distribution characteristics of each flooding range at different time points, and obtaining spatiotemporal prediction data of the flooding range;

[0067] Step S453: Determine the flood range change trend based on the spatiotemporal flood range prediction data to obtain flood range change trend information; perform multi-path planning of evacuation routes based on the flood range change trend information, plan multiple evacuation routes for each flood range, and obtain multiple evacuation route data;

[0068] Step S454: performing route feasibility assessment on the plurality of risk avoidance route data to obtain risk avoidance route feasibility data; determining an optimal risk avoidance route for the plurality of risk avoidance route data based on the risk avoidance route feasibility data to generate optimal risk avoidance route data;

[0069] Step S455: compile an early warning planning report based on the optimal risk avoidance route data to obtain a rural flood early warning planning report.

[0070] Based on a real-time flood forecasting and early warning model, the present invention predicts rural flood scenarios based on real-time monitoring data and delineates flooding areas based on the predicted data. This allows for real-time prediction of rural flood scenarios and precise delineation of flooding areas. The delineated flooding area data undergoes spatiotemporal analysis to analyze the spatial distribution characteristics of each flooded area at different time points, enabling dynamic monitoring of flooding area trends. This provides scientific spatiotemporal data support for subsequent evacuation route planning, enhancing the dynamic and adaptable nature of early warnings. The spatiotemporal flooding area prediction data is used to determine flooding area trends, allowing for the planning of multiple evacuation routes for each flooded area. Through trend determination and multi-path planning, evacuation routes can be scientifically planned, providing multiple route options for subsequent route feasibility assessments and improving the flexibility and reliability of evacuation. Route feasibility assessments are performed on multiple evacuation route data, and based on the evacuation route feasibility data, the optimal evacuation route is determined, enabling the selection of the most feasible evacuation route. The optimal evacuation route data is used to compile an early warning planning report, resulting in a rural flood early warning planning report. By compiling early warning planning reports, we can systematically integrate the data on optimal risk avoidance routes, provide comprehensive planning support for the management and early warning of rural flood disasters, and enhance the practicality and operability of risk warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 A flowchart of a method for risk assessment and early warning of rural waterlogged areas based on multi-source data;

[0072] Figure 2 for Figure 1 Detailed implementation steps of step S2 in FIG.

[0073] Figure 3 for Figure 1 Detailed implementation steps of step S4 in FIG.

[0074] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0075] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0076] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0077] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0078] To achieve this, please refer to Figures 1 to 3 A method for assessing and warning the risk level of waterlogged areas in rural areas based on multi-source data, comprising the following steps:

[0079] Step S1: Obtain rural waterlogged area geographic information, hydrological and flood data, and flood control and drainage engineering data, and perform multi-source data fusion to obtain rural flood multi-source data;

[0080] Step S2: Calculate river flood overflows on the rural flood multi-source data to obtain river flood overflow data; perform a two-dimensional hydrological and hydrodynamic quantitative assessment of waterlogging risk points on the rural flood multi-source data based on the river flood overflow data to obtain local waterlogging risk data; and construct a flood overflow and waterlogging risk model based on the river flood overflow data and the local waterlogging risk data.

[0081] Step S3: simulating the rural flood inundation process using a flood overtopping and waterlogging risk model to obtain flood inundation simulation results; dividing the flood inundation simulation results into watershed waterlogging areas to obtain watershed waterlogging area data; detecting rainstorm flood risk areas on the watershed waterlogging area data, and conducting a rural waterlogging risk level assessment to obtain waterlogging risk level data;

[0082] Step S4: Determine the rural flood risk warning index based on the waterlogging risk level data and generate a rural flood warning index system; construct a real-time flood forecast and warning model through the rural flood warning index system, and predict the rural flood risk area based on the real-time flood forecast and warning model, plan the risk avoidance and transfer route, and obtain a rural flood warning planning report.

[0083] The present invention obtains multi-source rural flood data by acquiring geographic information, hydrological and flood data, and flood control and drainage engineering data from rural waterlogging areas and performing multi-source data fusion. This multi-source data fusion method effectively integrates data from different sources, improving data integrity and accuracy. This provides a solid data foundation for subsequent flood overtopping calculations and waterlogging risk assessments, ensuring the reliability and effectiveness of model construction. River flood overtopping calculations are performed on the multi-source rural flood data to obtain river flood overtopping data. Based on the river flood overtopping data, a local two-dimensional hydrological and hydrodynamic quantitative assessment of waterlogging risk points is performed on the multi-source rural flood data to obtain local waterlogging risk data. This process accurately quantifies waterlogging risk points and provides detailed data support for the construction of flood overtopping and waterlogging risk models. This quantitative assessment allows for more accurate identification and assessment of waterlogging risk areas, improving the accuracy of risk warnings. Using the flood overtopping and waterlogging risk model, rural flood inundation processes are simulated to obtain flood inundation simulation results. The flood inundation simulation results are then divided into watershed waterlogging zones to obtain watershed waterlogging zone data. This simulation process dynamically displays the scope and process of flood inundation, providing a visual basis for detecting rainstorm flood risk areas and assessing rural waterlogging risk levels. By zoning watershed flood zones, risk areas can be more clearly defined, providing clear regional delineation for subsequent risk management and early warning. Rainstorm flood risk zones are detected using watershed flood zone data, and rural waterlogging risk levels are assessed to generate waterlogging risk level data. Based on this waterlogging risk level data, rural flood risk warning indicators are determined, generating a rural flood warning indicator system. This assessment and indicator system construction process systematically assesses waterlogging risk levels and provides a scientific basis for the construction of a real-time flood forecasting and early warning model. With a clear risk level classification and early warning indicator system, more effective flood risk early warning and management can be achieved. A real-time flood forecasting and early warning model is constructed using the rural flood warning indicator system. Based on this real-time flood forecasting and early warning model, rural flood risk areas are predicted and evacuation routes are planned, resulting in a rural flood warning planning report. This process enables real-time prediction of flood risk areas, providing scientific route planning for evacuation. Therefore, the present invention uses data processing technology, pattern recognition technology and deep learning technology to fuse and analyze rural multi-source characteristic data to realize the risk level assessment of rural waterlogged areas; construct a real-time flood forecast and warning model to quickly predict the scope of rural flooding and plan risk avoidance routes when heavy rain occurs or the meteorological department issues a heavy rain warning.

[0084] In the embodiment of the present invention, reference Figure 1 FIG. 1 is a flow chart of the steps of a method for assessing and warning the risk level of waterlogged areas in rural areas based on multi-source data according to the present invention. In this example, the method for assessing and warning the risk level of waterlogged areas in rural areas based on multi-source data includes the following steps:

[0085] Step S1: Obtain rural waterlogged area geographic information, hydrological and flood data, and flood control and drainage engineering data, and perform multi-source data fusion to obtain rural flood multi-source data;

[0086] In this embodiment of the present invention, geographic information system (GIS) technology is used to obtain topographic data, including elevation, slope, and land use type, for rural waterlogged areas from the National Tibetan Plateau Science Data Center. ArcGIS software is used to preprocess the acquired geographic information data, including data format conversion, coordinate system integration, and data clipping, to ensure data accuracy and consistency. Meteorological data, such as rainfall and evaporation, for rural waterlogged areas are obtained from a regional ground meteorological element-driven dataset (1979-2018). Real-time hydrological data, including river levels and flows, are obtained from data sources such as the hydrological network. Python is used for data cleaning, transformation, and analysis to remove outliers and missing values ​​and unify data at different time scales. Existing flood control (drainage) standards and corresponding peak flow and peak water level data are obtained from the Guidelines for the Preparation of Flood Control Evaluation Reports for Construction Projects within River Management Areas (Trial). Basic information, including the location, scale, and design standards of water conservancy (flood control) projects, such as rivers, embankments, reservoirs, culverts, and pumping stations, is obtained from reports on existing water conservancy projects and other facilities. Use Excel or database software to organize and classify flood control and drainage project data, generating structured data tables for subsequent analysis and integration. Employ multi-source data fusion techniques, such as Bayesian network-based multi-source information fusion methods, to fuse acquired geographic information, hydrological and flood data, and flood control and drainage project data. First, standardize data from different sources and convert them into a unified format and dimension. Then, using a Bayesian network model, fuse the data based on correlations and dependencies between the data to generate multi-source rural flood data. This fused rural flood data includes comprehensive information from geographic information, hydrological and flood data, and flood control and drainage project data. It can comprehensively reflect the flood risk characteristics of rural waterlogged areas and provide data support for subsequent risk level assessment and early warning.

[0087] Step S2: Calculate river flood overflows on the rural flood multi-source data to obtain river flood overflow data; perform a two-dimensional hydrological and hydrodynamic quantitative assessment of waterlogging risk points on the rural flood multi-source data based on the river flood overflow data to obtain local waterlogging risk data; and construct a flood overflow and waterlogging risk model based on the river flood overflow data and the local waterlogging risk data.

[0088] In this embodiment of the present invention, multi-source data fusion technology is used to integrate geographic information, hydrological and flood data, and flood control and drainage engineering data to generate multi-source rural flood data. A flood overflow calculation model based on the water level-discharge relationship is used to calculate river flood overflow range and depth based on river water level and flow data, combined with river cross-sectional morphology and flood control engineering parameters. Parameter settings: The flood overflow calculation time step is set to 10 minutes, the spatial resolution is 10 meters by 10 meters, and flood overflow conditions under different recurrence periods (e.g., 10, 20, and 50 years) are considered. Output: Generates river flood overflow data, including information such as overflow range and water depth distribution, providing basic data for subsequent urban flood risk assessment. Based on the river flood overflow data, combined with geographic information from the multi-source rural flood data and flood control and drainage engineering data, a local two-dimensional hydrological and hydrodynamic quantitative assessment of urban flood risk points is conducted. Using a two-dimensional hydrological and hydrodynamic model, such as SWMM (Storm Water Management Model) or MIKE21, hydrological and hydrodynamic parameters such as water depth and flow velocity at flood risk points are simulated. Parameter settings: The model time step is set to 5 minutes, and the spatial resolution is 5 meters by 5 meters. The model considers flood risk under different rainfall intensities (e.g., 10 mm / hour, 20 mm / hour) and drainage system operating conditions (e.g., normal operation, partial blockage). Output: Localized flood risk data, including information such as water depth and flow velocity distribution at flood risk points, is generated, providing data support for the construction of a flood overfall and flood risk model. River flood overfall data and localized flood risk data are fused, and multi-source data fusion techniques such as Bayesian networks are used to establish a correlation between flood overfall and flood risk. Based on this fused data, a flood overfall and flood risk model is constructed. This model considers factors such as flood overfall range, water depth, and flow velocity, as well as water depth and flow velocity at flood risk points, to comprehensively assess flood risk in rural waterlogged areas. Parameter Optimization: Model parameters are optimized using historical flood event data, and grid search and other methods are used to determine the optimal parameter combination for the model, thereby improving the model's prediction accuracy. Model Validation: The constructed model is validated using independent flood event data to assess its accuracy and reliability, ensuring that the model can effectively reflect the flood risk characteristics of rural waterlogged areas.

[0089] Step S3: simulating the rural flood inundation process using a flood overtopping and waterlogging risk model to obtain flood inundation simulation results; dividing the flood inundation simulation results into watershed waterlogging areas to obtain watershed waterlogging area data; detecting rainstorm flood risk areas on the watershed waterlogging area data, and conducting a rural waterlogging risk level assessment to obtain waterlogging risk level data;

[0090] In this embodiment of the present invention, a constructed flood overtopping and waterlogging risk model is combined with multi-source rural flood data, including geographic information, hydrological and flood data, and flood control and drainage engineering data. GIS-based flood inundation simulation technology, combined with a one-dimensional hydraulic model (such as HecRAS) and a two-dimensional hydrological and hydrodynamic model (such as SWMM), is used to simulate rural flood inundation processes. Parameter settings include a 5-minute simulation time step and a spatial resolution of 10 meters by 10 meters. Flood inundation scenarios are considered under different recurrence periods (e.g., 10, 20, and 50 years) and rainfall intensities (e.g., 10 mm / hour and 20 mm / hour). Output: Flood inundation simulation results are generated, including information such as inundation range, water depth distribution, and flow velocity distribution, providing basic data for subsequent watershed waterlogging zoning. Watershed waterlogging zoning is performed based on the flood inundation simulation results, combined with information such as topography and land use types from geographic information data. Delineation Method: Utilizing the DEM-based overland flow accumulation method (D8 algorithm), based on the gravitational effects of surface runoff under topography, the flow direction of each DEM grid is calculated to determine the cumulative runoff of its downstream grid. This method then extracts the river network and identifies the catchment area. Parameter Settings: The minimum cumulative runoff threshold for catchment area delineation is set at 1000 grids to ensure the rationality and accuracy of the delineation results. Output: Waterlogged areas within the watershed are generated, including their boundaries, area, and average water depth, providing data support for subsequent detection of flood risk areas. Based on this waterlogged area data, combined with water depth and flow velocity information from flood inundation simulation results, flood risk areas are detected and the waterlogging risk level is assessed. The Normalized Difference Water Index (NDWI) is used to extract flood inundation areas from satellite remote sensing imagery. Combined with historical flood event data, flood risk areas are then identified. A waterlogging risk assessment technique is used, taking into account waterlogging depth, flow velocity, and hazard parameters. The core calculation formula is RH = d × (v + 0.5) + df, where RH is the risk index, d is the waterlogging depth, v is the flow velocity, and df is the water depth hazard parameter. Parameter settings: df = 0.5 when the waterlogging depth d ≤ 0.2 m; df = 1.0 when d > 0.2 m. Based on the risk index RH, waterlogging risk levels are categorized into low (0-10), medium (10-20), and high (20-30). Output: Waterlogging risk data is generated, including the waterlogging risk level for each flooded area and the distribution of risk regions, providing a scientific basis for risk assessment and early warning in rural waterlogged areas.

[0091] Step S4: Determine the rural flood risk warning index based on the waterlogging risk level data and generate a rural flood warning index system; construct a real-time flood forecast and warning model through the rural flood warning index system, and predict the rural flood risk area based on the real-time flood forecast and warning model, plan the risk avoidance and transfer route, and obtain a rural flood warning planning report.

[0092] In an embodiment of the present invention, waterlogging risk level data is used in combination with geographic information, hydrological and flood data, and flood control and drainage engineering data from multi-source rural flood data. Based on the waterlogging risk level data, rural flood risk warning indicators are determined, including water depth, flow rate, flooding time, rainfall intensity, etc. Specifically, water depth: water depth less than 15 cm is a slight waterlogging point, and greater than or equal to 15 cm is a waterlogging point. Flow rate: areas with a flow rate exceeding 0.5 m / s are considered high-risk areas. Flooding time: areas with a flooding time exceeding 24 hours are considered high-risk areas. Rainfall intensity: areas with a rainfall intensity exceeding 20 mm / hour are considered high-risk areas. Result output: A rural flood warning indicator system is generated, including the above-mentioned indicators and their corresponding thresholds. Based on the rural flood warning indicator system, real-time meteorological data, hydrological data, and geographic information data are combined. A real-time flood forecasting and warning model based on neural networks, such as BP neural networks, convolutional neural networks (CNNs), and recurrent neural networks (RNNs), is employed and optimized using the stochastic gradient descent (SGD) algorithm, adaptive learning rate optimization methods (such as Adam), L2 regularization, and dropout techniques. Parameter settings include a 5-minute time step and a 10-meter x 10-meter spatial resolution. Flood scenarios with different return periods (e.g., 10, 20, and 50 years) and rainfall intensities (e.g., 10 mm / hour and 20 mm / hour) are considered. This model generates a real-time flood forecasting and warning model capable of predicting inundation extent, water depth, and flow velocity in rural flood-risk areas. The model integrates real-time meteorological, hydrological, and geographic data to predict rural flood risk areas, including inundation extent, water depth, and flow velocity at different future time points. The resulting rural flood risk prediction results include information such as the predicted time point, inundation extent, water depth distribution, and flow velocity distribution, providing data support for evacuation route planning. Based on the results of rural flood risk zone predictions and incorporating geographic information such as road networks and settlement distribution, the GIS Network Analysis function was used to analyze and determine the two optimal flood evacuation routes: the shortest time and the shortest distance, based on the road network data within the study area. Principles for the division of evacuation units were established, such as evacuation units in flood storage and detention areas being no larger than a natural village, while evacuation units in protection and protection areas could be up to the township level. A evacuation route planning map was generated, including information such as evacuation units, routes, and resettlement sites, providing specific content for the rural flood early warning planning report. The rural flood risk early warning indicator system, the real-time flood forecast and warning model, the rural flood risk zone prediction results, and the evacuation route planning results were integrated. A rural flood early warning planning report was compiled, including the warning indicator system, model construction methods, risk zone prediction results, and evacuation route planning, providing a scientific basis for early warning and response to rural flood disasters.

[0093] Preferably, step S1 includes the following steps:

[0094] Step S11: determining the rural flood area to be monitored;

[0095] Step S12: Collecting geographic information of the rural flood-prone areas to be monitored to obtain rural geographic information data;

[0096] Step S13: Perform hydrological and flood monitoring on the rural flood-prone area to be monitored to obtain hydrological and flood monitoring data;

[0097] Step S14: Obtaining flood control and drainage engineering data in the rural flood-prone area to be monitored, and performing structural digital conversion to generate flood control and drainage engineering data;

[0098] Step S15: pre-processing the rural geographic information data, hydrological and flood monitoring data, and flood drainage engineering data to obtain rural flood processing data, wherein the data pre-processing includes data cleaning, data format standardization, and feature extraction and conversion;

[0099] Step S16: performing data mapping and alignment on the rural flood processing data, and merging multi-source data to obtain rural flood multi-source data.

[0100] In an embodiment of the present invention, geographic information system (GIS) technology is used to combine historical flood data and topographic information to identify rural flood-prone areas for monitoring. Specifically, this involves using GIS software (such as ArcGIS) to load historical flood event data and topographic data, and using spatial analysis to identify flood-prone areas. Unmanned aerial vehicle (UAV) surveying technology, equipped with sensors such as high-definition cameras and lidar, is used to conduct aerial photogrammetry and remote sensing data acquisition of the identified rural flood-prone areas. The collected data, including orthophotos and oblique photogrammetric 3D models, is used for subsequent geographic information analysis. An IoT-based hydrological monitoring system utilizes a sensor network to monitor hydrological data such as water level, flow, and rainfall in the monitored areas in real time. Specifically, this involves deploying water level sensors, rainfall sensors, and flow velocity sensors, and transmitting the data via a wireless network to a monitoring center for real-time analysis. Flood control and drainage project data is collected in the monitored areas, including the location, scale, and design standards of embankments, reservoirs, pumping stations, and other projects. This data is digitized using GIS software to generate structured flood control and drainage project data for subsequent analysis and application. Data cleaning tools (such as the Pandas library) are used to clean the collected rural geographic information data, hydrological and flood monitoring data, and flood and drainage engineering data to remove duplicate data and outliers. Next, the data is standardized to unify the data format and dimension. Finally, feature extraction and transformation are used to extract features related to flood risk, such as water depth, flow velocity, and rainfall intensity. Utilizing GIS multi-source data fusion technology, the pre-processed rural geographic information data, hydrological and flood monitoring data, and flood and drainage engineering data are mapped and aligned. This involves unifying the spatial coordinates of the different data sources and then merging the multi-source data using data merging tools (such as the Merge tool in ArcGIS) to generate multi-source rural flood data, providing foundational data for subsequent risk assessment and early warning.

[0101] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0102] Step S21: determining river hydrological zones for the rural flood multi-source data, and performing river flood overflow calculations to obtain river flood overflow data;

[0103] Step S22: performing a two-dimensional hydrological and hydrodynamic quantitative assessment of local waterlogging risk points on the rural flood multi-source data based on the river flood overflow data to obtain local waterlogging risk data;

[0104] Step S23: constructing flood overflow and waterlogging risk based on the river flood overflow data and the local waterlogging risk data to obtain a flood overflow and waterlogging risk pre-model; training the flood overflow and waterlogging risk pre-model based on the river flood overflow data and the local waterlogging risk data to obtain a flood overflow and waterlogging risk training model;

[0105] Step S24: performing model cross-validation evaluation on the flood overflow and waterlogging risk training model to obtain training model evaluation data; adjusting model parameters of the flood overflow and waterlogging risk training model using the training model evaluation data to obtain a flood overflow and waterlogging risk model.

[0106] In this embodiment of the present invention, multi-source rural flood data, including geographic information data, hydrological and flood monitoring data, and flood control and drainage engineering data, is used. A clustering zoning method based on hydrological similarity theory is used, combining watershed characteristics and hydrological variables, to zoning the study area. Flood frequency distribution curve parameters, runoff characteristics, soil characteristics, and other factors are selected as zoning indicators. Fuzzy clustering is used for zoning. Consistency and uniformity tests are performed to generate a river hydrological zoning map, clearly defining the boundaries and hydrological characteristics of each zone. The MIKE 11 model is used to calculate river flood overtopping, combined with the MIKE 21 model for two-dimensional flood propagation simulation. The model time step is set to 5 minutes, and the spatial resolution is 10 meters by 10 meters. Flood overtopping data is generated for different recurrence periods (e.g., 10, 20, and 50 years) and rainfall intensities (e.g., 10 mm / hour and 20 mm / hour), including information such as overtopping range, water depth distribution, and flow velocity distribution. Based on river flood overtopping data, combined with geographic information from multi-source rural flood data and data from flood control and drainage projects, the SWMM model was used to quantitatively assess local two-dimensional hydrological and hydrodynamic parameters at waterlogging risk points, simulating hydrological and hydrodynamic parameters such as waterlogging depth and flow velocity at waterlogging risk points. The model was set with a 5-minute time step and a 5×5-meter spatial resolution. The model considered waterlogging risks under different rainfall intensities (e.g., 10 mm / hour, 20 mm / hour) and drainage system operating conditions (e.g., normal operation, partial blockage), generating local waterlogging risk data, including waterlogging depth and flow velocity distribution at waterlogging risk points. The river flood overtopping data and local waterlogging risk data were fused, and multi-source data fusion techniques such as Bayesian networks were used to establish a correlation between flood overtopping and waterlogging risk. Based on this fused data, a pre-model for flood overtopping and waterlogging risk was constructed. This model considers factors such as flood overtopping range, water depth, and flow velocity, as well as waterlogging depth and flow velocity at waterlogging risk points, to comprehensively assess flood risk in rural waterlogged areas. The pre-model was trained using historical flood event data to ensure it effectively reflects the flood risk characteristics of rural waterlogged areas. Stochastic gradient descent (SGD), adaptive learning rate optimization methods (such as Adam), L2 regularization, and dropout techniques were used for model optimization. This resulted in a flood overtopping and waterlogging risk training model that accurately predicts flood risk in rural waterlogged areas. The flood overtopping and waterlogging risk training model was evaluated using cross-validation metrics including mean squared error (MSE), root mean squared error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and coefficient of determination (R²). Training model evaluation data, including the values ​​of these metrics, was generated to assess the model's accuracy and generalization ability. Based on the training model evaluation data, model parameters were adjusted to optimize model performance.Specific adjustment methods include adjusting the learning rate, increasing the number of iterations, adjusting the regularization parameters, etc., to obtain an optimized flood overtopping and waterlogging risk model. The model shows good prediction performance on the training set, cross-validation set and test set.

[0107] Preferably, step S21 includes the following steps:

[0108] Step S211: extracting topographic and geomorphic features from multi-source rural flood data to obtain rural topographic and geomorphic features; determining river system distribution based on the rural topographic and geomorphic features to generate river system distribution data;

[0109] Step S212: performing hydrological similarity identification on the river system distribution data to obtain hydrological similarity; partitioning the river system distribution data into river hydrologically similar regions based on the hydrological similarity to generate river hydrological partitioning data;

[0110] Step S213: Extracting mountain rainfall characteristics from the multi-source rural flood data to obtain mountain rainfall characteristics; determining reservoir locations based on river hydrological zoning data, and calculating reservoir water storage at the reservoir locations based on the mountain rainfall characteristics to generate reservoir rainfall-water storage relationship data;

[0111] Step S214: Calculate the river network flow direction, flow velocity, and water level change based on the river hydrological division data according to the reservoir rainfall-water storage relationship data to obtain river network hydrodynamic data;

[0112] Step S215: Calculating the surface water infiltration intensity, propagation parameters, and turbulent viscosity coefficient based on the river network hydrodynamic data and the rural topographic features to obtain surface water dynamic data;

[0113] Step S216: Calculate the forces acting on river floods and surface waterlogging based on the river network hydrodynamic data and the surface hydrodynamic data to obtain river flood overflow data.

[0114] In this embodiment of the present invention, multi-source rural flood data, including geographic information data, hydrological and flood monitoring data, and flood control and drainage engineering data, is used. GIS technology, combined with digital elevation model (DEM) data, is used to extract topographic features of rural areas, such as elevation, slope, and aspect. The DEM data is set to a spatial resolution of 10 meters by 10 meters. Elevation, slope, and aspect features are extracted to generate rural topographic and geomorphic data, including elevation maps, slope maps, and aspect maps. Based on this extracted rural topographic and geomorphic data, a DEM-based flow accumulation analysis method is used to determine the distribution of river systems. The flow accumulation threshold is set at 1,000 grid cells to identify river networks and generate river system distribution data, including river network maps and river flow direction maps. This river system distribution data is combined with hydrological monitoring data and a cluster analysis method based on hydrological characteristics is used to identify the hydrological similarity of river systems. Flood frequency distribution curve parameters, runoff characteristics, and soil characteristics were selected as clustering indicators. Fuzzy clustering was used for similarity identification, resulting in hydrological similarity data, including a similarity matrix for each river system. Based on this hydrological similarity data, river system distribution data was partitioned according to hydrological similarity to generate river hydrological partitioning data. Partitioning thresholds were set to ensure consistency and uniformity in the partitioning results. River hydrological partitioning data, including partition maps and partition characteristic tables, were generated. Using meteorological data from multi-source rural flood data, including rainfall amount and rainfall intensity, a time series analysis-based approach was employed to extract rainfall characteristics in mountainous areas, such as rainfall intensity and duration. A time step of 5 minutes was set to extract rainfall intensity and duration characteristics, resulting in mountainous rainfall characteristic data, including rainfall intensity and duration maps. Based on river hydrological zoning data and combined with mountain rainfall data, reservoir locations are determined based on river hydrological zoning data. Based on mountain rainfall characteristics, reservoir water storage models are used to calculate water storage and generate reservoir rainfall-storage relationship data. Parameters such as the initial reservoir water level and target water level are set for water storage calculations, generating reservoir rainfall-storage relationship data, including reservoir storage curves and rainfall-storage relationship tables. Using reservoir rainfall-storage relationship data and combined with river hydrological zoning data, calculation methods based on hydrodynamic models, such as the MIKE 11 model, are used to calculate the flow direction, velocity, and water level changes in the river network. The model time step is set to 5 minutes and the spatial resolution is 10 meters by 10 meters, accounting for hydrodynamic changes under different rainfall intensities. Hydrodynamic data for the river network is generated, including flow direction maps, velocity maps, and water level change maps.Using river network hydrodynamic data, combined with rural topographic and geomorphological data, and employing a surface hydrodynamic model-based calculation method, we calculated the inflow and infiltration intensity, propagation parameters, and turbulent viscosity coefficient of surface water. Setting the model time step to 5 minutes and a spatial resolution of 5 meters by 5 meters, we considered surface hydrodynamic parameters under different terrain conditions and generated surface hydrodynamic data, including maps of inflow and infiltration intensity, propagation parameters, and turbulent viscosity coefficient. Using river network and surface hydrodynamic data, we employed a coupled model-based calculation method, such as the MIKE 11 and SWMM coupled model, to calculate the forces acting on river floods and surface waterlogging. Setting the model time step to 5 minutes and a spatial resolution of 10 meters by 10 meters, we considered the forces acting under different rainfall intensities and terrain conditions and generated river flood overtopping data, including overtopping range maps, water depth distribution maps, and flow velocity distribution maps.

[0115] Preferably, step S22 includes the following steps:

[0116] Step S221: Marking the flood overflow area of ​​the river flood overflow data to generate the flood overflow area; extracting the terrain elevation, slope, and distance from the river channel of the overflow area from the rural flood multi-source data based on the flood overflow area to obtain spatial feature data of the overflow area;

[0117] Step S222: quantifying the waterlogging risk index based on the spatial characteristic data of the overflow area, and calculating the waterlogging risk index based on the terrain elevation, slope, and distance from the river in the overflow area to obtain a waterlogging risk quantification index;

[0118] Step S223: measuring local two-dimensional hydrological and hydrodynamic parameters for the waterlogging risk quantification index, and calculating the water depth and flow velocity hydrodynamics of the local area to obtain local waterlogging hydrodynamic parameters;

[0119] Step S224: Evaluate the waterlogging risk change characteristics of local waterlogging hydrodynamic parameters to obtain local waterlogging risk data.

[0120] In an embodiment of the present invention, river flood overflow data is combined with geographic information data and GIS technology to perform spatial analysis on the river flood overflow data and mark flood overflow areas. A threshold for the overflow area is set, such as marking areas with a water depth greater than 0.2 meters as overflow areas, and a flood overflow area map is generated to clearly define the boundaries and locations of the overflow areas. Based on the marked flood overflow areas, geographic information data from multi-source rural flood data is combined. GIS extraction tools are used to extract spatial feature data such as terrain elevation, slope, and distance from the river channel in the overflow area. The extraction resolution is set to a spatial resolution of 10 meters x 10 meters. Features such as terrain elevation, slope, and distance from the river channel are extracted to obtain spatial feature data of the overflow area, including a terrain elevation map, a slope map, and a distance from the river channel map. The waterlogging risk index is quantified using the spatial characteristic data of the overflow area. The waterlogging risk index is calculated by combining the overflow area's terrain elevation, slope, and distance from the river to obtain a quantitative waterlogging risk index. Using the spatial characteristic data of the overflow area, including terrain elevation, slope, and distance from the river, a waterlogging risk level assessment model is used to quantify the waterlogging risk index by combining these characteristics. Parameters of the waterlogging risk level assessment model, such as waterlogging depth, flow velocity, and water depth hazard parameters, are set to obtain quantitative waterlogging risk indicators, including the risk level index. Based on the waterlogging risk quantification index, the model combines the overflow area's terrain elevation, slope, and distance from the river. The core calculation formula of the waterlogging risk level assessment model is used, such as RH = d × (v + 0.5) + df, where RH is the risk level index, d is the waterlogging depth, v is the flow velocity, and df is the water depth hazard parameter. Parameter settings: df = 0.5 for waterlogging depth d ≤ 0.2 m; df = 1.0 for d > 0.2 m. Local two-dimensional hydrodynamic parameters were measured for the waterlogging risk quantification indicators, and the water depth and flow velocity hydrodynamics of the local area were calculated to obtain local waterlogging hydrodynamic parameters. The waterlogging risk quantification indicators were combined with geographic information data and hydrological monitoring data. The SWMM model was used to measure local two-dimensional hydrodynamic parameters and simulate the water depth and flow velocity hydrodynamics of the local area. The model time step was set to 5 minutes, and the spatial resolution was 5 m × 5 m. The model considered the hydrodynamic changes under different rainfall intensities and drainage system operating conditions. Local waterlogging hydrodynamic parameters, including water depth and flow velocity distribution maps, were obtained. Based on the results of the local two-dimensional hydrodynamic parameter measurements, the SWMM model was used to calculate the water depth and flow velocity hydrodynamics of the local area. Parameter setting: The model time step is set to 5 minutes and the spatial resolution is 5 meters × 5 meters. The hydrodynamic changes under different rainfall intensities and drainage system working conditions are considered to obtain the local waterlogging hydrodynamic parameters, including the water depth and flow velocity distribution map of the local area.Using local waterlogging hydrodynamic parameters, including the water depth and flow velocity distribution map of the local area, and adopting the waterlogging risk change characteristic assessment model, combined with local waterlogging hydrodynamic parameters, the changing characteristics of waterlogging risk are assessed, and the parameters of the assessment model are set, such as the range of change of the risk level index, to obtain local waterlogging risk data, including the changing characteristic map of waterlogging risk.

[0121] Preferably, step S3 includes the following steps:

[0122] Step S31: setting initial conditions for the flood simulation process based on multi-source rural flood data, including setting initial rainfall intensity, initial water level, and topographic features, to generate simulation initial condition data;

[0123] Step S32: Based on the simulation initial condition data, the rural flood multi-source data is input into the flood overtopping and waterlogging risk model to simulate the rural flood inundation process and obtain the flood inundation simulation results;

[0124] Step S33: determining the flood depth based on the flood simulation result to obtain flood depth simulation data; identifying the flood area boundary based on the flood depth simulation data to obtain flood area boundary data;

[0125] Step S34: performing watershed mapping on the submerged area boundary data to obtain submerged watershed mapping data; performing waterlogging area division on the submerged watershed division data to obtain watershed waterlogging area data;

[0126] Step S35: Detect the rainstorm flood risk area on the watershed waterlogging area data, and evaluate the rural waterlogging risk level to obtain waterlogging risk level data.

[0127] In this embodiment of the present invention, multi-source rural flood data, including geographic information data, hydrological and flood monitoring data, and flood control and drainage engineering data, is used in conjunction with historical flood event data and topographic data to determine initial rainfall intensity, initial water level, and topographic relief characteristics. Parameter settings include: initial rainfall intensity: set to 10 mm / hour; initial water level: set to the average value of river water level monitoring data; and topographic relief characteristics: Digital elevation model (DEM) data is used to extract terrain features such as elevation, slope, and aspect. Simulation initial condition data, including initial rainfall intensity, initial water level, and topographic relief characteristics, is generated. This simulation initial condition data and multi-source rural flood data are input into a flood overtopping and waterlogging risk model. GIS-based flood inundation simulation technology is used, combining a one-dimensional hydraulic model (such as Hec-RAS) with a two-dimensional hydrological and hydrodynamic model (such as SWMM). Parameter settings include: time step: set to 5 minutes; and spatial resolution: set to 10 meters x 10 meters. Consider different recurrence periods: such as 10 years, 20 years, and 50 years; obtain flood inundation simulation results, including inundation range, water depth distribution, flow velocity distribution and other information. Use the flood inundation simulation results and GIS technology to perform spatial analysis on the flood inundation simulation results, extract inundation depth information, and obtain inundation depth simulation data, including inundation depth distribution map. Based on the inundation depth simulation data, adopt a threshold-based boundary recognition method, set the area with inundation depth greater than 0.2 meters as the inundation area, and obtain inundation area boundary data, including the inundation area boundary map. Perform watershed mapping on the inundation area boundary data to obtain inundation watershed mapping data; perform waterlogging area division on the inundation watershed division data to obtain watershed waterlogging area data. Use the inundation area boundary data and the watershed division tool of GIS to map the inundation area boundary data onto the watershed map to obtain inundation watershed mapping data, including the correspondence between the watershed map and the inundation area. Based on the inundation watershed mapping data. Using a zoning method based on topography and hydrological characteristics, combined with GIS cluster analysis tools, waterlogging zones were delineated, resulting in watershed waterlogging zone data, including boundary maps and characteristic tables. Using this watershed waterlogging zone data, a hydrological model-based rainstorm flood risk zone detection method, combined with GIS overlay analysis tools, was used to detect rainstorm flood risk zones, generating data including boundary maps and risk levels. Based on this rainstorm flood risk zone data, a waterlogging risk level assessment model was employed, combined with GIS multi-source data fusion techniques, to assess rural waterlogging risk levels. Parameter settings included: waterlogging depth: less than 15 cm is considered mild waterlogging, while greater than or equal to 15 cm is considered waterlogging. Flow velocity: areas with a flow velocity exceeding 0.5 m / s are considered high-risk. Inundation duration: areas with inundation lasting more than 24 hours are considered high-risk. Waterlogging risk level data was generated, including a waterlogging risk level distribution map and a risk area table.

[0128] Preferably, step S35 includes the following steps:

[0129] Step S351: performing multi-dimensional feature correlation analysis on the flood depth, flood range, and flood propagation speed of the watershed waterlogged area data to obtain multi-dimensional flood correlation data;

[0130] Step S352: locating the rainstorm range of the multi-dimensional flood-related data to obtain rainstorm range data; identifying the rainstorm risk area of ​​the multi-dimensional flood-related data based on the rainstorm range data to generate the rainstorm risk area;

[0131] Step S353: performing a dynamic simulation of the rainstorm and flood risk area, and analyzing the changes of the rainstorm and flood risk area at different time points to obtain dynamic change data of the risk area;

[0132] Step S354: Calculate the risk assessment index for the dynamic change data of the risk area to generate risk assessment index data; classify the risk assessment index data into waterlogging risk levels to obtain waterlogging risk level data.

[0133] In this embodiment of the present invention, watershed flooding data is used, including information such as inundation depth, inundation range, and flood propagation speed. GIS technology, combined with multi-source data fusion methods, is used to perform correlation analysis on inundation depth, inundation range, and flood propagation speed. Parameter settings include: Inundation depth: A threshold of 0.2 meters is set, and areas greater than 0.2 meters are considered flooded. Inundation range: A threshold of 100 square meters is set, and areas greater than 100 square meters are considered flooded. Flood propagation speed: A threshold of 0.5 meters per second is set, and areas greater than 0.5 meters per second are considered high-risk areas. Multidimensional flood correlation data is generated, including a correlation map of inundation depth, inundation range, and flood propagation speed. Using this multidimensional flood correlation data, a GIS-based rainstorm range location method is employed, combining meteorological and topographic data, to determine the rainstorm range. Parameter settings: A rainstorm intensity threshold of 10 mm / hour is set, and areas greater than 10 mm / hour are considered rainstorm ranges. Rainstorm range data, including a rainstorm range map, is generated. Based on the rainstorm range data, multidimensional flood correlation data is combined. A GIS-based method for identifying rainstorm flood risk areas was used to identify rainstorm flood risk areas, combining characteristics such as inundation depth, inundation range, and flood propagation speed. Rainstorm flood risk areas were generated, including risk area boundary maps and characteristic tables. Using rainstorm flood risk area data, a GIS-based dynamic simulation method was used, combined with time series analysis, to dynamically simulate the rainstorm flood risk areas. Parameter settings: A time step of 5 minutes was set to simulate the changes in risk areas at different time points. Dynamic risk area change data was generated, including risk area change maps at different time points. Based on the risk area dynamic change data, a GIS-based spatiotemporal analysis method was used to analyze the changes in rainstorm flood risk areas at different time points. Dynamic risk area change data was generated, including risk area change trend maps and change characteristic tables. Using the risk area dynamic change data, a GIS-based risk assessment indicator calculation method was used to calculate risk assessment indicators, combining characteristics such as inundation depth, inundation range, and flood propagation speed. Parameter settings: Waterlogging depth: Waterlogging depths less than 15 cm are considered mild waterlogging, while waterlogging depths greater than or equal to 15 cm are considered waterlogging. Flow velocity: Areas with a flow velocity exceeding 0.5 m / s are considered high-risk. Inundation duration: Areas with an inundation duration exceeding 24 hours are considered high-risk. Risk assessment indicator data is generated, including a risk assessment indicator map and indicator table. Based on this risk assessment indicator data, a GIS-based waterlogging risk classification method is used, combined with risk assessment indicators, to categorize waterlogging risk levels. Parameter settings: Risk level index: 0-10 represents low risk, 10-20 represents medium risk, and 20-30 represents high risk. Waterlogging risk level data is generated, including a waterlogging risk level distribution map and a risk area table.

[0134] As an example of the present invention, refer to Figure 3 As shown, in this example, step S4 includes:

[0135] Step S41: Arrange the waterlogging risk level data in reverse order of the risk level gradient to obtain risk level gradient data; map low-, medium-, and high-risk rural waterlogging areas based on the risk level gradient data to obtain waterlogging risk mapping areas;

[0136] Step S42: Preliminary selection of flood risk warning indicators for the waterlogging risk mapping area to obtain preliminary selected warning indicator data; performing correlation analysis on the preliminary selected warning indicator data to obtain warning indicator correlation data; performing sensitivity analysis on the preliminary selected warning indicator data to obtain warning indicator sensitivity data;

[0137] Step S43: screening flood risk warning indicators from the preliminary warning indicator data based on the warning indicator correlation data and the warning indicator sensitivity data, and constructing a warning indicator system to generate a rural flood warning indicator system;

[0138] Step S44: constructing a flood warning model through the rural flood warning indicator system to generate a real-time flood forecast warning model;

[0139] Step S45: Predict rural flood risk areas based on the real-time flood forecast and warning model to obtain rural flood risk area prediction data; plan risk avoidance transfer routes based on the rural flood risk area prediction data to generate a rural flood warning planning report.

[0140] In the embodiment of the present invention, the waterlogging risk level data is used and GIS technology is used to arrange the waterlogging risk level data in reverse order to generate risk level gradient data. Parameter setting: The risk levels are set from high to low as high risk (3), medium risk (2), and low risk (1). Risk level gradient data is obtained, including a risk level gradient map. Low, medium, and high risk rural waterlogging area mapping: Based on the risk level gradient data, the risk level gradient data is mapped onto the rural waterlogging area map using GIS mapping tools to generate waterlogging risk mapping areas. Waterlogging risk mapping areas are obtained, including distribution maps of low, medium, and high risk areas. Waterlogging risk mapping area data is used, including information such as flooding depth, flooding range, and flood propagation speed. Combined with historical flood data and expert experience, flood risk warning indicators are preliminarily selected. Common warning indicators include water depth, flow rate, rainfall intensity, flooding time, terrain elevation, slope, etc. Preliminary warning indicator data are obtained, including preliminarily selected warning indicators and their data. Based on the preliminary data of early warning indicators, statistical analysis methods, such as Pearson correlation coefficient analysis, are used to calculate the correlations between each early warning indicator. The details are as follows: Data standardization: Standardize the data of each early warning indicator to eliminate dimensionality effects. Correlation coefficient calculation: Use the Pearson correlation coefficient formula to calculate the correlation coefficients between each indicator. Output: Obtain early warning indicator correlation data, including a correlation coefficient matrix between each early warning indicator. Based on the preliminary data of early warning indicators, sensitivity analysis methods, such as single-factor sensitivity analysis, are used to calculate the sensitivity of each early warning indicator to urban waterlogging risk. The details are as follows: Setting a baseline scenario: Select a baseline scenario, such as the current rainfall intensity and terrain conditions. Changing factors one by one: Change each early warning indicator one by one, keeping other factors constant, and observe the changes in urban waterlogging risk. Calculating sensitivity indices: Calculate the sensitivity index of each early warning indicator based on the magnitude of the change in urban waterlogging risk. Obtain early warning indicator sensitivity data, including the sensitivity index of each early warning indicator. Using the early warning indicator correlation data and early warning indicator sensitivity data, combined with the correlation and sensitivity analysis results, screen out early warning indicators with high correlation and strong sensitivity. Screened early warning indicator data is obtained, including the screened early warning indicators and their data. Based on the screened warning indicator data, a rural flood warning indicator system is constructed using the Analytic Hierarchy Process (AHP) or Principal Component Analysis (PCA). This system is designed to generate a rural flood warning indicator system, including the weights of each warning indicator and a comprehensive evaluation formula. Using this system, machine learning models, such as random forests (RF) or support vector machines (SVM), are trained in conjunction with historical flood data. Parameter settings: Set model training parameters, such as the number of trees in the random forest and the kernel function of the support vector machine. A real-time flood forecast and warning model is generated, enabling real-time prediction of rural flood risks. Using this model, real-time forecasts are performed using the model, combined with real-time meteorological and hydrological data, to predict rural flood risk areas.Obtain forecast data for rural flood risk areas, including forecast time, risk zone scope, and risk level. Based on this data, GIS network analysis tools, combined with road networks and settlement distribution, plan evacuation routes. Generate a evacuation route map, including information on evacuation routes and resettlement sites. Integrate the rural flood risk area forecast data with the evacuation route planning results to compile a rural flood early warning planning report, including a warning indicator system, model construction methods, risk area forecast results, and evacuation route planning. This report provides a scientific basis for early warning and response to rural flood disasters.

[0141] Preferably, step S44 includes the following steps:

[0142] Step S441: determining meteorological conditions on rural flood multi-source data to obtain meteorological condition performance data; performing flood warning demand analysis on the meteorological condition performance data to generate flood warning demand data;

[0143] Step S442: constructing a flood warning framework based on the rural flood warning indicator system to obtain flood warning framework information; calibrating the flood warning framework information based on the flood warning demand data to obtain a warning calibration framework;

[0144] Step S443: constructing a flood warning model for the warning calibration framework to generate an initial flood warning model; performing model warning verification on the initial flood warning model to obtain warning performance verification data; optimizing response parameters of the initial flood warning model based on the warning performance verification data to obtain a flood warning response model;

[0145] Step S444: Acquire real-time meteorological monitoring data and real-time water level monitoring data to obtain real-time monitoring data; input the real-time monitoring data into the flood warning response model, and conduct real-time flood forecast and warning training to generate a real-time flood forecast and warning model.

[0146] In this embodiment of the present invention, multi-source data on rural floods and waterlogging is used, including meteorological data, topographic data, and hydrological monitoring data. Combined with data from meteorological monitoring stations, current meteorological conditions, such as rainfall intensity, rainfall duration, wind speed, and wind direction, are determined. Meteorological condition performance data, including rainfall intensity and duration maps, wind speed and wind direction maps, are obtained. Based on this meteorological condition performance data and combined with historical flood data and topographic data, the potential flood risk under current meteorological conditions is analyzed to determine the need for flood warnings. Flood warning demand data is generated, including warning levels, warning areas, and warning times. A rural flood warning indicator system, including indicators such as waterlogging depth, flow velocity, rainfall intensity, and inundation duration, is used. The Analytic Hierarchy Process (AHP) or Principal Component Analysis (PCA) is used to construct a flood warning framework, defining the weights of each warning indicator and a comprehensive evaluation formula. This flood warning framework information is obtained, including the weights of each warning indicator, a comprehensive evaluation formula, and warning thresholds. Based on and in conjunction with flood warning demand data, the flood warning framework was calibrated, adjusting the weights and thresholds of warning indicators to ensure the accuracy and reliability of the warning framework. This calibration framework, including the calibrated warning indicator weights, comprehensive evaluation formula, and warning thresholds, was obtained. Using the calibration framework, including the calibrated warning indicator weights, comprehensive evaluation formula, and warning thresholds, machine learning models, such as random forests (RF) or support vector machines (SVM), were trained in conjunction with historical flood data. This generated an initial flood warning model capable of providing preliminary predictions of rural flood risk. Based on this initial flood warning model, a cross-validation method was used to validate the initial model and evaluate its accuracy and generalization ability. Validation data for warning performance was obtained, including evaluation metrics such as mean squared error (MSE), root mean squared error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and coefficient of determination (R²). Based on the early warning performance validation data, the response parameters of the initial flood warning model were optimized, such as adjusting the learning rate, increasing the number of iterations, and adjusting the regularization parameter. This resulted in a flood warning response model that demonstrated good predictive performance on the training, cross-validation, and test sets. Using real-time data from meteorological and water level monitoring stations, IoT technology was used to acquire real-time meteorological and water level data, including rainfall intensity, duration, wind speed, wind direction, and water level. This data, including real-time rainfall intensity and water level maps, was then fed into the flood warning response model for real-time flood forecast and warning training. The model's parameters were optimized to improve its real-time prediction capabilities. The resulting real-time flood forecast and warning model can predict rural flood risks in real time, providing a scientific basis for early warning and response to flood disasters.

[0147] Preferably, step S45 includes the following steps:

[0148] Step S451: Based on the real-time flood forecasting and warning model, the real-time monitoring data is used to predict rural flood scenarios to obtain rural flood scenario prediction data; the rural flood scenario prediction data is used to delineate flooding ranges to obtain flooding range delineation data;

[0149] Step S452: performing spatiotemporal analysis on the flood range delineation data, analyzing the spatial distribution characteristics of each flood range at different time points, and obtaining spatiotemporal prediction data of the flood range;

[0150] Step S453: Determine the flood range change trend based on the spatiotemporal flood range prediction data to obtain flood range change trend information; perform multi-path planning of evacuation routes based on the flood range change trend information, plan multiple evacuation routes for each flood range, and obtain multiple evacuation route data;

[0151] Step S454: performing route feasibility assessment on the plurality of risk avoidance route data to obtain risk avoidance route feasibility data; determining an optimal risk avoidance route for the plurality of risk avoidance route data based on the risk avoidance route feasibility data to generate optimal risk avoidance route data;

[0152] Step S455: compile an early warning planning report based on the optimal risk avoidance route data to obtain a rural flood early warning planning report.

[0153] In an embodiment of the present invention, a real-time flood forecasting and warning model and real-time monitoring data, including meteorological data and water level data, are used. The real-time flood forecasting and warning model, combined with real-time monitoring data, is used to predict rural flood scenarios. Rural flood scenario prediction data is obtained, including information such as the predicted time point, inundation area, and water depth distribution. Based on this rural flood scenario prediction data, GIS technology is used in conjunction with topographic data to delineate inundation areas. Flood area delineation data, including inundation area maps, is obtained. Using this inundation area delineation data, GIS spatiotemporal analysis tools are used to analyze the spatial distribution characteristics of each inundation area at different time points. Spatiotemporal flood area prediction data, including inundation area distribution maps at different time points, is obtained. Trend analysis methods are used to determine the changing trends of the inundation area using this spatiotemporal flood area prediction data. Information on the changing trends of the inundation area is obtained, including the expansion or contraction of the inundation area. Based on this inundation area changing trend information, GIS network analysis tools are used in conjunction with road networks and settlement distribution to plan multiple evacuation routes for each inundation area. Obtain data on multiple evacuation routes, including multiple evacuation route maps. Using the data on multiple evacuation routes, employ GIS overlay analysis tools, and combine topography, road conditions, and settlement distribution to assess the feasibility of evacuation routes. Obtain feasibility data on evacuation routes, including feasibility assessment reports for each evacuation route. Based on the feasibility data on evacuation routes, select the optimal evacuation route. Generate optimal evacuation route data, including an optimal evacuation route map. Use the optimal evacuation route data, combined with GIS technology and text editing tools, to compile a rural flood warning planning report, including a warning indicator system, model building methods, risk area prediction results, and evacuation route planning. Obtain a rural flood warning planning report to provide a scientific basis for warning and response to rural flood disasters.

[0154] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0155] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for assessing and warning the risk level of rural waterlogged areas based on multi-source data, characterized in that: The following steps are involved: Step S1: Obtain rural waterlogged area geographic information, hydrological and flood data, and flood control and drainage engineering data, and perform multi-source data fusion to obtain rural flood multi-source data; Step S2: Calculate river flood overflows on the rural flood multi-source data to obtain river flood overflow data; perform a two-dimensional hydrological and hydrodynamic quantitative assessment of waterlogging risk points on the rural flood multi-source data based on the river flood overflow data to obtain local waterlogging risk data; and construct a flood overflow and waterlogging risk model based on the river flood overflow data and the local waterlogging risk data. The river flood overflow calculation is performed on the rural flood multi-source data to obtain the river flood overflow data including: Extract topographic and geomorphic features from multi-source data of rural floods to obtain rural topographic and geomorphic features; determine the river system distribution based on the rural topographic and geomorphic features to generate river system distribution data; Perform hydrological similarity recognition on river system distribution data to obtain hydrological similarity; partition river system distribution data into river hydrologically similar areas based on the hydrological similarity to generate river hydrological partition data; Extracting mountain rainfall characteristics from multi-source rural flood data to obtain mountain rainfall characteristics; determining reservoir locations based on river hydrological zoning data, and calculating reservoir storage based on mountain rainfall characteristics to generate reservoir rainfall-storage relationship data; Based on the rainfall-storage relationship data of the reservoir, the river network flow direction, flow velocity and water level change are calculated for the river hydrological division data to obtain the river network hydrodynamic data; Based on the river network hydrodynamic data, the surface water infiltration intensity, propagation parameters and turbulent viscosity coefficient are calculated for the rural topographic features to obtain the surface hydrodynamic data. Calculate the forces acting on river floods and surface waterlogging based on river network hydrodynamic data and surface hydrodynamic data to obtain river flood overflow data; Step S3: simulating the rural flood inundation process using a flood overtopping and waterlogging risk model to obtain flood inundation simulation results; dividing the flood inundation simulation results into watershed waterlogging areas to obtain watershed waterlogging area data; detecting rainstorm flood risk areas on the watershed waterlogging area data, and conducting a rural waterlogging risk level assessment to obtain waterlogging risk level data; Step S4: Determine rural flood risk warning indicators based on the waterlogging risk level data and generate a rural flood warning indicator system; construct a real-time flood forecast and warning model based on the rural flood warning indicator system, and predict rural flood risk areas based on the real-time flood forecast and warning model, plan risk avoidance and transfer routes, and obtain a rural flood warning planning report; constructing the real-time flood forecast and warning model based on the rural flood warning indicator system includes: Determine meteorological conditions based on multi-source data on rural floods and obtain meteorological condition performance data; perform flood warning demand analysis on meteorological condition performance data to generate flood warning demand data; Construct a flood warning framework based on the rural flood warning indicator system to obtain flood warning framework information; calibrate the flood warning framework information based on flood warning demand data to obtain a warning calibration framework; Construct a flood warning model for the warning calibration framework to generate an initial flood warning model; perform model warning verification on the initial flood warning model to obtain warning performance verification data; optimize the response parameters of the initial flood warning model based on the warning performance verification data to obtain a flood warning response model; Real-time meteorological monitoring data and real-time water level monitoring data are obtained to obtain real-time monitoring data; the real-time monitoring data is input into the flood warning response model, and real-time flood forecast and warning training is carried out to generate a real-time flood forecast and warning model.

2. The method for risk assessment and early warning of rural waterlogged areas based on multi-source data according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: determining the rural flood area to be monitored; Step S12: Collecting geographic information of the rural flood-prone areas to be monitored to obtain rural geographic information data; Step S13: Perform hydrological and flood monitoring on the rural flood-prone area to be monitored to obtain hydrological and flood monitoring data; Step S14: Obtaining flood control and drainage engineering data in the rural flood-prone area to be monitored, and performing structural digital conversion to generate flood control and drainage engineering data; Step S15: pre-processing the rural geographic information data, hydrological and flood monitoring data, and flood drainage engineering data to obtain rural flood processing data, wherein the data pre-processing includes data cleaning, data format standardization, and feature extraction and conversion; Step S16: performing data mapping and alignment on the rural flood processing data, and merging multi-source data to obtain rural flood multi-source data.

3. The method for risk assessment and early warning of rural waterlogged areas based on multi-source data according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: determining river hydrological zones for the rural flood multi-source data, and performing river flood overflow calculations to obtain river flood overflow data; Step S22: performing a two-dimensional hydrological and hydrodynamic quantitative assessment of local waterlogging risk points on the rural flood multi-source data based on the river flood overflow data to obtain local waterlogging risk data; Step S23: constructing flood overflow and waterlogging risk based on the river flood overflow data and the local waterlogging risk data to obtain a flood overflow and waterlogging risk pre-model; training the flood overflow and waterlogging risk pre-model based on the river flood overflow data and the local waterlogging risk data to obtain a flood overflow and waterlogging risk training model; Step S24: performing model cross-validation evaluation on the flood overflow and waterlogging risk training model to obtain training model evaluation data; adjusting model parameters of the flood overflow and waterlogging risk training model using the training model evaluation data to obtain a flood overflow and waterlogging risk model.

4. The method for risk assessment and early warning of rural waterlogged areas based on multi-source data according to claim 3 is characterized in that: Step S22 includes the following steps: Step S221: Marking the flood overflow area of ​​the river flood overflow data to generate the flood overflow area; extracting the terrain elevation, slope, and distance from the river channel of the overflow area from the rural flood multi-source data based on the flood overflow area to obtain spatial feature data of the overflow area; Step S222: quantifying the waterlogging risk index based on the spatial characteristic data of the overflow area, and calculating the waterlogging risk index based on the terrain elevation, slope, and distance from the river in the overflow area to obtain a waterlogging risk quantification index; Step S223: measuring local two-dimensional hydrological and hydrodynamic parameters for the waterlogging risk quantification index, and calculating the water depth and flow velocity hydrodynamics of the local area to obtain local waterlogging hydrodynamic parameters; Step S224: Evaluate the waterlogging risk change characteristics of local waterlogging hydrodynamic parameters to obtain local waterlogging risk data.

5. The method for risk assessment and early warning of rural waterlogged areas based on multi-source data according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: setting initial conditions for the flood simulation process based on multi-source rural flood data, including setting initial rainfall intensity, initial water level, and topographic features, to generate simulation initial condition data; Step S32: Based on the simulation initial condition data, the rural flood multi-source data is input into the flood overtopping and waterlogging risk model to simulate the rural flood inundation process and obtain the flood inundation simulation results; Step S33: determining the flood depth based on the flood simulation result to obtain flood depth simulation data; identifying the flood area boundary based on the flood depth simulation data to obtain flood area boundary data; Step S34: performing watershed mapping on the submerged area boundary data to obtain submerged watershed mapping data; performing waterlogging area division on the submerged watershed division data to obtain watershed waterlogging area data; Step S35: Detect the rainstorm flood risk area on the watershed waterlogging area data, and evaluate the rural waterlogging risk level to obtain waterlogging risk level data.

6. The method for risk assessment and early warning of rural waterlogged areas based on multi-source data according to claim 5 is characterized in that: Step S35 includes the following steps: Step S351: performing multi-dimensional feature correlation analysis on the flood depth, flood range, and flood propagation speed of the watershed waterlogged area data to obtain multi-dimensional flood correlation data; Step S352: locating the rainstorm range of the multi-dimensional flood-related data to obtain rainstorm range data; identifying the rainstorm risk area of ​​the multi-dimensional flood-related data based on the rainstorm range data to generate the rainstorm risk area; Step S353: performing a dynamic simulation of the rainstorm and flood risk area, and analyzing the changes of the rainstorm and flood risk area at different time points to obtain dynamic change data of the risk area; Step S354: Calculate the risk assessment index for the dynamic change data of the risk area to generate risk assessment index data; classify the risk assessment index data into waterlogging risk levels to obtain waterlogging risk level data.

7. The method for risk assessment and early warning of rural waterlogged areas based on multi-source data according to claim 1 is characterized in that: Step S4 includes the following steps: Step S41: Arrange the waterlogging risk level data in reverse order of the risk level gradient to obtain risk level gradient data; map low-, medium-, and high-risk rural waterlogging areas based on the risk level gradient data to obtain waterlogging risk mapping areas; Step S42: Preliminary selection of flood risk warning indicators for the waterlogging risk mapping area to obtain preliminary selected warning indicator data; performing correlation analysis on the preliminary selected warning indicator data to obtain warning indicator correlation data; performing sensitivity analysis on the preliminary selected warning indicator data to obtain warning indicator sensitivity data; Step S43: screening flood risk warning indicators from the preliminary warning indicator data based on the warning indicator correlation data and the warning indicator sensitivity data, and constructing a warning indicator system to generate a rural flood warning indicator system; Step S44: constructing a flood warning model through the rural flood warning indicator system to generate a real-time flood forecast warning model; Step S45: Predict rural flood risk areas based on the real-time flood forecast and warning model to obtain rural flood risk area prediction data; plan risk avoidance transfer routes based on the rural flood risk area prediction data to generate a rural flood warning planning report.

8. The method for risk assessment and early warning of rural waterlogged areas based on multi-source data according to claim 1 is characterized in that: Step S45 includes the following steps: Step S451: Based on the real-time flood forecasting and warning model, the real-time monitoring data is used to predict rural flood scenarios to obtain rural flood scenario prediction data; the rural flood scenario prediction data is used to delineate flooding ranges to obtain flooding range delineation data; Step S452: performing spatiotemporal analysis on the flooding range delineation data, analyzing the spatial distribution characteristics of each flooding range at different time points, and obtaining spatiotemporal prediction data of the flooding range; Step S453: Determine the flood range change trend based on the spatiotemporal flood range prediction data to obtain flood range change trend information; perform multi-path planning of evacuation routes based on the flood range change trend information, plan multiple evacuation routes for each flood range, and obtain multiple evacuation route data; Step S454: performing route feasibility assessment on the plurality of risk avoidance route data to obtain risk avoidance route feasibility data; determining an optimal risk avoidance route for the plurality of risk avoidance route data based on the risk avoidance route feasibility data to generate optimal risk avoidance route data; Step S455: compile an early warning planning report based on the optimal risk avoidance route data to obtain a rural flood early warning planning report.

Images