Model construction method, system and device for flood forecasting of small water conservancy facilities

Abstract

The present invention relates to a model construction method, system and device for flood forecasting of small-scale water conservancy facilities, and the method comprises: obtaining original water conservancy data of a target watershed of a small-scale water conservancy facility, and performing watershed analysis to obtain corresponding hydrological and meteorological data; obtaining geographic data, and performing characteristic parameter analysis to obtain a corresponding watershed characteristic parameter set; obtaining facility information, and setting parameters of a preset WRF-Hydro model in combination with the watershed characteristic parameter set to obtain an initial flood model; performing multi-objective parameter calibration processing on the initial flood model according to the hydrological and meteorological data to obtain a flood forecasting model; performing scenario simulation on the flood forecasting model to obtain scenario simulation results; optimizing the flood forecasting model according to the model optimization parameters to obtain an optimized flood forecasting model. The present invention can improve the accuracy of flood forecasting and provide a more reliable basis for disaster prevention and mitigation.

Description

Technical Field

[0001] The present invention relates to the technical field of model construction, and in particular to a model construction method, system and device for flood forecasting of small-scale water conservancy facilities. Background Art

[0002] As an important tool for disaster prevention and mitigation, the flood forecasting model of small-scale water conservancy facilities has important application value in water conservancy projects. With the frequent occurrence of climate change and extreme weather events, how to accurately forecast floods to protect people's lives and property has become one of the research focuses. Existing flood forecasting methods for small-scale water conservancy facilities often only consider a single factor, such as the total rainfall in the basin or historical flood data, while ignoring the combined effects of basin geographical characteristics, facility information, and multiple rainfall scenarios. This simplified method may lead to insufficient accuracy in flood forecasting, thereby affecting the effectiveness of disaster prevention and mitigation measures. Summary of the invention

[0003] The main purpose of the present invention is to provide a model construction method, system and device for flood forecasting of small-scale water conservancy facilities, which can improve the accuracy of flood forecasting and provide a more reliable basis for disaster prevention and mitigation.

[0004] To achieve the above object, the present invention provides a model construction method for flood forecasting of small water conservancy facilities, comprising:

[0005] Obtain the original water conservancy data of the target watershed of small water conservancy facilities, conduct watershed analysis, and obtain the corresponding hydrological and meteorological data;

[0006] Obtain geographic data and perform characteristic parameter analysis to obtain the corresponding watershed characteristic parameter set;

[0007] Acquire facility information, and set parameters of a preset WRF-Hydro model in combination with the watershed characteristic parameter set to obtain an initial flood model;

[0008] Performing multi-objective parameter calibration processing on the initial flood model according to the hydrological and meteorological data to obtain a flood forecasting model;

[0009] Performing scenario simulation on the flood forecasting model according to a plurality of preset rainfall scenarios and watershed conditions to obtain scenario simulation results;

[0010] The scenario simulation results are subjected to parameter optimization analysis according to preset independent flood event conditions to obtain corresponding model optimization parameters, and the flood forecasting model is optimized according to the model optimization parameters to obtain the optimized flood forecasting model.

[0011] Furthermore, the original water conservancy data of the target watershed of the small water conservancy facility is obtained, and the watershed analysis is performed to obtain the corresponding hydrological and meteorological data, including:

[0012] Acquire the original water conservancy data of the target watershed, and perform hydrological element analysis on the target watershed to obtain the main hydrological elements of the watershed;

[0013] Conduct basin rainfall-runoff analysis based on the main hydrological elements of the basin to obtain a set of hydrological and meteorological parameters;

[0014] Performing hydrological and meteorological information correlation analysis on the original water conservancy data to obtain basin hydrological and meteorological information data;

[0015] Multi-dimensional data fusion processing is performed on the basin hydrological and meteorological information data and the hydrological and meteorological parameter set to obtain the hydrological and meteorological data.

[0016] Furthermore, the acquisition of geographic data and analysis of characteristic parameters to obtain a corresponding watershed characteristic parameter set includes:

[0017] Acquire the geographic data of the small-scale water conservancy facilities, and perform digital elevation processing to obtain terrain elevation data;

[0018] Analyzing the slope and aspect of the target watershed according to the terrain elevation data to obtain slope data and aspect data;

[0019] Extracting river network from the terrain elevation data to obtain river network morphology data;

[0020] Calculating the river length and river density of the target basin according to the river network morphology data to obtain river length data and river density data;

[0021] Acquire regional soil type data of the target watershed, and perform soil permeability analysis to obtain soil permeability data;

[0022] The terrain elevation data, slope data, aspect data, river network morphology data, river length data, river density data and soil permeability data are integrated to obtain the watershed characteristic parameter set.

[0023] Furthermore, the acquisition of facility information and setting parameters of a preset WRF-Hydro model in combination with the watershed characteristic parameter set to obtain an initial flood model include:

[0024] Obtaining structural parameters and historical operation data of small water conservancy facilities, and integrating and processing them to obtain the facility information;

[0025] Inputting the watershed characteristic parameter set and the facility information into a preset WRF-Hydro model, and performing model parameter initialization processing to obtain a corresponding initial characteristic parameter set;

[0026] Performing basic parameter training on the WRF-Hydro model according to the initial characteristic parameter set to obtain the initial flood model;

[0027] Among them, basic parameter training includes:

[0028] Performing DEM processing on the initial characteristic parameter set to obtain the watershed slope and watershed area, analyzing the river network density, and training the watershed characteristic parameters of the WRF-Hydro model based on the watershed slope, the watershed area and the river network density;

[0029] Generating a meteorological parameter distribution map according to the initial characteristic parameter set, and performing meteorological parameter training;

[0030] Performing river basin analysis on the initial characteristic parameter set to obtain a river roughness coefficient and setting corresponding reservoir dispatching rules;

[0031] Training the time step and spatial grid size of the model according to the basin area, the river network density and the model computing capacity;

[0032] The inflow and outflow boundaries of the basin are analyzed for the basin area and the river network density, and boundary conditions are set based on the boundary parameters obtained from the analysis.

[0033] Furthermore, the multi-objective parameter calibration process is performed on the initial flood model according to the hydrological and meteorological data to obtain a flood forecast model, including:

[0034] Performing a spatiotemporal distribution analysis on the rainfall data in the hydrological and meteorological data to obtain the spatiotemporal distribution characteristics of rainfall;

[0035] Performing flow process line analysis on the flow data in the hydrological and meteorological data to obtain flow process line characteristics;

[0036] Performing soil moisture analysis on the soil moisture data in the hydrological and meteorological data to obtain soil moisture characteristics;

[0037] Comprehensively calibrating the initial flood model according to the temporal and spatial distribution characteristics of rainfall, the flow process line characteristics and the soil moisture characteristics to obtain a calibrated flood model;

[0038] Comparing the preset standard flood data with the calibration flood model to obtain the error result of the model;

[0039] The calibration flood model is subjected to secondary parameter adjustment according to the error result to obtain the flood forecast model.

[0040] Furthermore, the flood forecasting model is subjected to scenario simulation according to the preset multiple rainfall scenarios and watershed conditions to obtain scenario simulation results, including:

[0041] Inputting the multiple rainfall scenarios and the basin conditions into the flood forecasting model, performing data formatting and tensorization processing through the input data processing layer of the flood forecasting model to generate an input tensor;

[0042] Inputting the input tensor into a rainfall-runoff conversion layer, performing coupling analysis of rainfall input and geographical hydrological characteristics through the rainfall-runoff conversion layer, and obtaining the runoff time series information;

[0043] Performing local feature extraction and analysis on the runoff time series information through the local time feature extraction sublayer of the rainfall runoff conversion layer to obtain a local time feature tensor;

[0044] The local time characteristic tensor is input into the confluence simulation layer, and the water flow path in the basin is simulated through the confluence simulation layer to generate the flow time series tensor of the basin outlet;

[0045] Performing overall path analysis on the flow time series tensor through the overall confluence path analysis sublayer of the confluence simulation layer to obtain confluence characteristic information;

[0046] The flow time series tensor is input into the hydrodynamic simulation layer, and the water level change in the basin is simulated through the hydrodynamic simulation layer to obtain the water level change time series;

[0047] Extracting dynamic response features of the water level change time series through the dynamic response analysis sublayer of the hydrodynamic simulation layer to obtain water level change feature information;

[0048] Inputting the water level change time series into the flood process simulation layer, simulating the flood propagation process through the flood process simulation layer, and generating a flood inundation characteristic tensor;

[0049] The flood process simulation layer uses the regional inundation analysis sublayer to identify high-risk areas of the flood inundation feature tensor to obtain inundation area information;

[0050] Through the comprehensive analysis and result output layer of the flood forecasting model, the time series characteristic information of the runoff, the confluence characteristic information, the water level change characteristic information and the flooded area information are comprehensively analyzed to obtain the scenario simulation result.

[0051] Furthermore, the scenario simulation results are subjected to parameter optimization analysis according to the preset independent flood event conditions to obtain corresponding model optimization parameters, and the flood forecast model is optimized according to the model optimization parameters to obtain the optimized flood forecast model, including:

[0052] Performing feature screening on the simulation results according to preset independent flood event conditions to obtain flood event feature parameters;

[0053] Performing parameter sensitivity analysis on the flood forecasting model according to the characteristic parameters of the flood event to obtain a parameter sensitivity ranking;

[0054] Performing multi-objective parameter optimization analysis on the flood forecasting model according to parameter sensitivity ranking to obtain an optimized parameter set;

[0055] Recalibrating the initial flood model according to the optimized parameter set to obtain the recalibrated flood forecast model;

[0056] Conducting independent flood event simulation verification on the recalibrated flood forecasting model to obtain simulation verification results;

[0057] Perform error analysis on the simulation verification results to obtain error analysis information;

[0058] Adjusting the optimization parameter set according to the error analysis information to obtain final model optimization parameters;

[0059] The recalibrated flood forecast model is finally optimized according to the final model optimization parameters to obtain the optimized flood forecast model.

[0060] The present invention also provides a model construction system for flood forecasting of small water conservancy facilities, which is applied to any one of the above-mentioned model construction methods for flood forecasting of small water conservancy facilities, comprising:

[0061] A collection module, which is used to obtain the original water conservancy data of the target watershed of the small water conservancy facility, and perform watershed analysis to obtain the corresponding hydrological and meteorological data;

[0062] An analysis module, which is used to obtain geographic data and perform characteristic parameter analysis to obtain a corresponding watershed characteristic parameter set;

[0063] An association module, the association module is used to obtain facility information, and set parameters of a preset WRF-Hydro model in combination with the basin characteristic parameter set to obtain an initial flood model;

[0064] A processing module, the processing module is used to perform multi-objective parameter calibration processing on the initial flood model according to the hydrological and meteorological data to obtain a flood forecast model;

[0065] A control module, the control module is used to perform scenario simulation on the flood forecasting model according to a plurality of preset rainfall scenarios and watershed conditions to obtain scenario simulation results;

[0066] An execution module is used to perform parameter optimization analysis on the scenario simulation results according to preset independent flood event conditions to obtain corresponding model optimization parameters, and optimize the flood forecasting model according to the model optimization parameters to obtain the optimized flood forecasting model.

[0067] The present invention also provides a model building device for flood forecasting of small water conservancy facilities, comprising:

[0068] Memory, used to store programs;

[0069] The processor is used to execute the program to implement each step of the model building method for flood forecasting of small-scale water conservancy facilities described in any one of the above.

[0070] The present invention provides a model construction method, system and device for flood forecasting of small water conservancy facilities, which has the following beneficial effects:

[0071] By conducting a comprehensive basin analysis and characteristic parameter analysis on the target basins of small-scale water conservancy facilities and combining facility information, a flood forecasting model can be constructed more accurately, thereby improving the accuracy of flood forecasting and providing a more reliable basis for disaster prevention and mitigation. By performing multi-objective parameter calibration on the initial flood model, the flood evolution process under different hydrological and meteorological conditions can be more effectively reflected, which is helpful to achieve accurate flood forecasting and avoid the risks caused by forecast errors. Scenario simulation of the flood forecasting model based on a variety of rainfall scenarios and basin conditions can ensure that the model can operate efficiently under various extreme weather events, reduce forecast uncertainty, and improve the effectiveness of disaster prevention and mitigation measures. By performing parameter optimization analysis on the scenario simulation results and optimizing the flood forecasting model based on the model optimization parameters, a more reasonable flood control scheduling strategy can be formulated, and the optimized operation of the entire flood control system can be achieved through a comprehensive adjustment strategy, thereby effectively utilizing water resources and reducing disaster losses and environmental impacts. At the same time, by considering the impact of basin characteristic parameters and facility information on flood forecasting, the flood forecasting strategy can be flexibly adjusted according to the characteristics and demand changes of different basins, making the flood control system more adaptable to diverse demand scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 It is a flow chart of a model construction method for flood forecasting of small water conservancy facilities provided by the present invention;

[0073] Figure 2It is a structural diagram of a model building system for flood forecasting of small water conservancy facilities provided by the present invention;

[0074] Figure 3 The present invention provides a structural diagram of a model building device for flood forecasting of small-scale water conservancy facilities.

[0075] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0076] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0077] The present invention is further described below in conjunction with the accompanying drawings and specific implementation methods.

[0078] Reference Figure 1 As shown, the present invention provides a model construction method for flood forecasting of small water conservancy facilities, comprising:

[0079] Step S1: obtaining the original water conservancy data of the target watershed of the small water conservancy facility, and performing watershed analysis to obtain the corresponding hydrological and meteorological data;

[0080] Step S2: Obtain geographic data and perform characteristic parameter analysis to obtain a corresponding watershed characteristic parameter set;

[0081] Step S3: Obtain facility information, and set parameters for the preset WRF-Hydro model in combination with the basin characteristic parameter set to obtain an initial flood model;

[0082] Step S4: performing multi-objective parameter calibration processing on the initial flood model according to the hydrological and meteorological data to obtain a flood forecast model;

[0083] Step S5: simulating the flood forecasting model according to various preset rainfall scenarios and watershed conditions to obtain scenario simulation results;

[0084] Step S6: Optimize the parameters of the scenario simulation results according to the preset independent flood event conditions to obtain the corresponding model optimization parameters, optimize the flood forecasting model according to the model optimization parameters to obtain the optimized flood forecasting model.

[0085] Based on the above steps, the detailed process is as follows:

[0086] Step S1: Obtain the original water conservancy data of the target basin, including rainfall, flow, evaporation, soil moisture, groundwater level, etc. The original water conservancy data is obtained from long-term observation stations, remote sensing satellite data and historical records.

[0087] Determine the extent of the target watershed based on topographic data and hydrological characteristics.

[0088] The topographic and hydrological characteristics of the basin are extracted through DEM data, including basin area, river network density, slope and slope direction.

[0089] Combine meteorological data such as rainfall, temperature, and wind speed with the hydrological characteristics of the watershed to generate high-quality hydrometeorological data.

[0090] Step S2: Obtain the land use types in the watershed, such as farmland, forest, city, etc. These data can be obtained through remote sensing images and land use surveys. Obtain the soil types and characteristics in the watershed, such as soil texture, permeability, etc. Soil data can be obtained from soil survey reports and databases.

[0091] Topographic parameters include slope, aspect, river network density, etc. DEM data is used to analyze slope and aspect and extract river network morphological characteristics.

[0092] Through remote sensing image classification, the distribution and proportion of different land use types are determined.

[0093] Soil characteristic analysis: Analyze the soil's permeability, water retention and other characteristics, and evaluate the soil's impact on rainfall infiltration and runoff formation.

[0094] The terrain parameters, land use types and soil characteristic data are integrated to form a complete set of watershed characteristic parameters.

[0095] Step S3: Acquire facility information. Water conservancy facility data includes structural parameters and operation rules of facilities such as reservoirs, dams, and channels. The data source is obtained through engineering design drawings and operation records.

[0096] Facility operation data: Obtain the historical operation data of the facility, such as water level changes in the reservoir, flood discharge records, etc. These data will be used for parameter setting and calibration of the model.

[0097] Input the watershed characteristic parameter set and facility information into the WRF-Hydro model to initialize the model. Ensure that the model input parameters are accurate and comprehensive. Set the model's terrain parameters, such as slope, river network density, etc., based on the DEM data. Set the model's land use type and soil parameters based on the land use and soil characteristic data.

[0098] The model parameters were preliminarily calibrated using historical hydrological and meteorological data to ensure that the model could simulate the hydrological process of the basin. The model was initially run to verify the basic functions and calculation correctness of the model and obtain the initial flood model.

[0099] Step S4: Select calibration targets, such as flow, peak flow, flood duration, etc. Each target helps to evaluate the model performance in different aspects. According to the selected calibration target, define the corresponding objective function, such as minimizing flow error, peak error, etc.

[0100] Application of hydrometeorological data, using historical records of hydrometeorological data to ensure data integrity and representativeness. These data will be used as input for model calibration. Data from different time periods are calibrated to ensure the stability and applicability of the model under various meteorological conditions.

[0101] Adjust key parameters in the model, such as rainfall runoff coefficient, soil permeability, etc., to meet the multi-objective calibration goals. Intelligent algorithms such as genetic algorithms and particle swarm optimization can be used to optimize parameters and improve the accuracy of the model. Use independent validation data sets to verify the calibrated model to ensure the generalization ability of the model. Verify the accuracy and stability of the model by evaluating indicators such as the Nash efficiency coefficient (NSE) and root mean square error (RMSE) of the model.

[0102] Through the above process, the multi-objective parameter calibration of the initial flood model is completed to obtain the flood forecasting model.

[0103] Step S5: Design multiple rainfall scenarios, such as rainfall events of different intensities, durations, and spatial distributions. Consider factors that may affect land use changes and infrastructure construction in the basin, and design corresponding scenarios. Input the designed scenarios into the flood forecasting model and perform simulation runs. Record the simulation results for each scenario, including flood peak, flood volume, duration, etc.

[0104] The simulation results under different scenarios were compared, the response of the model to different rainfall and watershed conditions was analyzed, and the uncertainty of the model under different scenarios was evaluated to identify potential risks and response strategies to obtain scenario simulation results.

[0105] Step S6: Independent flood event condition setting, select typical flood events that occurred in the past as independent event conditions to evaluate the applicability and accuracy of the model. Define the key parameters of the event, such as rainfall, initial conditions of the basin, etc.

[0106] Comparison of simulation results with actual events: Compare the scenario simulation results with the actual data of independent flood events to identify deviations in model parameters. Based on the comparison results, adjust the model parameters to more accurately reflect the actual situation.

[0107] The optimized parameters are applied to update the flood forecasting model. The optimized model is validated with a new independent dataset.

[0108] The model's Nash efficiency coefficient (NSE), root mean square error (RMSE) and other indicators were evaluated again.

[0109] Analyze the adaptability of the optimized model under different watershed conditions.

[0110] The present invention provides a model construction method for flood forecasting of small water conservancy facilities. By conducting a comprehensive basin analysis and characteristic parameter analysis on the target basin of the small water conservancy facilities, and combining the facility information, a flood forecasting model can be constructed more accurately, thereby improving the accuracy of flood forecasting and providing a more reliable basis for disaster prevention and mitigation. By performing multi-objective parameter calibration processing on the initial flood model, the flood evolution process under different hydrological and meteorological conditions can be more effectively reflected, which is helpful to achieve accurate flood forecasting and avoid the risks caused by forecast errors. Based on a variety of rainfall scenarios and basin conditions, the flood forecasting model is simulated in a scenario, which can ensure that the model can operate efficiently under various extreme weather events, reduce the uncertainty of the forecast, and improve the effectiveness of disaster prevention and mitigation measures. By performing parameter optimization analysis on the scenario simulation results and optimizing the flood forecasting model according to the model optimization parameters, a more reasonable flood control scheduling strategy can be formulated, and the optimized operation of the entire flood control system can be achieved through a comprehensive adjustment strategy, thereby effectively utilizing water resources and reducing disaster losses and environmental impacts. At the same time, by considering the impact of basin characteristic parameters and facility information on flood forecasting, it is possible to flexibly adjust flood forecasting strategies according to the characteristics of different basins and changes in demand, making the flood control system more adaptable to diverse demand scenarios.

[0111] In one embodiment, the original water conservancy data of the target watershed of the small water conservancy facility is obtained, and the watershed analysis is performed to obtain the corresponding hydrological and meteorological data, including:

[0112] Obtain the original water conservancy data of the target basin. These data usually come from various hydrological monitoring points arranged in the basin, including basic hydrological elements such as rainfall, runoff, and water level. These monitoring points continuously record hydrological data through automated equipment to ensure the timeliness and accuracy of the data.

[0113] Based on the original water conservancy data obtained, the target basin is analyzed for hydrological elements. During the analysis, by comparing historical data, statistical analysis methods are used to identify the main hydrological elements in the basin, such as rainfall intensity, runoff coefficient, evaporation, etc. This step aims to determine the key factors affecting the hydrological process of the basin and provide a basis for subsequent analysis.

[0114] Based on the main hydrological elements of the basin, the basin rainfall-runoff analysis is carried out to simulate the runoff process after the rainfall event and generate a set of hydrological and meteorological parameters including rainfall, runoff volume, runoff time, etc. These parameter sets provide necessary data support for subsequent flood forecasting.

[0115] Conduct hydrological and meteorological information correlation analysis on the original water conservancy data. By constructing a hydrological and meteorological information database, the spatial and temporal correlation between hydrological elements such as rainfall, runoff, and evaporation in the basin and meteorological elements is comprehensively considered to extract the hydrological and meteorological information data of the basin. This step reveals the intrinsic connection between hydrological elements and meteorological elements through data mining technology.

[0116] Multi-dimensional data fusion processing is performed on the basin hydrological and meteorological information data and hydrological and meteorological parameter sets. This process uses data fusion techniques, such as Kalman filtering and principal component analysis, to comprehensively process multi-source data and extract the most representative hydrological and meteorological data.

[0117] This embodiment can accurately grasp the main hydrological elements in the basin by acquiring original water conservancy data and performing basin analysis on the target basin of small-scale water conservancy facilities, thereby providing a reliable data basis for flood forecasting. Using advanced hydrological models to conduct rainfall runoff analysis, the generated hydrological and meteorological parameter set can more accurately reflect the hydrological dynamics in the basin and improve the accuracy of the forecast. Through the correlation analysis of hydrological and meteorological information, the complex relationship between hydrological elements and meteorological elements is revealed, so that the forecast model can comprehensively consider a variety of influencing factors, thereby improving the comprehensiveness and accuracy of the forecast. Multi-dimensional data fusion processing further integrates multi-source data and model results, and the generated hydrological and meteorological data not only has high precision and high reliability, but also can adapt to different forecasting needs, effectively supporting flood control decisions of small-scale water conservancy facilities.

[0118] In one embodiment, geographic data is obtained and characteristic parameter analysis is performed to obtain a corresponding watershed characteristic parameter set, including:

[0119] Obtain geographic data of small water conservancy facilities and perform digital elevation processing to obtain terrain elevation data. Analyze the slope and aspect of the target basin based on the terrain elevation data to obtain slope data and aspect data. Extract the river network from the terrain elevation data to obtain river network morphology data. Calculate the river length and river density of the target basin based on the river network morphology data to obtain river length data and river density data.

[0120] The regional soil type data of the target watershed is obtained, and soil permeability analysis is performed to obtain soil permeability data. The terrain elevation data, slope data, slope direction data, river network morphology data, river length data, river density data and soil permeability data are integrated to obtain the watershed characteristic parameter set.

[0121] In the stage of geographic data acquisition and digital elevation processing, remote sensing technology and digital elevation model (DEM) data are used to ensure the accuracy of terrain elevation data. In the process of slope and aspect analysis, terrain elevation data is used to calculate through geographic information system (GIS) tools to generate accurate slope and aspect data.

[0122] In the river network extraction stage, the river network morphology is determined through terrain elevation data, and river length and river density data are calculated. In the process of soil type data acquisition and soil permeability analysis, field surveys and soil databases are combined to analyze soil permeability characteristics and generate soil permeability data.

[0123] All acquired data are integrated to form a watershed characteristic parameter set, providing basic data support for flood forecasting models.

[0124] This embodiment acquires geographic data of small-scale water conservancy facilities and performs characteristic parameter analysis to construct a flood forecasting model, thereby improving the accuracy and comprehensiveness of the data. The acquisition of terrain elevation data and digital elevation processing ensure the accuracy of terrain data through the application of remote sensing technology and digital elevation model (DEM) data. The use of geographic information system (GIS) tools for slope and aspect analysis provides more accurate slope and aspect data, which helps to more accurately describe the landform characteristics of the watershed. The above types of data are integrated to form a watershed characteristic parameter set, which provides complete and accurate data support for the flood forecasting model.

[0125] In one embodiment, the facility information is obtained, and the parameters of the preset WRF-Hydro model are set in combination with the watershed characteristic parameter set to obtain an initial flood model, including:

[0126] Obtain the structural parameters and historical operation data of small water conservancy facilities, including the height of the dam, water storage capacity, size and shape of the spillway, etc. These data are processed by integrating historical records and sensor data to obtain complete and detailed facility information.

[0127] Input the basin characteristic parameter set and the above-mentioned facility information into the preset WRF-Hydro model. The basin characteristic parameter set usually includes information such as topography, land use, soil type, and river network structure. This information can be obtained through geographic information system (GIS) and remote sensing technology and standardized. The parameter initialization processing of the model generates an initial characteristic parameter set by analyzing and calculating these data. The initial characteristic parameter set is the basis for the basic parameter training of the WRF-Hydro model.

[0128] When training basic parameters, the initial characteristic parameter set is processed by digital elevation model (DEM). DEM processing obtains the slope and basin area of ​​the basin, which are important components of basin characteristics. The river network density is analyzed, and the basin characteristic parameters of the WRF-Hydro model are trained based on the basin slope, basin area and river network density. The purpose of basin characteristic parameter training is to enable the model to accurately reflect the terrain and river network characteristics of the actual basin.

[0129] Generating meteorological parameter distribution maps is another work based on the initial feature parameter set. The meteorological parameter distribution map shows the meteorological conditions such as rainfall, temperature, wind speed, etc. in different areas of the basin. By statistically analyzing the historical data of these meteorological parameters and combining them with the current meteorological forecast data, the WRF-Hydro model is trained with meteorological parameters. The meteorological parameter training enables the model to accurately simulate the impact of future meteorological changes on the hydrological conditions of the basin.

[0130] The initial characteristic parameter set is analyzed for the river basin to obtain the river roughness coefficient. The river roughness coefficient reflects the resistance characteristics of the river and affects the flow velocity and water level. The corresponding reservoir dispatching rules are set and formulated based on historical data so that the reservoir's storage capacity can be reasonably used when floods come to reduce flood disasters.

[0131] The time step and spatial grid size of the model are trained based on the basin area, river network density and model computing capacity. The time step determines the calculation accuracy and efficiency of the model, and the spatial grid size affects the model's ability to describe the spatial heterogeneity of the basin. The training of these two parameters requires repeated trials and adjustments to find the best balance.

[0132] The inflow and outflow boundary analysis of the basin is carried out for the basin area and river network density. The purpose of boundary analysis is to determine the water inflow and outflow within the model calculation range. The boundary conditions are set according to the boundary parameters obtained by the analysis to ensure that the calculation results of the model at the boundary are consistent with the actual situation.

[0133] Through comprehensive processing and training of the above steps, the initial flood model is obtained.

[0134] This embodiment obtains the structural parameters and historical operation data of small-scale water conservancy facilities, and sets the parameters of the preset WRF-Hydro model in combination with the basin characteristic parameter set, which can accurately reflect the actual situation of the facilities and the basin and improve the accuracy of the model. By performing DEM processing on the initial characteristic parameter set and training the basin characteristic parameters, the model can accurately simulate the terrain and river network characteristics of the basin, and enhance the reliability of flood forecasting. Meteorological parameter training ensures that the model can dynamically reflect the impact of meteorological changes on the hydrological conditions of the basin, and improves the timeliness and accuracy of the forecast. The setting of the river roughness coefficient and the application of reservoir scheduling rules make the model closer to actual operation when simulating the flood process, and improve the scientificity and effectiveness of scheduling. The optimization of the time step and the spatial grid size helps to balance the calculation accuracy and efficiency of the model and ensure that the forecast results are timely and effective. The boundary condition setting ensures the overall consistency and accuracy of the model calculation through accurate inflow and outflow boundary analysis. Through the systematic model construction method, the initial flood model obtained can provide a reliable scientific basis for flood prevention and disaster reduction, and greatly improve the accuracy and practicality of flood forecasting.

[0135] In one embodiment, a multi-objective parameter calibration process is performed on the initial flood model according to the hydrological and meteorological data to obtain a flood forecasting model, including:

[0136] The spatial and temporal distribution of rainfall data in hydrological and meteorological data is analyzed to obtain the spatial and temporal distribution characteristics of rainfall. Through detailed analysis of rainfall data, the intensity, duration and spatial variability of rainfall are determined. These characteristics provide basic rainfall input information for flood forecasting models.

[0137] The flow process line analysis is performed on the flow data in the hydrological and meteorological data to obtain the flow process line characteristics. The flow process line analysis includes the analysis of flow change trends, peak flow and flow decay process. By mastering these characteristics, the dynamic changes of the flood process can be detected, providing a basis for the calibration of the model.

[0138] Soil moisture data in hydrological and meteorological data are analyzed to obtain soil moisture characteristics. Soil moisture analysis includes the analysis of parameters such as initial soil moisture state, rainfall infiltration rate, and soil saturation. By obtaining these characteristics, the soil's response to rainfall is evaluated, which in turn affects the formation and development of floods.

[0139] According to the above-mentioned rainfall spatiotemporal distribution characteristics, flow process line characteristics and soil moisture characteristics, the initial flood model is comprehensively calibrated to obtain a calibrated flood model. The comprehensive calibration process requires inputting various characteristic parameters into the model and adjusting the model parameters to make its output consistent with the actual observation data, thereby improving the accuracy of the model.

[0140] The preset standard flood data is compared with the calibration flood model to obtain the error results of the model. The error results reflect the degree of deviation of the calibration model in the process of simulating floods. According to the error results, the calibration flood model is adjusted for secondary parameters to further optimize the model parameters, reduce the error, and finally obtain the flood forecast model.

[0141] This embodiment improves the accuracy of flood forecasting by performing multi-objective parameter calibration on the initial flood model and accurately adjusting the model according to actual hydrological and meteorological data. The spatiotemporal distribution analysis of rainfall data can grasp the intensity and variability of rainfall in detail, and provide accurate rainfall input information for the model. The analysis of flow process lines can accurately reflect the flow change trend and peak flow, and enhance the model's dynamic simulation capability for flood processes. The analysis of soil moisture characteristics enables the model to consider the initial moisture state of the soil and the rainfall infiltration rate, thereby more accurately simulating the flood formation process. The comprehensive calibration process optimizes the model parameters through multiple iterations to ensure that the model output is highly consistent with the actual observation data, reduce errors, and improve the reliability of the forecast. Verification and secondary adjustment of the calibration model through standard flood data helps to further optimize the model and ensure its adaptability and stability under different conditions.

[0142] In one embodiment, a flood forecasting model is simulated based on a plurality of preset rainfall scenarios and watershed conditions to obtain scenario simulation results, including:

[0143] The flood forecasting model is simulated based on a variety of preset rainfall scenarios and watershed conditions. A variety of rainfall scenarios and watershed conditions are input into the flood forecasting model, and data formatting and tensor processing are performed through the model's input data processing layer. Data formatting refers to converting the input rainfall and watershed condition data into the format required by the model so that it can be understood and processed by the model. Tensor processing is to convert the formatted data into a tensor form for subsequent model calculations.

[0144] The generated input tensor is input into the rainfall-runoff conversion layer, which performs a coupling analysis of rainfall input and geographical and hydrological characteristics. Coupling analysis refers to combining rainfall input with the geographical and hydrological characteristics of the basin, taking into account the interaction between the two, and obtaining the runoff time series information. The runoff time series information refers to the change in the amount of runoff converted from rainfall in the basin at different time points.

[0145] The local time feature extraction sublayer performs local feature extraction analysis on the runoff time series information. Local feature extraction analysis refers to extracting representative and important local features from the runoff time series to obtain a local time feature tensor. The local time feature tensor refers to data in the form of a tensor that contains important local features in the runoff time series.

[0146] The local time characteristic tensor is input into the confluence simulation layer, and the water flow path in the basin is simulated through the confluence simulation layer. Water flow path simulation refers to simulating the flow path of water in the basin, thereby generating the flow time series tensor of the basin outlet. The flow time series tensor refers to the change of water flow at the basin outlet at different time points.

[0147] The overall confluence path analysis sublayer of the confluence simulation layer performs overall path analysis on the flow time series tensor. Overall path analysis refers to a comprehensive analysis of the flow path of the entire basin to obtain confluence characteristic information. Confluence characteristic information refers to information reflecting the overall path and characteristics of water flow in the basin.

[0148] The flow time series tensor is input into the hydrodynamic simulation layer, and the water level change simulation in the basin is performed through the hydrodynamic simulation layer. Water level change simulation refers to predicting the water level changes at different locations in the basin based on the flow time series to obtain the water level change time series. The water level change time series refers to the water level changes at various locations in the basin at different time points.

[0149] The dynamic response analysis sublayer of the hydrodynamic simulation layer extracts dynamic response features from the water level change time series. Dynamic response feature extraction refers to extracting information reflecting the dynamic characteristics of water level changes from the water level change time series to obtain water level change feature information. Water level change feature information refers to data reflecting the dynamic characteristics of water level changes.

[0150] The water level change time series is input into the flood process simulation layer, and the flood propagation process is simulated through the flood process simulation layer. Flood propagation process simulation refers to simulating the propagation and diffusion process of floods in the basin and generating flood inundation characteristic tensors. Flood inundation characteristic tensors refer to data reflecting the inundation situation of floods in the basin.

[0151] The regional inundation analysis sublayer of the flood process simulation layer identifies high-risk areas based on the flood inundation characteristic tensor. High-risk area identification refers to identifying high-risk areas in the basin that are susceptible to flooding based on flood inundation characteristics and obtaining inundation area information. Inundation area information refers to data reflecting the location and range of high-risk inundation areas in the basin.

[0152] Through the comprehensive analysis and result output layer of the flood forecasting model, the time series characteristic information of runoff, confluence characteristic information, water level change characteristic information and flooded area information are comprehensively analyzed. Comprehensive analysis refers to the comprehensive processing of the characteristic information obtained from the above layers to obtain the final scenario simulation results. The scenario simulation results refer to the predicted flood development and impact based on different rainfall scenarios and basin conditions.

[0153] This embodiment integrates multi-level simulation and analysis modules into the flood forecasting model, and can flexibly adjust the calculation process of the model under different rainfall scenarios and basin conditions. This design enables the model to process diverse input data without replacing different model structures, thereby improving the adaptability and practicality of the model. Through the feature extraction and analysis of modules at all levels, it is ensured that even under complex hydrological conditions, the key flood characteristics can be accurately captured, greatly improving the accuracy of the forecast. The combination of local and overall path analysis not only improves the understanding of the water flow dynamics in the basin, but also provides a more reliable basis for identifying high-risk flooding areas, thereby effectively supporting flood control decisions. At the same time, through the extraction and comprehensive analysis of dynamic response characteristics, the allocation of computing resources is optimized, the complexity of model operation is reduced, and flood forecasting is made more efficient and energy-saving.

[0154] In one embodiment, a parameter optimization analysis is performed on the scenario simulation results according to the preset independent flood event conditions to obtain corresponding model optimization parameters, and the flood forecasting model is optimized according to the model optimization parameters to obtain an optimized flood forecasting model, including:

[0155] The simulation results are screened based on the preset independent flood event conditions, which usually include key parameters such as the start time, peak flow, and duration of the flood. By screening these characteristic parameters, the characteristic parameters of flood events that meet the conditions can be extracted to provide basic data for subsequent analysis.

[0156] Based on the extracted characteristic parameters of flood events, parameter sensitivity analysis of flood forecasting model is carried out. The purpose of parameter sensitivity analysis is to find out the parameters that have the greatest impact on flood forecasting model and obtain parameter sensitivity ranking through analysis. This ranking helps to identify the key parameters that need to be adjusted during model optimization.

[0157] Based on the parameter sensitivity ranking, a multi-objective parameter optimization analysis is performed on the flood forecasting model to obtain the optimized parameter set. Multi-objective parameter optimization analysis involves balancing multiple objectives to achieve the best model performance. For example, a trade-off may be required between accuracy and computational efficiency. Through this multi-objective optimization analysis, a set of parameters that can comprehensively improve model performance can be determined.

[0158] Based on the optimized parameter set, the initial flood model is recalibrated. The recalibration process involves adjusting the parameters in the model to make it closer to the actual flood behavior, thereby improving the model's prediction accuracy. The recalibrated flood forecast model needs to be verified by independent flood event simulations to ensure its stability and reliability under different scenarios.

[0159] In the simulation verification phase, the recalibrated flood forecast model is simulated for independent flood events to obtain simulation verification results. The accuracy and reliability of the simulation verification results need to be evaluated through error analysis. Error analysis information can reveal the prediction errors and deviations of the model under different scenarios, thus providing a basis for further optimization.

[0160] According to the error analysis information, the optimization parameter set is adjusted accordingly to obtain the final model optimization parameters. The adjusted parameter set will further improve the prediction accuracy and stability of the model. Finally, according to the final model optimization parameters, the recalibrated flood forecasting model is finally optimized to obtain the optimized flood forecasting model.

[0161] This embodiment achieves accurate extraction and efficient forecasting of flood characteristics by gradually optimizing parameters and model calibration. Through feature screening and sensitivity analysis, key influencing factors are identified, so that resources can be concentrated to improve the accuracy of important parameters during model optimization, making the model more targeted and adaptable. Through multi-objective optimization analysis, different objectives are balanced to achieve the best combination of accuracy and efficiency, and error analysis and multiple calibration verifications are used to ensure the stability and reliability of the model in various flood scenarios. The optimization process of the model effectively reduces errors, significantly improves the accuracy of flood forecasts, can adapt to different flood event conditions in a short period of time, and achieve accurate early warning of small water conservancy facilities, which is conducive to the efficient management of water resources and the improvement of flood control safety.

[0162] Reference Figure 2 As shown, the present invention also provides a model construction system for flood forecasting of small water conservancy facilities, characterized in that the model construction method for flood forecasting of small water conservancy facilities applied to any one of the above items includes:

[0163] The acquisition module is used to obtain the original water conservancy data of the target watershed of the small water conservancy facilities, and conduct watershed analysis to obtain the corresponding hydrological and meteorological data;

[0164] Analysis module, which is used to obtain geographic data and perform characteristic parameter analysis to obtain a corresponding watershed characteristic parameter set;

[0165] The association module is used to obtain facility information and set parameters of the preset WRF-Hydro model in combination with the basin characteristic parameter set to obtain the initial flood model;

[0166] A processing module, which is used to perform multi-objective parameter calibration processing on the initial flood model according to the hydrological and meteorological data to obtain a flood forecasting model;

[0167] A control module, the control module is used to perform scenario simulation on the flood forecasting model according to a plurality of preset rainfall scenarios and watershed conditions to obtain scenario simulation results;

[0168] The execution module is used to perform parameter optimization analysis on the scenario simulation results according to the preset independent flood event conditions, obtain the corresponding model optimization parameters, optimize the flood forecasting model according to the model optimization parameters, and obtain the optimized flood forecasting model.

[0169] The present invention provides a model construction system for flood forecasting of small water conservancy facilities. By conducting a comprehensive basin analysis and characteristic parameter analysis on the target basin of the small water conservancy facilities, and combining the facility information, a flood forecasting model can be constructed more accurately, thereby improving the accuracy of flood forecasting and providing a more reliable basis for disaster prevention and mitigation. By performing multi-objective parameter calibration processing on the initial flood model, the flood evolution process under different hydrological and meteorological conditions can be more effectively reflected, which is helpful to achieve accurate flood forecasting and avoid the risks caused by forecast errors. Based on a variety of rainfall scenarios and basin conditions, the flood forecasting model is simulated in a scenario, which can ensure that the model can operate efficiently under various extreme weather events, reduce the uncertainty of the forecast, and improve the effectiveness of disaster prevention and mitigation measures. By performing parameter optimization analysis on the scenario simulation results and optimizing the flood forecasting model according to the model optimization parameters, a more reasonable flood control scheduling strategy can be formulated, and the optimized operation of the entire flood control system can be achieved through a comprehensive adjustment strategy, thereby effectively utilizing water resources and reducing disaster losses and environmental impacts. At the same time, by considering the impact of basin characteristic parameters and facility information on flood forecasting, it is possible to flexibly adjust flood forecasting strategies according to the characteristics of different basins and changes in demand, making the flood control system more adaptable to diverse demand scenarios.

[0170] Reference Figure 3 As shown, the present invention also provides a model building device for flood forecasting of small water conservancy facilities, comprising:

[0171] Memory, used to store programs;

[0172] The processor is used to execute the program to implement each step of any one of the above-mentioned model building methods for flood forecasting of small-scale water conservancy facilities.

[0173] In this embodiment, the processor and the memory may be connected via a bus or other means. The memory may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a read-only memory, a flash memory, a hard disk, or a solid-state drive. The processor may be a general-purpose processor, such as a central processing unit, a digital signal processor, an application-specific integrated circuit, or one or more integrated circuits configured to implement the embodiments of the present invention.

[0174] It should be noted that technicians in the relevant technical field can clearly understand that for the convenience and conciseness of description, the specific working process of the system and each module described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0175] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A model construction method for flood forecasting of small water conservancy facilities, characterized in that: include: Obtain the original water conservancy data of the target watershed of small water conservancy facilities, conduct watershed analysis, and obtain the corresponding hydrological and meteorological data; Obtain geographic data and perform characteristic parameter analysis to obtain the corresponding watershed characteristic parameter set; Acquire facility information, and set parameters of a preset WRF-Hydro model in combination with the watershed characteristic parameter set to obtain an initial flood model; Performing multi-objective parameter calibration processing on the initial flood model according to the hydrological and meteorological data to obtain a flood forecasting model; Performing scenario simulation on the flood forecasting model according to a plurality of preset rainfall scenarios and watershed conditions to obtain scenario simulation results; Performing parameter optimization analysis on the scenario simulation results according to the preset independent flood event conditions to obtain corresponding model optimization parameters, and optimizing the flood forecasting model according to the model optimization parameters to obtain the optimized flood forecasting model; The acquisition of geographic data and analysis of characteristic parameters to obtain a corresponding watershed characteristic parameter set includes: Acquire the geographic data of the small-scale water conservancy facilities, and perform digital elevation processing to obtain terrain elevation data; Analyzing the slope and aspect of the target watershed according to the terrain elevation data to obtain slope data and aspect data; Extracting river network from the terrain elevation data to obtain river network morphology data; Calculating the river length and river density of the target basin according to the river network morphology data to obtain river length data and river density data; Acquire regional soil type data of the target watershed, and perform soil permeability analysis to obtain soil permeability data; Integrate terrain elevation data, slope data, aspect data, river network morphology data, river length data, river density data and soil permeability data to obtain a watershed characteristic parameter set; The flood forecasting model is subjected to scenario simulation according to the preset multiple rainfall scenarios and watershed conditions to obtain scenario simulation results, including: Inputting the multiple rainfall scenarios and the basin conditions into the flood forecasting model, performing data formatting and tensorization processing through the input data processing layer of the flood forecasting model to generate an input tensor; Inputting the input tensor into a rainfall-runoff conversion layer, performing coupling analysis of rainfall input and geographical hydrological characteristics through the rainfall-runoff conversion layer, and obtaining the runoff time series information; Performing local feature extraction and analysis on the runoff time series information through the local time feature extraction sublayer of the rainfall runoff conversion layer to obtain a local time feature tensor; The local time characteristic tensor is input into the confluence simulation layer, and the water flow path in the basin is simulated through the confluence simulation layer to generate the flow time series tensor of the basin outlet; Performing overall path analysis on the flow time series tensor through the overall confluence path analysis sublayer of the confluence simulation layer to obtain confluence characteristic information; The flow time series tensor is input into the hydrodynamic simulation layer, and the water level change in the basin is simulated through the hydrodynamic simulation layer to obtain the water level change time series; Extracting dynamic response features of the water level change time series through the dynamic response analysis sublayer of the hydrodynamic simulation layer to obtain water level change feature information; Inputting the water level change time series into the flood process simulation layer, simulating the flood propagation process through the flood process simulation layer, and generating a flood inundation characteristic tensor; The flood process simulation layer uses the regional inundation analysis sublayer to identify high-risk areas of the flood inundation feature tensor to obtain inundation area information; Through the comprehensive analysis and result output layer of the flood forecasting model, the runoff time series information, the confluence characteristic information, the water level change characteristic information and the flooded area information are comprehensively analyzed to obtain the scenario simulation result.

2. The model construction method for flood forecasting of small water conservancy facilities according to claim 1 is characterized in that: The original water conservancy data of the target watershed of the small water conservancy facility is obtained, and the watershed analysis is performed to obtain the corresponding hydrological and meteorological data, including: Acquire the original water conservancy data of the target watershed, and perform hydrological element analysis on the target watershed to obtain the main hydrological elements of the watershed; Conduct basin rainfall-runoff analysis based on the main hydrological elements of the basin to obtain a set of hydrological and meteorological parameters; Performing hydrological and meteorological information correlation analysis on the original water conservancy data to obtain basin hydrological and meteorological information data; Multi-dimensional data fusion processing is performed on the basin hydrological and meteorological information data and the hydrological and meteorological parameter set to obtain the hydrological and meteorological data.

3. The model building method for flood forecasting of small water conservancy facilities according to claim 1 is characterized in that: The acquiring of facility information and setting of parameters of a preset WRF-Hydro model in combination with the basin characteristic parameter set to obtain an initial flood model include: Obtaining structural parameters and historical operation data of small water conservancy facilities, and integrating and processing them to obtain the facility information; Inputting the watershed characteristic parameter set and the facility information into a preset WRF-Hydro model, and performing model parameter initialization processing to obtain a corresponding initial characteristic parameter set; Performing basic parameter training on the WRF-Hydro model according to the initial characteristic parameter set to obtain the initial flood model; Among them, basic parameter training includes: Performing DEM processing on the initial characteristic parameter set to obtain the watershed slope and watershed area, analyzing the river network density, and training the watershed characteristic parameters of the WRF-Hydro model based on the watershed slope, the watershed area and the river network density; Generating a meteorological parameter distribution map according to the initial characteristic parameter set, and performing meteorological parameter training; Performing river basin analysis on the initial characteristic parameter set to obtain a river roughness coefficient and setting corresponding reservoir dispatching rules; Training the time step and spatial grid size of the model according to the basin area, the river network density and the model computing capacity; The inflow and outflow boundaries of the basin are analyzed for the basin area and the river network density, and boundary conditions are set based on the boundary parameters obtained from the analysis.

4. The model building method for flood forecasting of small water conservancy facilities according to claim 1 is characterized in that: The multi-objective parameter calibration process is performed on the initial flood model according to the hydrological and meteorological data to obtain a flood forecast model, including: Performing a spatiotemporal distribution analysis on the rainfall data in the hydrological and meteorological data to obtain the spatiotemporal distribution characteristics of rainfall; Performing flow process line analysis on the flow data in the hydrological and meteorological data to obtain flow process line characteristics; Performing soil moisture analysis on the soil moisture data in the hydrological and meteorological data to obtain soil moisture characteristics; Comprehensively calibrating the initial flood model according to the temporal and spatial distribution characteristics of rainfall, the flow process line characteristics and the soil moisture characteristics to obtain a calibrated flood model; Comparing the preset standard flood data with the calibration flood model to obtain the error result of the model; The calibration flood model is subjected to secondary parameter adjustment according to the error result to obtain the flood forecast model.

5. The model building method for flood forecasting of small water conservancy facilities according to claim 1 is characterized in that: The method comprises: performing parameter optimization analysis on the scenario simulation results according to the preset independent flood event conditions to obtain corresponding model optimization parameters, optimizing the flood forecast model according to the model optimization parameters to obtain the optimized flood forecast model, including: Performing feature screening on the simulation results according to preset independent flood event conditions to obtain flood event feature parameters; Performing parameter sensitivity analysis on the flood forecasting model according to the characteristic parameters of the flood event to obtain a parameter sensitivity ranking; Performing multi-objective parameter optimization analysis on the flood forecasting model according to parameter sensitivity ranking to obtain an optimized parameter set; Recalibrating the initial flood model according to the optimized parameter set to obtain the recalibrated flood forecast model; Conducting independent flood event simulation verification on the recalibrated flood forecasting model to obtain simulation verification results; Perform error analysis on the simulation verification results to obtain error analysis information; Adjusting the optimization parameter set according to the error analysis information to obtain final model optimization parameters; The recalibrated flood forecast model is finally optimized according to the final model optimization parameters to obtain the optimized flood forecast model.

6. A model building system for flood forecasting of small water conservancy facilities, characterized in that: A model construction method for flood forecasting of small water conservancy facilities applied to any one of claims 1 to 5 above, comprising: A collection module, which is used to obtain the original water conservancy data of the target watershed of the small water conservancy facility, and perform watershed analysis to obtain the corresponding hydrological and meteorological data; An analysis module, which is used to obtain geographic data and perform characteristic parameter analysis to obtain a corresponding watershed characteristic parameter set; An association module, the association module is used to obtain facility information, and set parameters of a preset WRF-Hydro model in combination with the basin characteristic parameter set to obtain an initial flood model; A processing module, the processing module is used to perform multi-objective parameter calibration processing on the initial flood model according to the hydrological and meteorological data to obtain a flood forecast model; A control module, the control module is used to perform scenario simulation on the flood forecasting model according to a plurality of preset rainfall scenarios and watershed conditions to obtain scenario simulation results; An execution module is used to perform parameter optimization analysis on the scenario simulation results according to preset independent flood event conditions to obtain corresponding model optimization parameters, and optimize the flood forecasting model according to the model optimization parameters to obtain the optimized flood forecasting model.

7. A model building device for flood forecasting of small water conservancy facilities, characterized in that: include: Memory, used to store programs; A processor is used to execute the program to implement the various steps of the model building method for flood forecasting of small-scale water conservancy facilities as described in any one of claims 1-5.

Images