Partition flood prevention early warning method and system fusing multi-source data

Abstract

The invention provides a partition flood control early warning method and system fusing multi-source data, and relates to the technical field of flood control early warning, and the method comprises the steps: obtaining a multi-source monitoring data set in a target drainage basin; preprocessing the multi-source monitoring data set to obtain a standardized data set; respectively constructing a flood routing feature sub-model of each partition in the target drainage basin based on the standardized data set; carrying out coupling calculation on the output result of the flood routing characteristic sub-model of each partition by adopting a lightweight physical-empirical hybrid model to obtain a coupling calculation result; analyzing the historical flood case data based on a machine learning algorithm, and generating a dynamic early warning threshold adaptive to the flood routing feature sub-model of each partition; and generating partition early warning information based on the coupling calculation result and the dynamic early warning threshold. According to the method, multi-source data can be fused, and the threshold value is dynamically adjusted and efficiently calculated according to partition characteristics.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of flood warning, and in particular to a partitioned flood warning method and system fusing multi-source data. BACKGROUND

[0002] Flood disaster is one of the main natural disasters threatening the safety of people's lives and property in the basin. Precise and timely flood warning is the key to reducing disaster losses. Some traditional flood warning methods rely on fixed site monitoring data, such as rain stations and water level gauges, resulting in limited data coverage and insufficient timeliness. In addition, the flood evolution laws of different geographical units in the basin, such as mountainous areas, alluvial fans and plains, are significantly different, and some flood warning methods use uniform parameters for calculation, which cannot reflect the characteristics of each partition. And the warning threshold in the traditional method is based on historical extreme value or experience, which cannot be dynamically adjusted with real-time hydrological conditions and weather changes, and is prone to false positives or false negatives. SUMMARY

[0003] To overcome the problems in the related art, the present application provides a partitioned flood warning method and system fusing multi-source data, which is efficient in calculating and dynamically adjusting the threshold for the characteristics of each partition, in order to improve the warning accuracy and timeliness.

[0004] According to a first aspect of an embodiment of the present application, a partitioned flood warning method fusing multi-source data is provided, comprising: obtaining a multi-source monitoring data set in a target basin, wherein the multi-source monitoring data set includes rain station monitoring data, snow depth monitoring device data, flow meter data, water level gauge data, Beidou positioning mobile monitoring device data and satellite remote sensing data; preprocessing the multi-source monitoring data set to obtain a standardized data set; based on the standardized data set, constructing a flood evolution feature sub-model for each partition in the target basin; wherein the partitions include mountainous areas, alluvial fans and plains; using a lightweight physical-empirical hybrid model to couple and calculate the output results of the flood evolution feature sub-models of each partition to obtain a coupling calculation result; based on a machine learning algorithm to analyze historical flood case data, generating a dynamic warning threshold adapted to the flood evolution feature sub-model of each partition; based on the coupling calculation result and the dynamic warning threshold, generating partition warning information.

[0005] In some example embodiments of the present application, based on the foregoing scheme, based on the standardized data set, a flood evolution feature sub-model for each partition in the target basin is constructed, comprising: Screening geographical parameters, hydrological characteristics and underlying surface attribute characteristics unique to each subregion from the standardized data set, and establishing a subregion characteristic vector library; Based on the subregion characteristic vector library and historical flood process data of each subregion, the initial submodel of each subregion is trained, and a dynamic correction module is embedded to obtain a flood evolution characteristic submodel of each subregion.

[0006] In some example embodiments of the present application, based on the foregoing scheme, a light physical-empirical hybrid model is used to couple and calculate the output results of the flood evolution characteristic submodel of each subregion, to obtain coupling calculation results including: Based on the basic law of basin hydrodynamics, a water exchange equation between subregions is constructed; The output results of each submodel are taken as the boundary input of the water exchange equation between subregions, and the hydraulic connection relationship of adjacent subregions is solved by simplifying the Saint-Venant equation set to obtain a preliminary coupled flood propagation process; An empirical coefficient extracted from historical flood processes is introduced, and a machine learning method is used to correct the deviation of the preliminary coupled flood propagation process to output a corrected subregion flood hydrograph; For the output results of each submodel of different time scales and spatial scales, a spatiotemporal interpolation algorithm is used for scale unification to generate a consistent flood evolution spatiotemporal distribution data of the whole basin, and the consistent flood evolution spatiotemporal distribution data of the whole basin is taken as the coupling calculation result output; Among them, through the Monte Carlo simulation method, the uncertainty in the coupling calculation process is transmitted and analyzed, and the confidence interval and probability distribution of the coupling calculation result are output to quantify the uncertainty in the coupling process.

[0007] In some example embodiments of the present application, based on the foregoing scheme, the dynamic early warning threshold of the flood evolution characteristic submodel of each subregion is generated based on machine learning algorithm analysis of historical flood case data, including: Extracting key influencing factors of each subregion from historical flood case data to construct a subregion feature matrix; According to the actual disaster degree of each subregion in the historical flood event, and combining with the flood control standard specification, labeling the corresponding early warning level label for each historical case; Using gradient boosting tree, random forest or deep learning model, taking the subregion feature matrix as input and the early warning level label as output to train the model, to obtain the trained model; Embedding a real-time update module in the trained model, taking the current hydrological monitoring data, short-term weather forecast data and underlying surface change information as dynamic input, and outputting the specific early warning threshold of each subregion under the current situation by the model; The dynamic early warning threshold generated by the model is verified in combination with historical extreme events and expert experience, to obtain a dynamic early warning threshold of the flood evolution characteristic sub-model adapted to each sub-region.

[0008] In some example embodiments of the present application, based on the foregoing scheme, the sub-region early warning information comprises one or more combinations of sub-region identification, early warning level, predicted risk duration, predicted flood arrival time, impact range, and risk avoidance suggestion.

[0009] In some example embodiments of the present application, based on the foregoing scheme, based on the coupling calculation result and the dynamic early warning threshold, generating sub-region early warning information comprises: When the coupling calculation result is greater than 50% of the dynamic early warning threshold, early warning information comprising early warning level, impact range, predicted risk duration, and risk avoidance suggestion is generated; When the coupling calculation result is greater than 20% of the dynamic early warning threshold or when the coupling calculation result is less than 20% of the dynamic early warning threshold, early warning information comprising sub-region identification, early warning level, predicted flood arrival time, impact range, and risk avoidance suggestion is generated; when the coupling calculation result is less than 20% of the dynamic early warning threshold, early warning information comprising sub-region identification, early warning level, and impact range is generated.

[0010] In some example embodiments of the present application, based on the foregoing scheme, after generating sub-region early warning information based on the coupling calculation result and the dynamic early warning threshold, the sub-region flood control early warning method of fusing multi-source data further comprises: Publishing the sub-region early warning information through at least two of the following: short message, broadcast, application program push, and display screen.

[0011] According to a second aspect of an embodiment of the present application, a sub-region flood control early warning system of fusing multi-source data is provided, comprising: A data acquisition module is configured to acquire a multi-source monitoring data set in a target river basin, wherein the multi-source monitoring data set comprises rain gauge monitoring data, snow depth monitoring device data, flow meter data, water level gauge data, Beidou positioning mobile monitoring device data, and satellite remote sensing data. A data preprocessing module is configured to preprocess the multi-source monitoring data set to obtain a standardized data set. A sub-model construction module is configured to construct a flood evolution characteristic sub-model for each sub-region in the target river basin based on the standardized data set; wherein the sub-regions include mountainous areas, alluvial fans, and plains. A coupling calculation module is configured to perform coupling calculation on the output results of the flood evolution characteristic sub-models for each sub-region using a lightweight physical-empirical hybrid model to obtain a coupling calculation result. A dynamic threshold generation module is configured to generate a dynamic early warning threshold of a flood evolution feature sub-model of each subregion based on machine learning algorithm analysis of historical flood case data; An early warning information generation module is configured to generate subregion early warning information based on the coupling calculation result and the dynamic early warning threshold.

[0012] According to a third aspect of the embodiments of the present application, an electronic device is provided, comprising a processor and a memory having computer readable instructions stored thereon, the computer readable instructions being executed by the processor to implement the method in the first aspect.

[0013] According to a fourth aspect of the embodiments of the present application, a computer readable storage medium having a computer program stored thereon is provided, the computer program being executed by a processor to implement the method in the first aspect.

[0014] The technical solutions provided by the embodiments of the present application can include the following beneficial effects: In the embodiments of the present application, on the one hand, in view of the differences in flood evolution laws of mountainous areas, alluvial fans and plains in a river basin, the flood evolution feature sub-models of each subregion are customized based on the unique geographical parameters, hydrological characteristics and underlying surface properties of each subregion, which can avoid the simulation deviation caused by the lack of adaptability of the traditional unified model, and can more accurately depict the flood generation and propagation process in different regions. On the other hand, by using a lightweight physical-empirical hybrid model, the water exchange relationship between subregions is constructed based on the law of water dynamics, and then the output results of the subregion sub-models are efficiently coupled by combining empirical correction and space-time scale unification to form global coherent flood evolution data. This coupling method not only retains the accuracy of the subregion model, but also solves the problem of each subregion being independent, while greatly reducing the computational complexity and meeting the efficiency requirements of real-time early warning.

[0015] Secondly, by comparing the coupling results of the subregions with the dynamic early warning thresholds adapted to each subregion, the early warning level, impact range, risk time and targeted risk avoidance suggestions for each subregion are determined, which can directly match the emergency disposal needs of different regions, so that the management department and the affected population can quickly obtain key information, and the accuracy and efficiency of disaster response are improved.

[0016] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0017] The accompanying drawings, which are incorporated into the specification and constitute a part of the present application, illustrate embodiments consistent with the present application and, together with the specification, serve to explain the principles of the present application.

[0018] Figure 1A schematic diagram of a system architecture of an exemplary application environment of a partitioned flood warning method and system fusing multi-source data to which embodiments of the present application can be applied is shown. Figure 2 A flowchart of a partitioned flood warning method fusing multi-source data according to some embodiments of the present application is schematically shown. Figure 3 A schematic diagram of a partitioned flood warning system fusing multi-source data according to some embodiments of the present application is schematically shown. Figure 4 A structural schematic diagram of a computer system of an electronic device according to some embodiments of the present application is schematically shown. Figure 5 A schematic diagram of a computer readable storage medium according to some embodiments of the present application is schematically shown. DETAILED DESCRIPTION

[0019] The exemplary embodiments will be described in detail herein with reference to the attached drawings. The description of the exemplary embodiments is intended to apply to all alternative embodiments, unless otherwise indicated. The following description is not meant to limit the application to any one or more particular embodiments, but to provide examples and teaching to enable others skilled in the art to make and use the application as defined by the appended claims. Therefore, to the extent that there is any conflict between what is described in the disclosure and what is disclosed in the appended claims, the latter will control.

[0020] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0021] It is to be understood that the singular forms "a", "an", and "the" include plural referents unless the context clearly dictates otherwise. It is to be understood that the term "and / or" as used herein encompasses all possible combinations of particular items listed apart from disjunctively worded limitations of various claims. It is further to be understood that the use of "approximately", "substantially", or "about" in describing the embodiments of the application are intended to mean within 10% of the value stated.

[0022] Figure 1 A schematic diagram of a system architecture of an exemplary application environment of a partitioned flood warning method and system fusing multi-source data to which embodiments of the present application can be applied is shown.

[0023] As Figure 1As shown, the system architecture 100 can include one or more of terminal devices, such as a desktop computer 101, a portable computer 102, a smart phone 103, etc., a network 104 and a server 105. The network 104 is a medium for providing a communication link between the terminal devices and the server 105. The network 104 can include various connection types, such as wired, wireless communication links or fiber optic cables, etc. The terminal devices can be various electronic devices with data processing functions, which have a display screen for showing the warning information of each sub-area in the target flow field to the user, including but not limited to the desktop computer, the portable computer, the smart phone, etc. It should be understood that Figure 1 The number of terminal devices, networks and servers in the system architecture 100 is only illustrative. According to the implementation needs, there can be any number of terminal devices, networks and servers. For example, the server 105 can be a sub-server cluster composed of multiple sub-servers, etc.

[0024] The sub-area flood control warning method for fusing multi-source data provided by the embodiments of the present application can generally be executed by the terminal device, and accordingly, the sub-area flood control warning device for fusing multi-source data is generally arranged in the terminal device. However, it is easily understood by those skilled in the art that the sub-area flood control warning method for fusing multi-source data provided by the embodiments of the present application can also be executed by the server 105, and accordingly, the sub-area flood control warning device for fusing multi-source data can also be arranged in the server 105, which is not specially limited in the present exemplary embodiment.

[0025] In addition, it should be understood that the sub-area flood control warning method for fusing multi-source data of the present embodiment can be configured as a software module. In some implementation scenarios, the sub-area flood control warning method for fusing multi-source data of the present application can be deployed separately to be able to display the warning information of each sub-area in different target flow fields, or to display the warning information of each sub-area in the target flow field individually. In other implementation scenarios, the sub-area flood control warning method for fusing multi-source data of the present application can be deployed in other software as a functional module of the software, such as being deployed in the analysis software of underground pipelines, and the application mode of the sub-area flood control warning method for fusing multi-source data of the present application is not specially limited.

[0026] Next, the embodiments of the present application will be described in detail.

[0027] As shown in Figure 2 As shown in Figure 2 is a flow chart of a sub-area flood control warning method for fusing multi-source data according to an exemplary embodiment of the present application, which includes the following steps: S210: Obtain a multi-source monitoring data set in a target basin, wherein the multi-source monitoring data set comprises rainfall station monitoring data, snow depth monitoring device data, flow meter data, water level gauge data, Beidou positioning mobile monitoring device data and satellite remote sensing data; S220: Preprocess the multi-source monitoring data set to obtain a standardized data set; S230: Based on the standardized data set, a flood evolution feature sub-model for each partition in the target basin is constructed respectively; wherein the partitions include mountainous areas, alluvial fans and plains; S240: A light physical-empirical hybrid model is used to couple and calculate the output results of the flood evolution feature sub-models of each partition to obtain coupling calculation results; S250: Based on machine learning algorithm analysis of historical flood case data, a dynamic early warning threshold suitable for the flood evolution feature sub-model of each partition is generated; S260: Based on the coupling calculation results and the dynamic early warning threshold, partition early warning information is generated.

[0028] In the embodiment of the application, on the one hand, in view of the differences in flood evolution rules of mountainous areas, alluvial fans and plains in the basin, the flood evolution feature sub-models of each partition are customized based on the unique geographical parameters, hydrological characteristics and underlying surface properties of each partition, which can avoid the simulation deviation caused by the lack of adaptability of traditional unified models, and can more accurately depict the flood generation and propagation process in different regions. On the other hand, through the light physical-empirical hybrid model, the water exchange relationship between partitions is first constructed based on hydrodynamic rules, and then the output results of each partition sub-model are efficiently coupled by combining empirical correction and space-time scale unification to form coherent flood evolution data in the whole region. This coupling method not only retains the accuracy of the partition model, but also solves the problem of each partition being independent, while greatly reducing the computational complexity and meeting the efficiency requirements of real-time early warning.

[0029] Secondly, based on the comparison between the partition coupling results and the dynamic early warning threshold adapted to each partition, the early warning level, impact range, risk time and targeted risk avoidance suggestions for each partition are determined, which can directly match the emergency disposal needs of different regions, so that the management department and the affected population can quickly obtain key information, and the accuracy and efficiency of disaster response are improved.

[0030] In S210, a multi-source monitoring data set in a target basin is obtained, wherein the multi-source monitoring data set comprises rainfall station monitoring data, snow depth monitoring device data, flow meter data, water level gauge data, Beidou positioning mobile monitoring device data and satellite remote sensing data.

[0031] The target basin refers to an independent geographical unit with a complete hydrological cycle system, which is designated as the core research or monitoring object in specific research and application scenarios such as hydrology, water resources management, ecological protection, or disaster prevention and control. Its boundary is usually divided by topography and geomorphology, such as mountains, plains, and valleys, and its internal structure covers the complete hydrological process of precipitation, runoff, confluence, and discharge, including the main stream, tributaries, reservoirs, lakes, groundwater systems, and corresponding land areas.

[0032] The division of the target basin needs to be combined with specific application requirements. For example, if it is used for mountain flood disaster warning, the target basin may be a small river and its catchment area prone to mountain floods; if it is used for water resources allocation, the target basin may be the catchment area of a reservoir; if it is used for ecological protection, the target basin may be the habitat of a rare aquatic organism. Its core feature is to have a clear spatial range and complete hydrological function, and it is the basic spatial carrier for multi-source monitoring, data integration, and subsequent analysis and application.

[0033] In the embodiments of the present application, the basin is first divided into several monitoring grids based on the topography and geomorphology, hydrological units, and climate zones of the target basin using GIS spatial analysis techniques, ensuring that each grid covers at least one core monitoring indicator. The core monitoring indicators include rainfall station monitoring data, snow depth monitoring device data, flow meter data, water level gauge data, Beidou positioning mobile monitoring device data, and satellite remote sensing data, avoiding data collection blind spots.

[0034] Rainfall station monitoring data can be obtained by setting a rain gauge in the rainfall station. For example, in some embodiments, a fixed tipping bucket rain gauge can be set up according to the rainfall distribution in the basin (such as rainy areas and dry areas), topographic features (such as open areas and avoiding building / tree cover), and key areas (such as mountain flood-prone areas and reservoir upstream areas) to ensure coverage without blind spots.

[0035] The rain gauge automatically measures rainfall by the number of tipping bucket turns, with a sampling frequency of 10-30 minutes / turn, and simultaneously records the timestamp, rainfall, and device status. It uses 4G / 5G wireless transmission or LoRa low-power transmission technology to upload real-time collected data to the data center of the target basin.

[0036] Snow depth monitoring device data can be obtained by ultrasonic snow depth monitors arranged in the snow accumulation area of the target basin. For example, in some embodiments, for areas with an altitude of ≥ 2500 m or continuous snow in winter, ultrasonic snow depth monitors are installed on stable supports 1.5-2 m from the ground to avoid the influence of ground freeze-thaw on measurement accuracy; the distance from the ground / snow surface to the sensor is measured by ultrasonic ranging principle, and the snow depth is calculated, and the snow surface temperature is collected synchronously, and the sampling frequency can be set to 1-2 hours / time; finally, the data is uploaded to the data center in real time by LoRa low-power transmission (adapted to high-altitude signal weak environment), ensuring data continuity during snowmelt and snowfall periods.

[0037] Flow meter data can be obtained by fixed Doppler ultrasonic flow meters arranged at key sections of the target basin river. In some embodiments, representative sections of the main stream and main tributaries of the basin can be selected, such as the outlet of the basin, the inlet and outlet of the reservoir, and the water intake of the town, and the fixed Doppler ultrasonic flow meter is installed at the thalweg position of the section, i.e. the position with the most turbulent flow and the most representative flow speed, to ensure that the sensor is completely submerged in water and avoids silt accumulation areas; the water flow speed is measured by Doppler effect, and the instantaneous flow is calculated by combining the section area, and the average flow speed of the section is recorded synchronously, and the sampling frequency can be set to 5-10 minutes / time; finally, the data is transmitted to the data acquisition terminal at the section site by wired cable, and then uploaded to the basin data center by the terminal through 4G / 5G, avoiding the influence of wireless signal obstruction by the surrounding terrain of the river, thereby ensuring the stability of data transmission.

[0038] Water level gauge data can be obtained by fixed and submerged static pressure water level gauges arranged in rivers, reservoirs and other water bodies in the target basin. In some embodiments, the water level gauge can be submerged to a fixed depth in the water on the stable bedrock or concrete platform on the left / right bank of the river section, with the Yellow Sea elevation or local datum elevation as the reference datum, and preferably arranged on the same side of the flow meter to facilitate the spatio-temporal matching of flow and water level data; the water level elevation is calculated by measuring the static pressure of the water body, and the sampling frequency is set to 1-5 minutes / time, and high-frequency collection is used to capture the process of sudden rise and fall of water level, such as during flood period; the data acquisition terminal is shared with the flow meter, and the data is uploaded to the data center together with the flow data after being transmitted to the terminal by wired transmission, realizing the synchronous correlation of water level and flow data.

[0039] The Beidou positioning mobile monitoring equipment data can be carried by a UAV or a person to carry a mobile monitoring terminal integrated with a Beidou dual-mode positioning system to obtain in a target basin. For example, in some embodiments, the mobile terminal is integrated with a Beidou satellite navigation system + global positioning system dual-mode positioning module and a portable hydrological sensor. The portable hydrological sensor, such as a handheld rain gauge and a small flowmeter, has an IP68 waterproof level to adapt to rainy days and water operations and has a storage capacity of greater than or equal to 16 GB. For areas where fixed equipment coverage is insufficient, such as remote tributaries and temporary engineering areas, 2-3 mobile monitoring routes are planned per month to ensure coverage of all monitoring grids without fixed equipment control. The terminal is stopped at the center point and the edge point of the river in each monitoring grid along the route, the latitude and longitude and the altitude of the sampling point are obtained through Beidou positioning, the instantaneous rainfall (when it rains) and the near-surface humidity are collected through the integrated sensor, and the surface coverage type (such as farmland and forest land) is manually recorded. After daily monitoring, the terminal data is uploaded to the data center, and if there is offline data, it is supplemented through a USB data line to ensure that there is no data loss.

[0040] Satellite remote sensing data is obtained by screening and downloading remote sensing image data of multiple types of satellites after preprocessing. In some embodiments, different satellite data can be selected according to monitoring needs, such as optical remote sensing (Landsat-9, Sentinel-2, etc., for obtaining surface coverage and vegetation coverage), microwave remote sensing (Sentinel-1, etc., for obtaining snow cover area and water body boundaries in cloudy and rainy weather), and hydrological special remote sensing (GRACE-FO, etc., for obtaining changes in underground water storage); through official platforms, image data covering the target basin and having no or little cloud cover is automatically downloaded or manually screened; preliminary preprocessing is performed using radiation calibration, atmospheric correction, geometric correction, and basin cropping to form a remote sensing data set.

[0041] The present application integrates multiple types of data such as rain gauge stations, snow depth equipment, Beidou mobile monitoring, and satellite remote sensing, covers ground fixed, mobile patrol, and space remote sensing full-scene monitoring dimensions, and can effectively fill in the monitoring blanks in remote areas and complex terrains; at the same time, data preprocessing is performed to realize format unification and quality optimization, avoid calculation deviation caused by data confusion or abnormality, and provide more comprehensive and reliable data source support for subsequent modeling and analysis.

[0042] S220: Preprocessing the multi-source monitoring data set to obtain a standardized data set; The preprocessing of the multi-source monitoring data set includes, but is not limited to, outlier rejection, missing value completion, and format standardization processing of the multi-source monitoring data set.

[0043] The outlier elimination needs to be targeted at the characteristics of different data, combined with the target basin, such as the geographical environment and physical laws of the mountainous area, alluvial fan and plain in Xinjiang region, and adopts a three-step process of preliminary identification, secondary verification and final elimination to avoid deleting effective data. The preliminary identification can use the Pandas library of Python or the statistical analysis tool of MATLAB to calculate the mean μ and standard deviation σ of single-class numerical data (such as the hourly rainfall data of a certain rainfall station), and mark the data exceeding the range of [μ-3σ, μ+3σ] as suspected outliers. For example, the μ of the hourly rainfall data of a certain mountainous rainfall station in July is 8 mm, and the σ is 5 mm, so the normal range is [-7 mm, 23 mm], and the data with hourly rainfall > 23 mm or 0 mm (rainfall has no negative value) is marked as suspected outliers.

[0044] The secondary verification process can combine the historical extreme value of the target basin and the physical law to develop the domain threshold of each data type, such as by querying the historical data of the basin hydrology yearbook or the local water conservancy department. The secondary verification process can be to compare the suspected outliers identified in the preliminary identification with the domain threshold. If both exceed the 3σ range and the domain threshold (such as hourly rainfall = 55 mm, which exceeds both [μ-3σ, μ+3σ] and the domain threshold of 50 mm in mountainous area), it is marked as a confirmed outlier. In some embodiments, the domain threshold can include: The maximum single-hour rainfall in Xinjiang mountainous area is usually not more than 50 mm, the maximum single-day snow depth increment in plain area is not more than 15 cm; the maximum range of flow meter in alluvial fan area is 500 m³ / s, and the flow data exceeding the range can be directly determined as abnormal.

[0045] Finally, the confirmed outliers are eliminated.

[0046] The operation process of other data can be similar to that of rainfall data, with the difference being the 3σ range and the domain threshold, which can be calculated and set by those skilled in the art according to the actual situation, and the present application does not make specific limitations.

[0047] The missing value completion can be processed according to the classification of missing scenarios. In this process, the missing situation can be first classified, such as by missing duration: short time ≤ 1 hour, medium time 1-24 hours, long time > 24 hours, or by data type: continuous numerical value, discrete position, image pixel, etc., and then combined with the data application scenario to select the adaptive method to ensure that the deviation rate of the completed data from the actual situation is less than or equal to the preset threshold.

[0048] The preset threshold can be 3% to 8%, which can be set according to the actual situation, and the present application does not make limitations.

[0049] In some embodiments, the same regional and same type data can be used to complete the target missing data first. If the same regional and same type data is insufficient, the detection data under similar conditions in the same period can be combined to build a random forest regression model through the Python Scikit-learn library, and the detection data under similar conditions in the same period is input to predict and complete. After completion, when the equipment resumes data collection (e.g., Beidou equipment re-uploads data), compare the completed value with the actual value, calculate the deviation rate; if the deviation rate > 8%, adjust the model parameters (such as increasing the number of decision trees of the random forest model) or replace the reference data source, and re-complete to ensure the reliability of the completed data.

[0050] Format standardization processing needs to solve the format difference, dimension difference, and space-time difference of different equipment data. In some embodiments, for numerical data conversion, the TXT format of the rain gauge and the Excel format of the flow meter data are uniformly converted to CSV format. In Python, the to_csv() function of the Pandas library is used to set the fields: equipment number (such as YL-001, YL represents rain gauge), monitoring time (UTC+8, format "YYYY-MM-DD HH:MM:SS"), data type (such as "hourly rainfall"), numerical value (such as "12.5"), unit (such as "mm"), and data quality identifier (such as "valid", "completed", "after exception removal").

[0051] For spatial data conversion, the JSON format position data of the Beidou positioning equipment is converted to GeoJSON format. In QGIS software, the JSON data is imported, the coordinate system is set to WGS84 (EPSG: 4326), the spatial feature type (point feature) is defined, and the attribute fields are added: equipment number (BD-008), monitoring time, position coordinates (longitude and latitude), and data quality identifier. The GeoJSON file is output.

[0052] For remote sensing image data conversion, different satellites (such as Landsat-8, Sentinel-2) remote sensing images are uniformly retained in TIFF format. In ENVI software, radiation calibration, atmospheric correction, and uniform resolution (terrain data 10m, land cover data 30m) are performed. Metadata such as shooting time, sensor type, resolution, and correction method are added to ensure that image data can be directly used for subsequent terrain analysis.

[0053] After that, the converted data is time-dimensioned and physically unitized. For time-dimensioning, the rain station data 5 minutes / time, the water level meter 10 minutes / time, and the satellite remote sensing 1 day / time can be unified to 1 hour / time by using high-frequency downsampling and low-frequency upsampling. For physically uniting, the physical quantity unit of rainfall data can be unified to millimeter, the physical quantity unit of snow depth data can be unified to centimeter, the physical quantity unit of flow data can be unified to cubic meter per second, the physical quantity of water level data can be unified to meter, the position coordinate can be unified to degree with 6 decimal places, the runoff velocity can be unified to meter per second, and the remote sensing image resolution can be unified to meter. The units are uniformly marked in the CSV format data to avoid unit confusion during calculation.

[0054] S230: Based on the standardized data set, a flood evolution feature sub-model of each subarea in the target basin is respectively constructed; wherein the subareas include mountainous area, alluvial fan and plain; In some example embodiments of the present application, based on the foregoing scheme, based on the standardized data set, a flood evolution feature sub-model of each subarea in the target basin is respectively constructed, including: From the standardized data set, filtering the geographical parameters (such as terrain slope, river curvature, and basin area) unique to each subarea, hydrological characteristics (such as historical flood peak flow, and concentration time), and underlying surface attribute characteristics (such as vegetation coverage and soil permeability coefficient), a subarea feature vector library is established; Based on the subarea feature vector library and the historical flood process data of each subarea, the initial sub-model of each subarea is trained, and a dynamic correction module is embedded to obtain the flood evolution feature sub-model of each subarea.

[0055] Here, the established subarea feature vector library can be aligned to a unified index to form a subarea feature vector : Among them, represents a static feature vector of geographical parameters, represents a time-varying feature vector driven by hydrological characteristics and historical processes, represents an underlying surface attribute feature vector, is a subarea identifier is a mountainous area, is an alluvial fan, is a plain, is a unit ID, is a time step.

[0056] The vector library is stored in a columnar database or a binary file, and the fields include: subarea identifier, unit ID, time index, feature name, feature value and quality mark (missing / interpolation flag).

[0057] The mountain flood routing characteristic submodel is composed of a graph neural network based on river network / slope topography and a differentiable runoff generation unit (e.g., Green-Ampt or SCS-CN approximation), with a loss function using weighted Huber (with extra weight on peak time to peak volume).

[0058] For each node, the output of each time step is: where, is the soil infiltration volume of unit in time step ; is the rainfall intensity / volume of unit in time step ; is the set of soil properties of unit ; is the soil water content of surface or effective infiltration layer at time ; is the integrated effect of soil matric suction on the advancement of infiltration front, of order m, varying with soil; is the upper limit of hydraulic conductivity of soil mass under saturated condition, high for sandy soil and low for clay soil; generally m / s; represents the runoff of unit in time step ; is the snowmelt volume of unit in time step , is the simplified evapotranspiration or ignored term.

[0059] Loss function of mountain flood routing characteristic submodel is: where, represents the basic regression loss and , represents the difference between predicted and observed flow, represents, represents the predicted flow, represents the observed flow; represents the peak volume loss and , represents the observed peak and , represents the predicted peak and ; represents the time to peak loss and , Indicates the estimated peak time. Indicates the time of the observed peak; Indicates physical consistency regularity and ; Indicates the mass conservation constraint and , , and These represent different weighting coefficients. In time step Internal control volume Actual / estimated total water volume change For local losses, To control volume, The step size for discrete-time integration; Indicates L2 regularization terms. This represents the set of parameters that participated in training and were regularized. and These represent different loss weights.

[0060] The flood evolution characteristic sub-model of an alluvial fan can be designed to include a dual-topological diffusion map and retention-infiltration units. The principal edges of the dual-topological diffusion map are the fan ribs, oriented from the inlet along the maximum slope or fan rib line to the fan edge to simulate the main flood discharge path. The set is denoted as... A bidirectional edge is established between the horizontal edge and its 4 / 8 neighbors to simulate the diffusion and overflow of the flow. Let set [set name missing]. Thus, the resulting bitopological graph is: .

[0061] The outflow function of the storage-infiltration unit is: in, Representation unit The storage-discharge outflow coefficient, Representation unit outflow power and and Related to roughness, roughness resistance, and surface roughness type. Indicates at time step Inside, unit The water depth, Representation unit The micro-topography affects the water depth.

[0062] Infiltration function and characteristic sub-model of mountain flood evolution at time step Inside, unit The soil has the same amount of water infiltration.

[0063] After horizontal swapping: where, denotes the flux of water from cell to cell at time step ; denotes the flux of water from cell to cell at time step , denotes the water depth of cell at time step , denotes the water depth of cell at time step , denotes the ground elevation of cell , denotes the ground elevation of cell .

[0064] The loss function of the alluvial fan flood routing sub-model is: where, , and are the weight coefficients corresponding to different loss functions, denotes the flow fitting loss, denotes the water depth fitting loss and , is the set of observed nodes, denotes the number of time steps used to calculate the loss, denotes the predicted water depth of cell in time step , denotes the observed water depth of cell in time step ; denotes the physical consistency regularization and ; denotes the mass conservation constraint and , , and are different weight coefficients, is the actual / estimated total water volume change of control volume in time step , is the local loss of cell in time step , is the control volume, is the step size of the discrete time integration; denotes an L2 regularization term, denotes the set of parameters participating in training and being regularized.

[0065] The plain flood routing feature sub-model can adopt a lightweight U-Net network model. The lightweight U-Net network model is to reduce the number of channels of each layer of the original U-Net network model to 1 / 2-1 / 4 of the standard U-Net.

[0066] In some embodiments, the loss function of the plain flood routing feature sub-model is: wherein, denotes a water depth fitting loss, and , is a set of observed nodes, denotes a time step for calculating the loss, denotes a predicted water depth of a cell at time step , denotes an observed water depth of a cell at time step ; is a weight of the water depth fitting loss; denotes an intersection over union, denotes an intersection over union weight, denotes a mass conservation constraint, and , denotes a weight coefficient of the mass conservation constraint, is an actual / estimated total water volume change of a control volume at time step , is a local loss of a cell at time step , is a control volume, is a step size of a discrete time integration; denotes an L2 regularization term, denotes the set of parameters participating in training and being regularized.

[0067] The present application customizes and constructs a flood routing feature sub-model according to the geographical conditions, hydrological characteristics and underlying surface properties unique to each sub-region, so that the model structure and parameter settings are more suitable for the actual hydrological process of the sub-region (for example, the mountainous area focuses on the simulation of confluence speed, and the plain focuses on the description of flood detention effect), greatly reduces the simulation deviation caused by one-size-fits-all modeling, and makes the restoration and prediction of the flood routing process more consistent with the actual situation.

[0068] S240: coupling calculation is performed on the output results of the flood evolution characteristic sub-model of each partition by using a lightweight physical-empirical hybrid model to obtain coupling calculation results; In some example embodiments of the present application, based on the foregoing scheme, coupling calculation is performed on the output results of the flood evolution characteristic sub-model of each partition by using a lightweight physical-empirical hybrid model to obtain coupling calculation results, including: Based on the basic law of basin hydrodynamics, a water exchange equation between partitions is constructed.

[0069] The water exchange equation between partitions includes a control volume , the mass conservation and momentum approximation is established; and for the boundaries of mountainous areas and alluvial fans, alluvial fans and plains, plains and downstream outlet, etc., a water head difference driving or open channel formula approximation is used.

[0070] The mass conservation and momentum approximation is: The water head difference driving or open channel formula approximation is: Wherein, represents the control volume of the water storage at the next time step, represents the control volume of the water storage at the current time step, represents the total flow into the control volume at time step , represents the total flow out of the control volume at time step , represents the volume flow of the surface source converted to the control volume at time step , represents the volume flow converted to the control volume at time step , represents the step length of discrete time integration, represents the flux from the unit to the unit at time step , represents the effective transmission coefficient of the unit to the unit , represents the water depth of the unit , represents the ground elevation of the unit , represents the unit The water depth, Representation unit Ground elevation.

[0071] The output results of each sub-model are used as the boundary input of the water exchange equation between the sub-regions. By simplifying the Saint-Venant equations, the hydraulic connection relationship between adjacent sub-regions is solved, and the preliminary coupled flood propagation process (such as the propagation speed of flood waves at the boundary of the sub-regions and the water level connection value) is obtained.

[0072] The simplified Saint-Venant equations are as follows: in, For time steps, For instantaneous water passage, For vertical spatial coordinates, For water flow section, For hydraulic terms related to water level, For bed surface slope, For roughness loss and , For hydraulic radius, It is the acceleration due to gravity. For lateral inflow, The effective velocity of the lateral incoming water in the mainstream direction.

[0073] The hydraulic connection between adjacent zones is solved by simplifying the Saint-Venant equations. Specifically, the boundary quantities (usually flow rates) given by the mountain flood evolution characteristic sub-model, alluvial fan flood evolution characteristic model, and plain flood evolution characteristic model at the zone boundaries are obtained. Water level / depth and possible lateral inflow Using the simplified Saint-Venant equations as boundary conditions, a one-dimensional hydraulic calculation is performed near the boundary to solve for the hydraulic state and exchange flux that should be consistent on both sides of the boundary at the same time, thus obtaining the preliminary coupled flood propagation process.

[0074] An experience coefficient refined in a historical flood process is introduced, and a machine learning method is combined to correct the deviation of the preliminary coupled flood propagation process, focusing on correcting the physical model error caused by complex underlying surfaces such as urban building groups and wetlands, and outputting the corrected subarea flood hydrograph (including flood peak flow, peak time, and flood duration). Here, the experience coefficient includes but is not limited to the equivalent roughness correction coefficient, the diffusion / lagging correction coefficient, and the boundary energy loss coefficient, and the spatial effective range is limited according to the adjacent near field (such as 200-1000m upstream and downstream). The machine learning method can be a neural network model (such as MLP, RNN, TCN, Transformer, etc.), a tree model (GBDT, XGBoost, LightGBM, CatBoost, etc.), and the like, and the present application does not make specific limitations. The correction object is the residual error between the preliminary coupling output and the observation ( ), or directly output the proportional factor of the correction amount ( ).

[0075] For the output results of each sub-model of different time scales (such as minute-level real-time monitoring and hour-level prediction) and spatial scales (such as river cross-sections and sub-basin units), a spatiotemporal interpolation algorithm (such as Kriging interpolation and linear interpolation) is used to unify the scales, generate consistent flood evolution spatiotemporal distribution data (such as flood range and water depth distribution at different times) in the whole basin, and output the flood evolution spatiotemporal distribution data in the whole basin as the coupling calculation result.

[0076] Among them, through the Monte Carlo simulation method, the uncertainty in the coupling calculation process is analyzed, and the confidence interval and probability distribution of the coupling calculation result are output, so as to quantify the uncertainty in the coupling process.

[0077] The present application adopts a lightweight hybrid model of physical mechanism and empirical correction, quickly solves the hydraulic correlation between subareas through simplifying the hydrodynamic equation, and combines historical experience and machine learning to correct deviation, which greatly shortens the coupling calculation time on the premise of ensuring calculation accuracy; in addition, through spatiotemporal scale unification processing, global coherent flood evolution data is generated, which provides complete and intuitive calculation basis for early warning judgment.

[0078] S250: Analyzing historical flood case data based on a machine learning algorithm to generate dynamic early warning thresholds of flood evolution feature sub-models adapted to each subarea; In some example embodiments of the present application, based on the foregoing scheme, analyzing historical flood case data based on a machine learning algorithm to generate dynamic early warning thresholds of flood evolution feature sub-models adapted to each subarea includes: Key influencing factors of each sub-area are extracted from historical flood case data, including hydrological parameters (such as historical flood peak flow, flood hydrograph shape, confluence time), geographical features (such as terrain slope, river flow capacity, levee elevation), meteorological conditions (such as rainfall, rainfall intensity, rainfall duration) and socio-economic data (such as population density, asset distribution), to construct a sub-area feature matrix; According to the actual disaster degree (such as flood loss rate, danger level) of each sub-area in historical flood events, combined with flood control standards and specifications, the corresponding warning level label (such as blue, yellow, orange, red warning) is labeled for each historical case; Gradient boosting tree (such as GBDT), random forest or deep learning model (such as LSTM) is used to train the model with sub-area feature matrix as input and warning level label as output to obtain the trained model; In the trained model, real-time update module is embedded, and current hydrological monitoring data (such as real-time water level, flow), short-term weather forecast data and underlying surface change information (such as land use change, water conservancy project scheduling information) are used as dynamic input, and specific warning threshold (such as critical water level value, flow value corresponding to red warning in a certain sub-area) under the current situation of each sub-area is output by the model; The dynamic warning threshold generated by the model is verified by combining historical extreme value events and expert experience to obtain the dynamic warning threshold of the flood evolution characteristic sub-model adapted to each sub-area. By setting threshold upper and lower limits, deviation correction coefficient and other ways, it is ensured that the threshold meets the data analysis results of the machine learning model and meets the safety redundancy requirements of actual flood control work.

[0079] In some embodiments, the dynamic warning threshold can be directly output by the trained model: taking the sub-area feature matrix as input, the prediction function of the critical water level / flow for different warning levels is trained to obtain The dynamic warning threshold generated by the model for the current situation vector is calculated , and the safety margin and deviation correction are superimposed to obtain the final threshold . Among them and are updated according to historical extreme value events, expert knowledge and recent observations to ensure business safety redundancy and robustness. Among them, represents the dynamic warning threshold generated by the model of the sub-area unit , represents the correction deviation term of the unit , represents the safety margin coefficient of the sub-area unit .

[0080] The application is based on machine learning analysis of historical cases, and current hydrological monitoring data, short-term weather forecast, and underlying surface changes (such as water conservancy dispatching and land use adjustment) are taken into account in threshold calculation to generate dynamic warning thresholds that adapt to real-time scenarios in each subzone. At the same time, the threshold is checked by historical extreme values and expert experience to ensure its reasonableness, making the warning judgment more in line with the current actual situation and effectively reducing the warning deviation caused by fixed thresholds.

[0081] S260: Based on the coupling calculation result and the dynamic warning threshold, generate subzone warning information.

[0082] In some embodiments, the subzone warning information can include one or a combination of the following: warning subzone identification, warning level, predicted risk duration, predicted flood arrival time, impact range, and risk avoidance suggestions.

[0083] Through the existence of the warning subzone identification, not only can the resource waste caused by traditional large-scale vague warning be solved, effectively improving the efficiency of resource allocation in the target subzone, but also the specific villages and road sections on the map can be directly mapped, providing precise spatial coordinate basis for subsequent regional blocking and personnel transfer, reducing the emergency delay caused by fuzzy regional positioning. The warning level can provide a gradientized technical basis for emergency response, avoiding the phenomenon of over-response or insufficient response caused by unclear risk level, and in the case of multiple regional warning, the system can automatically sort the priority based on the warning level, for example, the information of red warning subzone is preferentially pushed to the command terminal, ensuring that the emergency instructions of high-risk areas are preferentially implemented, and improving the overall emergency decision-making efficiency. The predicted risk duration, predicted flood arrival time, and impact range provide technical support for risk avoidance from the time and geographical dimensions. The risk avoidance suggestions not only improve the public's emergency response ability, but also unify the emergency action standards, avoid chaos in emergency action, and improve the overall emergency coordination efficiency of communities, villages, and other areas.

[0084] In some embodiments, the warning subzone identification can be an administrative unit identification (such as directly associating with mature management units such as streets, towns, villages, and communities based on existing administrative divisions, facilitating rapid connection at the grassroots level), a geographic network identification (such as dividing the area into standardized grids, such as 1km×1km or 500m×500m, and identifying them with latitude and longitude or grid codes), or a key facility identification (such as taking high-risk targets such as bridges, tunnels, low-lying road sections, and large communities as the core and marking their surrounding warning range).

[0085] In some embodiments, the warning level can include first-level warning, second-level warning, third-level warning, or in other embodiments, fourth-level warning, fifth-level warning, and so on. The warning identification can be distinguished by color, sound, and other means, for example, when the warning level of a subzone is first-level, the warning identification can be red, and the warning sound is sharp and long in duration.

[0086] In some embodiments, the predicted risk duration can be expressed in hours or days. The predicted flood arrival time can be expressed in hours or minutes. The impact range can be accurate to specific areas, key points, etc. The risk avoidance suggestion can include a risk avoidance suggestion for ordinary residents, a risk avoidance suggestion for grassroots emergency personnel, and a risk avoidance suggestion for specific posts (such as the traffic department and the water conservancy department). For example, in some embodiments, the risk avoidance suggestion for the traffic department is to close XX Road and XX Road (flood risk road sections) before 00:00 on X day, set up 'no entry' signs at the intersection, and guide vehicles to detour XX Road and XX Road.

[0087] In some example embodiments of the present application, based on the foregoing scheme, based on the coupling calculation result and the dynamic early warning threshold, generating partition early warning information includes: When the coupling calculation result is greater than 50% of the dynamic early warning threshold, generating early warning information containing a warning level, an impact range, a predicted risk duration, and a risk avoidance suggestion; When the coupling calculation result is greater than 20% of the dynamic early warning threshold or when the coupling calculation result is less than 20% of the dynamic early warning threshold, generating early warning information containing a warning partition identifier, a warning level, a predicted flood arrival time, an impact range, and a risk avoidance suggestion; and when the coupling calculation result is less than 20% of the dynamic early warning threshold, generating early warning information containing a warning partition identifier, a warning level, and an impact range.

[0088] Based on the comparison of the partition coupling result and the dynamic threshold, the present application generates customized early warning information containing a partition identifier, a warning level, a risk time, an impact range, and a risk avoidance suggestion, which can directly match the response priorities of different areas (such as focusing on personnel transfer paths in mountainous areas and focusing on dike patrol priorities in plains); combined with multi-channel release, it can ensure that the early warning information quickly reaches the target population and management departments, provide accurate guidance for disaster emergency disposal, and improve response efficiency.

[0089] In some example embodiments of the present application, based on the foregoing scheme, after generating partition early warning information based on the coupling calculation result and the dynamic early warning threshold, the partition flood prevention early warning method of fusing multi-source data further includes: Publishing the partition early warning information through at least two of the following: short message, broadcast, application program push, and display screen.

[0090] According to the second aspect of the embodiment of the present application, a partition flood prevention early warning system of fusing multi-source data is also provided, which, as shown in Figure 3 The partition flood prevention early warning system of fusing multi-source data includes: The data acquisition module 310 is configured to acquire a multi-source monitoring data set in a target basin, wherein the multi-source monitoring data set comprises rainfall station monitoring data, snow depth monitoring device data, flow meter data, water level gauge data, Beidou positioning mobile monitoring device data, and satellite remote sensing data. The data preprocessing module 320 is configured to preprocess the multi-source monitoring data set to obtain a standardized data set. The sub-model construction module 330 is configured to construct a flood evolution feature sub-model for each partition in the target basin based on the standardized data set, wherein the partitions include a mountainous area, an alluvial fan, and a plain. The coupling calculation module 340 is configured to perform coupling calculation on the output results of the flood evolution feature sub-models for the partitions by using a lightweight physical-empirical hybrid model to obtain coupling calculation results. The dynamic threshold generation module 350 is configured to generate a dynamic early warning threshold for the flood evolution feature sub-models of the partitions based on machine learning algorithm analysis of historical flood case data. The early warning information generation module 360 is configured to generate partition early warning information based on the coupling calculation results and the dynamic early warning threshold.

[0091] It should be noted that although several modules of the partition flood prevention early warning system fusing multi-source data are mentioned in the foregoing detailed description, such division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or sub-modules described above can be embodied in one module or unit. Conversely, the features and functions of one module described above can be further divided into a plurality of modules or sub-modules.

[0092] In addition, in the exemplary embodiments of the present application, an electronic device capable of implementing the above-mentioned partition flood prevention early warning method fusing multi-source data is also provided.

[0093] Those skilled in the art can understand that each aspect of the present application can be implemented as a system, a method or a program product. Therefore, each aspect of the present application can be embodied as a complete hardware embodiment, a complete software embodiment (including firmware, microcode, etc.), or an embodiment combining hardware and software aspects, which can be collectively referred to as "circuitry", "module" or "system" here.

[0094] The electronic device 400 according to this embodiment of the present application will be described below with reference to Figure 4 The electronic device 400 shown is merely an example and should not impose any limitation on the functions and use range of the embodiments of the present application. Figure 4 The electronic device 400 shown is merely an example and should not impose any limitation on the functions and use range of the embodiments of the present application.

[0095] As Figure 4As shown, the electronic device 400 is in the form of a general computing device. Components of the electronic device 400 can include, but are not limited to, the at least one processing unit 410 described above, the at least one storage unit 420 described above, a bus 430 that connects the various system components, including the storage unit 420 and the processing unit 410, a display unit 440.

[0096] The storage unit stores program code that can be executed by the processing unit 410 such that the processing unit 410 performs the steps described in the above "Exemplary Method" section according to various exemplary embodiments of the present application. For example, the processing unit 410 can execute the program code as shown in Figure 2 As shown in S210 in FIG. 2, a plurality of monitoring data sets in a target basin are acquired, wherein the plurality of monitoring data sets include rainfall station monitoring data, snow depth monitoring device data, flow meter data, water level gauge data, Beidou positioning mobile monitoring device data, and satellite remote sensing data; S220: the plurality of monitoring data sets are preprocessed to obtain a standardized data set; S230: based on the standardized data set, a flood evolution feature sub-model of each partition in the target basin is constructed respectively; wherein the each partition includes a mountainous area, an alluvial fan, and a plain; S240: a light physical-empirical hybrid model is used to perform coupling calculation on output results of the flood evolution feature sub-models of the each partition to obtain a coupling calculation result; S250: based on a machine learning algorithm, historical flood case data are analyzed to generate a dynamic early warning threshold that is adapted to the flood evolution feature sub-models of the each partition; and S260: based on the coupling calculation result and the dynamic early warning threshold, partition early warning information is generated.

[0097] The storage unit 420 can include a readable medium in the form of volatile storage such as random access memory (RAM) 421 and / or cache memory 422, and can further include a read-only memory (ROM) 423.

[0098] The storage unit 420 can further include program / utility 424 having a set of program modules 425, including but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which can include implementation of a network environment, individually or in some combination.

[0099] The bus 430 can be representative of one or more of several types of bus structures, including a storage unit bus or storage unit controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of a variety of bus structures.

[0100] Electronic device 400 can also communicate with one or more external devices 470 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable a user to interact with electronic device 400, and / or with any device that enables electronic device 400 to communicate with one or more other computing devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 450. Furthermore, electronic device 400 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 460. As shown, network adapter 460 communicates with other modules of electronic device 400 via bus 430. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 400, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0101] Through the description of the above embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions of the embodiments of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, portable hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the methods according to the embodiments of the present invention.

[0102] In exemplary embodiments of the present invention, a computer-readable storage medium is also provided, on which a program product capable of implementing the methods described above is stored. In some possible embodiments, various aspects of the present invention may also be implemented as a program product comprising program code, which, when the program product is run on a terminal device, causes the terminal device to perform the steps of the various exemplary embodiments of the present invention described in the "Exemplary Methods" section above.

[0103] refer to Figure 5 As shown, a program product 500 for implementing the above-described method for merging multi-source data for regional flood warning according to an embodiment of the present invention is described. It may employ a portable compact disc read-only memory (CD-ROM) and include program code, and may run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited thereto. In the present invention, the readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0104] The program product can take any combination of one or more readable storage media. The readable storage media can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any combination thereof. More specific examples (a non-exhaustive list) of the readable storage media include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0105] The program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++, or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server. In the latter scenario, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computing device, such as through the Internet using an Internet Service Provider. The present application can be implemented as a computer program product, which can include a computer-readable medium having stored computer program code.

[0106] Furthermore, the above-described diagrams are merely schematic illustrations of the processes included in the method according to the exemplary embodiments of the present application, and are not intended for limiting purposes. It is readily understood that the processes shown in the above-described diagrams do not indicate or limit the time sequence of the processes. In addition, it is readily understood that the processes can be executed synchronously or asynchronously, for example, in a plurality of modules.

[0107] From the above description of the embodiments, it is readily understood by those skilled in the art that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a U disk, a mobile hard disk, or the like) or on a network, and includes a plurality of instructions to cause a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiments of the present application.

[0108] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the application encompass any and all variations or modifications of the application which come within the scope of the overall concept thereof and that the true scope and spirit of the application be indicated by the claims and can include other concepts, aspects and features disclosed herein. The specification and examples are to be considered exemplary only, with the true scope and spirit of the application indicated by the claims.

[0109] It is to be understood that the application is not limited to the precise construction here described and illustrated in the accompanying drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application should only be limited by the claims appended hereto.

Claims

1. A method for regional flood prevention and early warning that integrates multi-source data, characterized in that, include: Obtain multi-source monitoring datasets for each zone within the target watershed, wherein the multi-source monitoring datasets include rain gauge monitoring data, snow depth monitoring equipment data, flow meter data, water level gauge data, BeiDou positioning mobile monitoring equipment data, and satellite remote sensing data; The multi-source monitoring dataset is preprocessed to obtain a standardized dataset; Based on the standardized dataset, flood evolution feature sub-models for each zone within the target watershed are constructed; wherein, each zone includes mountainous areas, alluvial fans, and plains; A lightweight physics-empirical hybrid model was used to couple the output results of the flood evolution characteristic sub-models of each zone to obtain the coupled calculation results; Based on machine learning algorithms, analyze historical flood case data to generate dynamic early warning thresholds for flood evolution feature sub-models adapted to each zone; Based on the coupling calculation results and the dynamic early warning threshold, zonal early warning information is generated.

2. The method for regional flood prevention and early warning by fusing multi-source data according to claim 1, characterized in that, Based on the standardized dataset, flood evolution feature sub-models for each zone within the target watershed are constructed, including: The unique geographical parameters, hydrological features, and underlying surface attributes of each region are selected from the standardized dataset to establish a regional feature vector library. Based on the regional feature vector library and historical flood process data of each region, the initial sub-model of each region is trained and a dynamic correction module is embedded to obtain the flood evolution feature sub-model of each region.

3. The method for regional flood prevention and early warning by fusing multi-source data according to claim 1, characterized in that, A lightweight physics-empirical hybrid model was used to couple the output results of the flood evolution characteristic sub-models for each zone, and the resulting coupled calculations included: Based on the fundamental laws of watershed hydrodynamics, a water exchange equation between different intervals is constructed. The output results of each sub-model are used as the boundary input of the water exchange equation between the sub-regions. The hydraulic connection relationship between adjacent sub-regions is solved by simplifying the Saint-Venant equations, and the preliminary coupled flood propagation process is obtained. By introducing empirical coefficients extracted from historical flood processes and combining them with machine learning methods, the bias of the initially coupled flood propagation process is corrected, and the corrected zonal flood process lines are output. For the output results of each sub-model at different time and spatial scales, a spatiotemporal interpolation algorithm is used to unify the scale, generate consistent spatiotemporal distribution data of flood evolution across the entire watershed, and output the consistent spatiotemporal distribution data of flood evolution across the entire watershed as the coupled calculation result; Specifically, the Monte Carlo simulation method is used to perform propagation analysis on the uncertainty in the coupled calculation process, and output the confidence interval and probability distribution of the coupled calculation results to quantify the uncertainty in the coupling process.

4. The method for regional flood prevention and early warning by fusing multi-source data according to claim 1, characterized in that, Based on machine learning algorithms, historical flood case data is analyzed to generate dynamic early warning thresholds for flood evolution characteristic sub-models adapted to each region, including: Key influencing factors for each region were extracted from historical flood case data, and a region feature matrix was constructed. Based on the actual degree of damage to each zone in historical flood events, and in accordance with flood control standards and specifications, a corresponding warning level label is marked for each historical case; The model is trained by using gradient boosting tree, random forest or deep learning model, with the partition feature matrix as input and the warning level label as output. A real-time update module is embedded in the trained model, taking current hydrological monitoring data, short-term meteorological forecast data and underlying surface change information as dynamic inputs, and the model outputs the specific warning thresholds for each zone under the current situation. By combining historical extreme events and expert experience, the dynamic early warning thresholds generated by the model are verified to obtain dynamic early warning thresholds that are adapted to the flood evolution characteristic sub-models of each zone.

5. The method for regional flood prevention and early warning by fusing multi-source data according to claim 1, characterized in that, The zoning early warning information includes one or more of the following: early warning zone identifier, early warning level, expected duration of risk, expected time of flood arrival, scope of impact, and risk avoidance recommendations.

6. The method for regional flood prevention and early warning by fusing multi-source data according to claim 1, characterized in that, Based on the coupling calculation results and the dynamic early warning threshold, the generated zonal early warning information includes: When the coupling calculation result is greater than 50% or more of the dynamic early warning threshold, early warning information including early warning level, scope of impact, expected duration of risk and risk avoidance suggestions is generated. When the coupling calculation result is greater than 20% of the dynamic early warning threshold, or when the coupling calculation result is less than 20% of the dynamic early warning threshold, an early warning message is generated that includes the early warning zone identifier, early warning level, expected flood arrival time, impact range, and evacuation advice; when the coupling calculation result is less than 20% of the dynamic early warning threshold, an early warning message is generated that includes the early warning zone identifier, early warning level, and impact range.

7. The method for regional flood prevention and early warning by fusing multi-source data according to claim 1, characterized in that, After generating zonal early warning information based on the coupled calculation results and the dynamic early warning threshold, the zonal flood prevention early warning method that integrates multi-source data further includes: Distribute zone alert information through at least two of the following methods: SMS, broadcast, app push, and display screen.

8. A regional flood control early warning system integrating multi-source data, characterized in that, include: The data acquisition module is used to acquire multi-source monitoring datasets within the target watershed, wherein the multi-source monitoring datasets include rain gauge monitoring data, snow depth monitoring equipment data, flow meter data, water level gauge data, BeiDou positioning mobile monitoring equipment data, and satellite remote sensing data; The data preprocessing module is used to preprocess the multi-source monitoring dataset to obtain a standardized dataset; The sub-model construction module is used to construct flood evolution characteristic sub-models for each partition within the target watershed based on the standardized dataset; wherein, each partition includes mountainous areas, alluvial fans, and plains; The coupling calculation module is used to perform coupling calculations on the output results of the flood evolution characteristic sub-models of each partition using a lightweight physical-empirical hybrid model, and obtain the coupling calculation results. The dynamic threshold generation module is used to analyze historical flood case data based on machine learning algorithms and generate dynamic early warning thresholds that are adapted to the flood evolution feature sub-models of each zone. The early warning information generation module is used to generate zoned early warning information based on the coupled calculation results and the dynamic early warning threshold.

9. An electronic device, characterized in that, include: processor; as well as A memory storing computer-readable instructions that, when executed by the processor, implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images