Power grid flood risk early warning method and device, storage medium and program product

Abstract

The invention discloses a power grid flood risk early warning method and device, a storage medium and a program product. The method comprises the steps that multi-source data are acquired, a physical process simulation model is established according to the multi-source data, and the physical process simulation model is used for representing hydrological characteristics of areas of different underlying surface types under different rainfall capacities; acquiring a machine learning model; integrating the physical process simulation model and the machine learning model to obtain a flood risk prediction model; and determining a power grid flood risk level of the monitoring area by adopting a flood risk prediction model according to the weather forecast data, the underlying surface type and the voltage level of the power grid equipment corresponding to the monitoring area. According to the method and the device, the technical problem that the flood risk prediction effect is poor due to the fact that the flood risk prediction model in the related technology often neglects the influence of different underlying surfaces on rainwater permeation, evaporation and runoff is solved.

Description

Technical Field

[0001] This application relates to the technical field of risk warning, and in particular, to a method, device, storage medium, and program product for warning of power grid flood risk. Background Art

[0002] Flood disasters are one of the natural disasters that exist globally. Especially in the case of frequent extreme weather events, the impact of flood disasters on human society is becoming more and more serious. As an important infrastructure of modern society, the safe operation of the power grid is directly related to social stability and development. The threat of flood disasters to the power grid cannot be ignored. Especially in heavy rainfall weather, transmission lines may fail due to flood attacks, and even cause large-scale power outages.

[0003] However, the flood risk prediction methods in related technologies often ignore the influence of different underlying surfaces on rainwater infiltration, evaporation, and runoff, and there are technical problems with poor flood risk prediction effects.

[0004] At present, no effective solution has been proposed for the above problems. Summary of the Invention

[0005] The embodiments of this application provide a method, device, storage medium, and program product for warning of power grid flood risk, so as to at least solve the technical problem of poor flood risk prediction effects caused by the fact that the flood risk prediction models in related technologies often ignore the influence of different underlying surfaces on rainwater infiltration, evaporation, and runoff.

[0006] According to one aspect of the embodiments of this application, a method for warning of power grid flood risk is provided, including: obtaining multi-source data, and establishing a physical process simulation model based on the multi-source data, where the physical process simulation model is used to characterize the hydrological characteristics of areas with different underlying surface types under different rainfall amounts; obtaining a machine learning model, where the machine learning model is trained based on the historical flood event data corresponding to the monitoring area and is used to predict the inundation depth of the monitoring area under different land and precipitation conditions; integrating the physical process simulation model and the machine learning model to obtain a flood risk prediction model; using the flood risk prediction model to determine the power grid flood risk level of the monitoring area based on the weather forecast data, underlying surface type, and voltage level of the power grid equipment corresponding to the monitoring area, where the power grid flood risk level is used to characterize the degree of influence of the monitoring area and the power grid equipment by floods.

[0007] Optionally, the multi-source data includes: satellite remote sensing data, land classification data, topographic and geomorphic data, urban pipe network data, rainfall data, power grid geographic information system data. Among them, the satellite remote sensing data is used to characterize the distribution of surface water bodies and the distribution of rainfall areas; the land classification data is used to characterize the underlying surface types of the surface in different regions; the topographic and geomorphic data is used to characterize the slope and aspect of different regions; the urban pipe network data is used to characterize the distribution and drainage capacity of the drainage pipe network; the rainfall data is used to characterize the temporal and spatial distribution of rainfall; the power grid geographic information system data is used to characterize the location distribution and voltage level of power grid equipment.

[0008] Optionally, before establishing a physical process simulation model based on the multi-source data, the method further includes: performing data preprocessing on the multi-source data, and performing normalization processing on the multi-source data after data preprocessing. Among them, data preprocessing includes: outlier removal and missing value filling; performing data fusion operations on the multi-source data after normalization processing. Among them, data fusion operations include: spatial fusion, temporal fusion, and attribute fusion. Spatial fusion is used to superimpose spatial data from different sources in the multi-source data onto the same coordinate system; temporal fusion is used to match and integrate data with different time scales in the multi-source data; attribute fusion is used to merge attribute information from different sources in the multi-source data.

[0009] Optionally, the physical process simulation model includes at least one of the following: rainfall-runoff model, surface runoff model, inundation simulation model; establishing a physical process simulation model based on the multi-source data includes: determining a rainfall-runoff model based on satellite remote sensing data, land classification data, and rainfall data. Among them, the rainfall-runoff model is used to simulate the process of rainfall converting into surface runoff. In the rainfall-runoff model, the runoff curve numbers and infiltration parameters corresponding to different underlying surface types are different. The underlying surface types include at least one of the following: impervious ground type, forest land ground type, bare land ground type; determining a surface runoff model based on land classification data, topographic and geomorphic data, and urban pipe network data. Among them, the surface runoff model is used to simulate the flow process of surface water in regions with different underlying surface types; determining an inundation simulation model based on rainfall data, topographic and geomorphic data, and the surface runoff model. Among them, the inundation simulation model is used to simulate the diffusion process of floods in regions with different underlying surface types and predict the inundation range and depth.

[0010] Optionally, obtaining a machine learning model includes: obtaining historical flood event data, where the historical flood event data includes: historical rainfall data of the monitoring area, terrain data of the monitoring area, underlying surface type of the monitoring area, and historical inundation depth; using a machine learning algorithm to extract data features from the historical flood event data and training an initial model based on the data features to obtain a machine learning model, where the machine learning algorithm includes at least one of the following: random forest algorithm, gradient boosting decision tree algorithm, support vector machine algorithm.

[0011] Optionally, using a flood risk prediction model to determine the grid flood risk level of the monitoring area based on the weather forecast data, underlying surface type, and voltage level of the grid equipment corresponding to the monitoring area includes: obtaining environmental data collected by sensor devices in the monitoring area, where the sensor devices include at least one of the following: rain gauge, water level gauge, flow meter, soil moisture sensor; using the flood risk prediction model to predict the inundation depth and drainage system drainage capacity corresponding to the monitoring area based on the environmental data and weather forecast data; using the analytic hierarchy process to determine the weight coefficients corresponding to the inundation depth, underlying surface type, voltage level of the grid equipment, and drainage system drainage capacity corresponding to the monitoring area; and determining the grid flood risk level of the monitoring area based on the inundation depth, underlying surface type, voltage level of the grid equipment, drainage system drainage capacity, and corresponding weight coefficients corresponding to the monitoring area.

[0012] Optionally, after determining the grid flood risk level of the monitoring area, the method further includes: generating a warning message based on the grid flood risk level and sending the warning message to a mobile terminal device located in the monitoring area.

[0013] According to another aspect of the embodiments of the present application, there is also provided a grid flood risk warning device, including: a physical process simulation module for obtaining multi-source data and establishing a physical process simulation model based on the multi-source data, where the physical process simulation model is used to characterize the hydrological characteristics of areas with different underlying surface types under different rainfall amounts; a machine learning optimization module for obtaining a machine learning model, where the machine learning model is trained based on historical flood event data corresponding to the monitoring area and is used to predict the inundation depth of the monitoring area under different land and precipitation conditions; a prediction model integration module for integrating the physical process simulation model and the machine learning model to obtain a flood risk prediction model; and a flood risk warning module for using the flood risk prediction model to determine the grid flood risk level of the monitoring area based on the weather forecast data, underlying surface type, and voltage level of the grid equipment corresponding to the monitoring area, where the grid flood risk level is used to characterize the degree of influence of the monitoring area and grid equipment by floods.

[0014] According to another aspect of the embodiments of the present application, a non-volatile storage medium is provided, the non-volatile storage medium includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the power grid flood risk warning method by running the computer program.

[0015] According to another aspect of the embodiments of the present application, a computer program product is provided, including a computer program, which implements the steps of the power grid flood risk warning method when the computer program is executed by a processor.

[0016] In an embodiment of the present application, multi-source data is obtained, and a physical process simulation model is established based on the multi-source data, wherein the physical process simulation model is used to characterize the hydrological characteristics of areas with different underlying surface types under different rainfall amounts; a machine learning model is obtained, wherein the machine learning model is trained based on historical flood event data corresponding to the monitoring area, and is used to predict the flooding depth of the monitoring area under different land and precipitation conditions; the physical process simulation model and the machine learning model are integrated to obtain a flood risk prediction model; the flood risk prediction model is used to determine the grid flood risk level of the monitoring area based on the weather forecast data, the underlying surface type, and the voltage level of the power grid equipment corresponding to the monitoring area, wherein the grid flood risk level is used to characterize the degree to which the monitoring area and the power grid equipment are affected by floods. By combining physical process simulation and machine learning optimization, risk assessment is performed for different underlying surface types, thereby improving the accuracy of flood risk prediction and achieving the purpose of real-time monitoring and early warning of grid flood risks, thereby solving the technical problem that the flood risk prediction model in the related art often ignores the influence of different underlying surfaces on rainwater infiltration, evaporation and runoff, resulting in poor flood risk prediction effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 It is a hardware structure block diagram of a computer terminal (or electronic device) for implementing a method for early warning of flood risk in a power grid provided in an embodiment of the present application;

[0019] Figure 2 It is a schematic diagram of a method flow for early warning of flood risk in a power grid provided according to an embodiment of the present application;

[0020] Figure 3 It is a schematic diagram of a technical route of a power grid flood risk prediction and early warning method for different underlying surface types provided according to an embodiment of the present application;

[0021] Figure 4It is a schematic diagram of an SCS-CN model processing flow provided by an embodiment of the present application;

[0022] Figure 5 It is a schematic diagram of a HEC-HMS model processing flow provided by an embodiment of the present application;

[0023] Figure 6 It is a schematic structural diagram of a power grid flood risk warning device provided by an embodiment of the present application. Detailed implementation manners

[0024] In order to enable those skilled in the art to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0025] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units is not necessarily limited to those clearly listed steps or units, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0026] For the convenience of those skilled in the art to better understand the embodiments of the present application, some technical terms or noun explanations related to the embodiments of the present application are as follows:

[0027] SCS-CN hydrological model (Soil Conservation Service Curve Number method): It is a watershed hydrological model used to predict rainfall runoff. This model was initially developed by the Soil Conservation Service of the US Department of Agriculture (now the Natural Resources Conservation Service of the US Department of Agriculture) to evaluate soil erosion and flood prevention projects.

[0028] HEC-HMS (The Hydrologic Engineering Center’s-Hydrologic Modeling System): A basin-scale flood simulation system developed by the Hydrologic Engineering Center of the U.S. Army Corps of Engineers, which is a semi-distributed sub-rainfall runoff model with physical concepts. According to the formation process of rainfall runoff, it is divided into four calculation parts: net rainfall process, direct runoff process, base flow, and channel confluence, and is mainly used for simulating the runoff process of dendritic basins.

[0029] The grid flood risk management methods in related technologies mainly rely on traditional hydrometeorological monitoring means and historical disaster data. Although these methods can, to a certain extent, predict the occurrence of flood disasters and their impacts on the power grid, due to the lack of detailed distinction and targeted analysis of underlying surface types, it is often difficult to provide accurate risk warnings. Moreover, the flood risk prediction methods in related technologies mostly focus on the construction and optimization of the overall urban flood control system, and the main technologies involved include hydrological models, meteorological prediction models, and Geographic Information System (GIS), etc. For example, some studies use remote sensing data and land classification data to simulate the distribution of urban flood risks and combine rainfall forecast data for risk warnings. These studies have achieved remarkable results in urban flood control, but their applications in grid flood risk prediction and warning are relatively few. For the flood risk assessment of grid facilities, most methods rely on the statistical analysis of historical data and lack real-time and forward-looking nature. Some studies attempt to predict flood risks by constructing hydrological models of the surrounding environment of transmission lines, but these methods usually ignore the impact of underlying surface types on flood risks and are difficult to accurately reflect the risk differences under different underlying surface conditions.

[0030] Specifically, there are mainly the following deficiencies:

[0031] 1) Current flood risk prediction models often rely on historical data and limited meteorological information, and the prediction accuracy for sudden heavy rainfall events is not high. This is mainly because existing models are difficult to comprehensively capture and simulate the occurrence and development process of rainfall when facing complex and changeable meteorological conditions. In addition, most models do not fully consider the influence of different underlying surfaces on rainwater infiltration, evaporation, and runoff during prediction, resulting in large deviations in prediction results. The response mechanism of underlying surface types (such as impervious surfaces, forestlands, bare soils, etc.) to rainfall is complex, and there are significant differences in flood risks under different underlying surface conditions.

[0032] 2) Although satellite remote sensing data, land classification data, etc. provide rich information, how to effectively integrate these multi-source heterogeneous data to form a unified analysis framework remains a major challenge. The spatio-temporal resolution, accuracy, and data formats of various data sources are different, and problems such as data inconsistency, redundancy, and conflicts need to be solved during the data fusion process. The acquisition and processing of some data (such as satellite data) have time delays and it is difficult to meet the requirements of real-time warning. To achieve real-time warning, it is necessary to improve the speed of data collection, transmission, and processing, shorten the time interval of data update, and ensure the timeliness and accuracy of data.

[0033] 3) The existing risk assessment methods have a single evaluation index, often focusing on flood levels or inundation areas, and ignoring factors such as underlying surface types and power grid characteristics, resulting in incomplete evaluation results. Flood risk assessment should comprehensively consider multiple factors such as inundation depth, underlying surface type, transmission line voltage level, and drainage systems to provide more accurate and comprehensive risk assessment results. Moreover, the existing warning methods may lack customized warning strategies for specific regions and specific facilities (such as power grids), making it difficult to provide accurate risk warnings. The warning methods should be flexible and scalable, capable of adjusting warning parameters and strategies according to the characteristics of different regions and facilities to improve the pertinence and effectiveness of warnings.

[0034] To solve the above problems, relevant solutions are provided in the embodiments of this application, which are described in detail below.

[0035] According to the embodiments of this application, a method embodiment for power grid flood risk warning is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0036] The method embodiments provided by the embodiments of this application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Figure 1 A hardware structure block diagram of a computer terminal (or electronic device) for implementing the power grid flood risk warning method is shown. As Figure 1As shown, the computer terminal 10 (or electronic device) may include one or more processors 102 (shown as 102a, 102b, ……, 102n in the figure) (the processor 102 may include, but is not limited to, processing devices such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may further include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply, and / or a camera. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 may further include more or fewer components than those Figure 1 shown in, or have a different configuration from that Figure 1 shown.

[0037] It should be noted that the above one or more processors 102 and / or other data processing circuits are generally referred to as "data processing circuits" herein. The data processing circuit may be embodied in software, hardware, firmware, or any combination thereof, in whole or in part. In addition, the data processing circuit may be a single independent processing module, or be incorporated in whole or in part into any one of the other elements in the computer terminal 10 (or electronic device). As involved in the embodiments of the present application, the data processing circuit is used for processor control (such as the selection of a variable resistance terminal path connected to an interface).

[0038] The memory 104 may be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the power grid flood risk warning method in the embodiments of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the above-mentioned power grid flood risk warning method. The memory 104 may include a high-speed random access memory, and may further include a non-volatile memory, such as one or more magnetic storage devices, a flash memory, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely set relative to the processor 102, and these remote memories may be connected to the computer terminal 10 through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0039] The transmission device 106 is used to receive or send data via a network. Specific examples of the above-mentioned network may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a Radio Frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0040] The display can be, for example, a touch-screen liquid crystal display (LCD), which enables users to interact with the user interface of the computer terminal 10 (or electronic device).

[0041] Under the above operating environment, an embodiment of the present application provides a method for early warning of power grid flood risk. Figure 2 It is a schematic diagram of the method flow for early warning of power grid flood risk provided by an embodiment of the present application, as Figure 2 shown, the method includes the following steps:

[0042] Step S202, obtain multi-source data, and establish a physical process simulation model based on the multi-source data, where the physical process simulation model is used to characterize the hydrological characteristics of areas with different underlying surface types under different rainfall amounts;

[0043] Step S204, obtain a machine learning model, where the machine learning model is trained based on the historical flood event data corresponding to the monitoring area and is used to predict the inundation depth of the monitoring area under different land and precipitation conditions;

[0044] Step S206, integrate the physical process simulation model and the machine learning model to obtain a flood risk prediction model;

[0045] Step S208, use the flood risk prediction model to determine the power grid flood risk level of the monitoring area based on the weather forecast data, underlying surface type, and voltage level of the power grid equipment corresponding to the monitoring area, where the power grid flood risk level is used to characterize the degree of influence of the monitoring area and power grid equipment by floods.

[0046] Through the above steps, by combining physical process simulation and machine learning optimization, risk assessment is carried out for different underlying surface types, achieving the purpose of improving the accuracy of flood risk prediction and realizing real-time monitoring and early warning of power grid flood risk. Furthermore, it solves the technical problem that the flood risk prediction effect is not good because the flood risk prediction models in the related technologies often ignore the influence of different underlying surfaces on rainwater infiltration, evaporation, and runoff.

[0047] The following further introduces the power grid flood risk warning method in steps S202 to S208 of the embodiments of the present application.

[0048] Figure 3 It is a schematic diagram of the technical route of a power grid flood risk prediction and warning method for different underlying surface types provided by the embodiments of the present application. As Figure 3 shown, the embodiments of the present application propose a flood risk prediction and assessment method integrating multi-source data. Combining satellite remote sensing data, land classification data, topographic and geomorphic data, urban pipe network data, rainfall forecast data, power grid GIS data, and historical flood event data, a multi-source data fusion platform is constructed through data cleaning, normalization, and fusion technologies. Physical process simulation methods such as the SCS-CN model and the HEC-HMS model are used, combined with the Digital Elevation Model (DEM) for surface runoff simulation, and the SWMM (Storm Water Management Model) and Flood Modeller models are used for inundation simulation. The model is optimized through machine learning algorithms, key features are extracted, and model training and integration are carried out. For different underlying surface types (impervious surfaces, forest land, bare soil), specific rainfall-runoff, surface runoff, and inundation simulation methods are used to predict flood risks. The Analytic Hierarchy Process (AHP) and comprehensive scoring method are used to evaluate flood risks, combined with environmental data collected by real-time sensor networks, for dynamic assessment and early warning information release to ensure that relevant departments and the public can obtain flood risk information in a timely manner and take protective measures and emergency responses. The following is a specific introduction.

[0049] First, multi-source data is obtained. In some embodiments of the present application, the multi-source data includes: satellite remote sensing data, land classification data, topographic and geomorphic data, urban pipe network data, rainfall data, and power grid geographic information system data. Among them, the satellite remote sensing data is used to characterize the distribution of surface water bodies and the distribution of rainfall areas, and the land classification data is used to characterize the underlying surface types of the surfaces in different regions; the topographic and geomorphic data is used to characterize the slope and aspect of different regions; the urban pipe network data is used to characterize the distribution of drainage pipe networks and drainage capabilities, the rainfall data is used to characterize the temporal and spatial distribution of rainfall, and the power grid geographic information system data is used to characterize the location distribution and voltage levels of power grid equipment.

[0050] Specifically, in the embodiments of the present application, the data sources of multi-source data include: 1) satellite remote sensing data, the data content of which includes surface images, rainfall, surface water body distribution, etc., which are used to provide large-scale and high-resolution surface information, help identify rainfall areas and water body diffusion conditions, and support the identification and classification of underlying surface types; 2) land classification data, the data content of which includes land use types (i.e., underlying surface types), which are used to determine different land use types on the surface, including impervious surfaces (such as urban roads, buildings), forest land (forests, grasslands), bare soil (wasteland, cultivated land), etc. Corresponding to different rainwater infiltration, evaporation and runoff characteristics, it helps to adjust the parameters of the flood risk model; 3) topographic and geomorphic data, the data content of which includes digital elevation model (DEM), slope, aspect, river network, etc. It is used for topographic analysis, simulating the flow path and convergence process of surface runoff, and determining the flood flow direction and speed; 4) urban pipe network data, the data content of which includes data of urban planning departments and drainage facility management units, including the distribution of drainage pipe networks, drainage capacity, pump station locations and capacities, and the maintenance status of drainage systems. It is used to evaluate the capacity and bottlenecks of urban drainage systems, conduct flood simulations in combination with rainfall data, and predict urban flood risks; 5) rainfall data, the content of which includes future rainfall, rainfall intensity, rainfall time and spatial distribution, as well as historical rainfall data, etc., which are used to provide real-time and historical rainfall information and serve as important input data for flood risk prediction models; 6) power grid GIS data, the content of which includes power grid GIS data and power company line data, including the specific locations of transmission lines, voltage levels (low voltage, medium voltage, high voltage), and the distribution of overhead lines and underground lines, etc., which are used to determine the scope and degree of the impact of flood risks on power grid facilities and help formulate targeted protection measures and emergency plans; 7) historical flood event data, the content of which includes records of local governments and disaster prevention and mitigation agencies, recording historical flood events, including occurrence time, location, inundation depth (recording the water depth in each area during floods, the specific range includes 0 - 0.5 meters, 0.5 - 1 meter, 1 - 2 meters, above 2 meters), affected range and loss situation, etc. It is used to provide training data for machine learning models, help optimize model parameters, and improve prediction accuracy.

[0051] After that, the obtained multi-source data is fused and processed, and the specific steps are as follows.

[0052] In some embodiments of the present application, before establishing a physical process simulation model based on multi-source data, the method further includes the following steps: performing data preprocessing on the multi-source data, and performing normalization processing on the multi-source data after data preprocessing, wherein the data preprocessing includes: outlier removal and missing value filling; performing data fusion operations on the multi-source data after normalization processing, wherein the data fusion operations include: spatial fusion, temporal fusion, and attribute fusion. Spatial fusion is used to superimpose spatial data from different sources in the multi-source data onto the same coordinate system. Temporal fusion is used to match and integrate data with different time scales in the multi-source data. Attribute fusion is used to combine attribute information from different sources in the multi-source data.

[0053] Specifically, first, perform data cleaning on the multi-source data to remove noise and errors in the data and ensure data quality. In this embodiment, techniques such as outlier detection, missing value filling, and data consistency verification can be used. Among them, outlier detection can use statistical methods (such as the 3σ principle) to identify and eliminate abnormal data points; missing value filling can use methods such as interpolation and regression analysis to fill in missing data; data consistency verification is used to ensure the consistency of time, space, and attribute information in different data sources.

[0054] Then perform data normalization to convert data with different dimensions to the same scale for easy comparison and fusion. In this embodiment, methods such as min-max normalization and Z-score standardization can be used. Among them, min-max normalization can scale the data proportionally to the [0, 1] interval. Z-score standardization can convert the data into a standard normal distribution (mean of 0 and standard deviation of 1).

[0055] Finally, perform data fusion of the multi-source data, including: 1) Spatial fusion: Perform data matching and fusion based on geographical coordinates, and superimpose spatial data from different sources onto the same coordinate system. Use GIS software and spatial databases (such as PostGlS) to perform spatial overlay analysis to generate a comprehensive spatial data layer. 2) Temporal fusion: Match and integrate data with different time scales to form time series data. Use time interpolation methods (such as linear interpolation and spline interpolation) to fill in and match data with inconsistent time resolutions. 3) Attribute fusion: Combine and synthesize attribute information from different data sources to form a unified attribute database. Use multi-attribute decision analysis (such as the analytic hierarchy process and grey system theory) to comprehensively evaluate and assign weights to attribute data.

[0056] In addition, as an optional real-time approach, in the embodiments of the present application, a multi-source data fusion platform can also be constructed. Specifically, a modular design can be adopted to build the platform architecture, including modules such as data collection, cleaning, normalization, fusion, and analysis, and technologies such as big data technology, distributed computing, and stream processing are used to implement it. Among them, big data technology can utilize big data technologies such as Hadoop and Spark to improve the efficiency and scalability of data processing. Distributed computing can adopt a distributed computing framework to achieve parallel processing of large-scale data and enhance the speed of data fusion and analysis. Stream processing technology can introduce stream processing technologies such as Apache Kafka and Flink to achieve efficient processing and fusion of real-time data and meet the requirements of real-time early warning.

[0057] In the embodiments of the present application, by combining satellite remote sensing data, land classification data, impervious surface data, urban pipe network data, rainfall forecast data, and power grid GIS data, effective fusion and comprehensive analysis of multi-source data are achieved, thereby improving the accuracy of flood risk prediction.

[0058] After that, based on the multi-source data after processing and fusion, a physical process simulation model can be established. The specific steps are as follows.

[0059] In some embodiments of the present application, the physical process simulation model includes at least one of the following: rainfall-runoff model, surface runoff model, inundation simulation model; establishing a physical process simulation model based on multi-source data includes the following steps: determining a rainfall-runoff model based on satellite remote sensing data, land classification data, and rainfall data, where the rainfall-runoff model is used to simulate the process of rainfall being converted into surface runoff. In the rainfall-runoff model, the runoff curve numbers and infiltration parameters corresponding to different underlying surface types are different, and the underlying surface types include at least one of the following: impervious ground type, forest land ground type, bare land ground type; determining a surface runoff model based on land classification data, topographic and geomorphic data, and urban pipe network data, where the surface runoff model is used to simulate the flow process of surface water in areas with different underlying surface types; determining an inundation simulation model based on rainfall data, topographic and geomorphic data, and the surface runoff model, where the inundation simulation model is used to simulate the diffusion process of floods in areas with different underlying surface types and predict the inundation range and depth.

[0060] Specifically, in the embodiments of the present application, the physical process simulation model includes: rainfall-runoff model, surface runoff model, inundation simulation model, etc., which are introduced separately below.

[0061] For the rainfall-runoff model, in the embodiments of the present application, in order to accurately simulate the process of rainfall being converted into surface runoff, hydrological models such as the SCS-CN model or the HEC-HMS model can be used. These models set specific parameters for different underlying surface types to reflect their influence on the conversion of rainfall into runoff. The following will be described by taking three types of underlying surfaces, namely the impervious surface type, the forest land surface type, and the bare land surface type, as examples.

[0062] Taking the SCS-CN model as an example of the rainfall-runoff model, as Figure 4 shown, the SCS-CN model can estimate the runoff generated by rainfall by calculating the Curve Number (CN) based on factors such as soil type, land use, and underlying surface conditions. The specific parameter settings are as follows: Impervious surface: High CN value (85 - 100), indicating that most of the rainfall is converted into runoff, with less infiltration and evaporation. Forest land: Low CN value (30 - 55), considering higher rainfall interception and evaporation, and a higher infiltration coefficient. Bare soil: Medium CN value (55 - 80), with an infiltration coefficient between that of the impervious surface and the forest land, and a relatively high runoff coefficient.

[0063] Taking the HEC-HMS model as an example of the rainfall-runoff model, as Figure 5 shown, the HEC-HMS model is a comprehensive hydrological simulation system used to simulate the rainfall-runoff process and can handle complex watershed characteristics. The specific parameter settings are as follows: Impervious surface: Set a high runoff curve number and a low infiltration parameter. Forest land: Set a low runoff curve number and a high infiltration parameter. Bare soil: Set a medium runoff curve number and infiltration parameter.

[0064] For the surface runoff model, in the embodiments of the present application, a Digital Elevation Model (DEM) can be used for terrain analysis to simulate the flow path and concentration process of surface runoff. Specifically, the Digital Elevation Model can provide high-precision terrain data for analyzing the flow path and direction of surface water. A continuous elevation surface is generated through an interpolation algorithm, and detailed terrain analysis is carried out in combination with slope and aspect data. The terrain analysis includes slope and aspect analysis, and river network extraction. Among them, slope and aspect analysis: Calculate the slope and aspect of the surface using DEM data to determine the flow direction of surface runoff. River network extraction: Identify and extract the main river network through terrain data to simulate the concentration and diffusion process of runoff.

[0065] For the surface runoff model, in the embodiments of the present application, a two-dimensional hydrodynamic model (such as SWMM, Flood Modeler) can be used in combination with rainfall, topography, and runoff paths to construct a model for simulating the flood diffusion process. Among them, SWMM is a model used to simulate urban rainfall-runoff and drainage systems. Model parameters are set according to different underlying surface types to simulate the generation and diffusion process of floods. It is mainly used in urban areas to evaluate the capabilities and bottlenecks of urban drainage systems and predict urban flood risks; Flood Modeler is a professional two-dimensional hydrodynamic model used to simulate the diffusion and propagation process of floods. Combining DEM data and rainfall forecast data, it simulates the flood diffusion process in areas with different underlying surface types and predicts the inundation range and depth. It is applicable to forest land and bare soil areas to simulate the propagation path and speed of floods in natural environments.

[0066] On the other hand, in order to further improve the accuracy of flood risk prediction, the embodiments of the present application can also select a variety of suitable machine learning algorithms for optimization. The specific steps are as follows.

[0067] In some embodiments of the present application, obtaining a machine learning model includes the following steps: obtaining historical flood event data, where the historical flood event data includes: historical rainfall data of the monitoring area, topographic data of the monitoring area, the underlying surface type of the monitoring area, and historical inundation depth; using a machine learning algorithm to extract data features from the historical flood event data and training an initial model based on the data features to obtain a machine learning model. The machine learning algorithm includes at least one of the following: random forest algorithm, gradient boosting decision tree algorithm, and support vector machine algorithm.

[0068] Specifically, in this embodiment, a variety of suitable machine learning algorithms can be selected for optimization. For example, Random Forest: It has strong capabilities in dealing with high-dimensional data and non-linear relationships and can handle multi-source heterogeneous data. It is used to extract and synthesize the characteristics of rainfall, topography, underlying surface, and historical flood data for flood risk prediction; Gradient Boosting Decision Tree (GBDT): It has high prediction accuracy and can effectively handle complex data relationships. It is used to optimize and adjust the parameters of the flood risk prediction model to improve the prediction accuracy; Support Vector Machine (SVM): It is suitable for small sample learning and has good generalization ability. It is used for classification and regression analysis, especially with remarkable effects when dealing with non-linear problems.

[0069] In this embodiment, key features can be extracted from multi-source data such as meteorological data, topographic and geomorphic data, underlying surface data, and historical flood event data, such as rainfall intensity, slope, land use type, historical inundation depth, etc., as the input of the machine learning model. Then, the historical flood event data is used to train the machine learning model. The dataset is divided into a training set and a test set, and the cross-validation method is used to evaluate the model performance. The grid search method is adopted to optimize the model parameters to improve the prediction accuracy and generalization ability of the model.

[0070] After that, the physical process simulation results (physical process simulation model) and the machine learning model are combined to construct an integrated model (flood risk prediction model). The process of simulating rainfall being converted into runoff and inundation is carried out to generate preliminary prediction results. The prediction results of the physical model are further optimized and corrected to improve the overall prediction accuracy.

[0071] Through the flood risk prediction model, prediction and early warning for different underlying surface types can be achieved. Specifically, 1) For the impervious ground type: Rainfall-runoff: Assume that most rainfall is converted into runoff, and high runoff coefficients and low infiltration coefficients are adopted. Surface runoff: Combine urban drainage system data to evaluate drainage capacity and bottlenecks, and simulate urban flood risks. Inundation simulation: Real-time monitor rainfall and the status of the urban drainage system, issue flood warning information, and focus on low-lying areas and areas with poor drainage capacity. 2) For the forest ground type, Rainfall-runoff: Consider higher rainfall interception and evaporation, and adopt low runoff coefficients and high infiltration coefficients. Surface runoff: Simulate the rainfall interception and infiltration processes in the forest, and evaluate the propagation path and speed of floods in the forest. Inundation simulation: Real-time monitor rainfall and soil moisture in the forest, issue flood warning information, and focus on areas with larger slopes and areas prone to landslides. 3) For the bare soil ground type, Rainfall-runoff: Adopt medium runoff coefficients and infiltration coefficients, considering the rainfall infiltration and runoff characteristics of bare soil. Surface runoff: Combine the topographic and geomorphic data of bare soil to simulate the diffusion process of floods in the bare soil area. Inundation simulation: Real-time monitor rainfall and runoff conditions in the bare soil area, issue flood warning information, and focus on areas with larger bare soil areas and areas prone to debris flows.

[0072] The application scenarios of this flood risk prediction model include: 1) Urban areas: It is mainly applied to urban impervious surface areas. Through real-time monitoring and prediction, flood warnings are provided to reduce the impact of urban flood disasters on transmission lines. 2) Forest areas: In forest areas, by monitoring rainfall and soil moisture, flood risks are predicted to prevent the threat of floods and landslides to transmission lines. 3) Bare soil areas: In bare soil areas, through real-time rainfall and runoff monitoring, debris flow and flood risks are warned to protect the safe operation of transmission lines.

[0073] The process of using the flood risk prediction model for risk assessment and grading is introduced as follows.

[0074] In some embodiments of the present application, when using the flood risk prediction model to determine the grid flood risk level of the monitoring area based on the corresponding weather forecast data, underlying surface type, and voltage level of grid equipment, the following steps are included: Obtain the environmental data collected by the sensor devices in the monitoring area, where the sensor devices include at least one of the following: rain gauge, water level gauge, flow meter, soil moisture sensor; Use the flood risk prediction model to predict the corresponding inundation depth and drainage capacity of the drainage system in the monitoring area based on the environmental data and weather forecast data; Use the analytic hierarchy process to determine the weight coefficients corresponding to the inundation depth, underlying surface type, voltage level of grid equipment, and drainage capacity of the drainage system in the monitoring area respectively; Determine the grid flood risk level of the monitoring area based on the inundation depth, underlying surface type, voltage level of grid equipment, drainage capacity of the drainage system, and the corresponding weight coefficients in the monitoring area.

[0075] Specifically, in this embodiment, the main evaluation factors include: inundation depth, underlying surface type, voltage level of transmission lines (voltage level of grid equipment), and drainage capacity of the drainage system. Among them, the inundation depth is one of the key indicators for evaluating flood risk and directly reflects the threat degree of floods to the surface and facilities. The inundation depth is divided into multiple levels: 0 - 0.5 meters, 0.5 - 1 meter, 1 - 2 meters, and above 2 meters. Shallow inundation depths may only affect surface traffic and light facilities, while deeper inundation depths may pose a serious threat to building foundations, power facilities, and personal safety.

[0076] The underlying surface type has an important impact on the evaluation of flood risk. Different types of underlying surfaces have different effects on rainfall interception, infiltration, and runoff. Among them, impervious surfaces: such as urban roads, buildings, etc. Most of the rainfall in these areas is converted into runoff, and the flood risk is relatively high. Woodlands: Have high interception and infiltration capabilities, and the flood risk is relatively low. Bare soil: The infiltration ability is between impervious surfaces and woodlands, and the flood risk is medium.

[0077] The voltage level of grid equipment, such as the voltage level of transmission lines, is an important factor affecting the degree of its susceptibility to floods. Among them, low-voltage lines: Are relatively less affected by floods, but are prone to problems such as short circuits and power outages. Medium-voltage lines: Are more affected by floods and need to be protected with emphasis. High-voltage lines: Once affected by floods, it may lead to large-scale power outages and significant economic losses, and need to be monitored and protected with emphasis.

[0078] The drainage capacity of the drainage system directly affects the formation and duration of floods. Among them, an efficient drainage system, such as advanced drainage facilities in cities, can quickly drain accumulated water and reduce the impact of floods. A general drainage system, such as ordinary drainage facilities in rural areas, has limited drainage capacity and is prone to water accumulation. A non-drainage system, such as some remote areas, lacks effective drainage facilities and has a high flood risk.

[0079] In this embodiment, the analytic hierarchy process can be used to determine the weight coefficients corresponding to the above evaluation factors, as follows. First, establish a hierarchical structure model, including an objective layer, a criterion layer, and an index layer. Among them, the objective layer: flood risk assessment. The criterion layer: inundation depth, underlying surface type, transmission line voltage level, drainage system. The index layer: specific classification and grading standards. Then, organize experts in the fields of hydrology, geography, electricity, etc. to determine the relative importance of each factor through scoring. According to experience and expert opinions, construct a comparison judgment matrix between the factors, that is, convert the scoring results of the experts into a judgment matrix. a ij represents the importance comparison value between the i-th factor and the j-th factor, a ij > 1 indicates that i is more important than j, a ij < 1 indicates that i is less important than j, a ij = 1 indicates that i and j are equally important.

[0080] After that, perform eigenvalue decomposition on the judgment matrix to obtain the eigenvector. The normalized result of the eigenvector is the weight of each factor. The weight reflects the relative importance of each factor in the flood risk assessment. In addition, it is also necessary to judge the consistency of the judgment matrix through the consistency ratio (CR) to ensure the reliability of the judgment. If the CR value is less than 0.1, the judgment matrix has good consistency. Among them, the formula for the consistency index (CI) is as follows:

[0081]

[0082] Among them, λ max is the largest eigenvalue of the judgment matrix, and n is the order of the judgment matrix.

[0083] Random consistency index (RI): Look up the random consistency index according to the order n of the judgment matrix.

[0084] Consistency ratio (CR): Calculate the consistency ratio, and the formula is as follows:

[0085]

[0086] At the same time, set the comprehensive scoring criteria, as follows.

[0087] Inundation depth: 0 - 0.5 meters gets 1 point, 0.5 - 1 meter gets 2 points, 1 - 2 meters gets 3 points, and more than 2 meters gets 4 points.

[0088] Underlying surface type: Impervious surface (such as urban roads, buildings) gets 3 points, bare soil (cultivated land, exposed surface) gets 2 points, and forest land (forest, shrub) gets 1 point.

[0089] Transmission line voltage level: Low-voltage line (35KV and below) gets 1 point, medium-voltage line (110KV) gets 2 points, and high-voltage line (220KV and above) gets 3 points.

[0090] Drainage system: An efficient drainage system (with strong drainage capacity) gets 1 point, a general drainage system (with general drainage capacity) gets 2 points, and no drainage system (with weak drainage capacity) gets 3 points.

[0091] Finally, a weighted comprehensive score is calculated: Comprehensive score = ∑(factor score × weight). The scores of each evaluation factor are weighted according to the weight to calculate the comprehensive scores of each region and facility. And risk grading is carried out according to the final comprehensive score. For example, low risk: comprehensive score of 1 - 5 points, indicating a relatively low flood risk. Medium risk: comprehensive score of 6 - 10 points, indicating a certain flood risk. High risk: comprehensive score of 11 - 15 points, indicating a relatively high flood risk. Extremely high risk: comprehensive score above 16 points, indicating a very serious flood threat.

[0092] By establishing a perfect flood risk grading standard, comprehensively considering multiple factors such as flood inundation depth, underlying surface type, different voltage levels of transmission lines, and drainage systems, scientific and reasonable risk grading is carried out to improve the accuracy and practicality of early warning information. And for the power grid transmission lines, considering the differences in different voltage levels, geographical environment, and underlying surface type of the transmission lines, a targeted flood risk prediction model is proposed to improve the accuracy and reliability of the prediction.

[0093] The embodiments of the present application can update the risk level of each monitoring area in real time and issue early warning information, that is, according to the flood risk level of the power grid, early warning information can be generated and sent to the mobile terminal devices located in the monitoring area.

[0094] In this embodiment, the early warning method can be flexibly extended: According to the characteristics of different regions and facilities, the early warning parameters and strategies are adjusted. For example, in high-risk areas, the monitoring frequency and early warning threshold settings are strengthened; around key facilities, special early warning parameters are set. Combining the flood risk prediction model and the risk grading standard, the flood risk early warning information is updated and released in real time. For example, the early warning information is released through multiple channels such as mobile phone APPs, text messages, and social media to improve the public's response speed and awareness of prevention. For high-risk regions and facilities, specific protection measures are formulated, such as establishing an emergency response plan, equipping necessary flood control facilities and equipment, strengthening public education and training, and improving the accuracy and practicality of the early warning.

[0095] Specifically, environmental data can be monitored in real time by deploying devices such as rain gauges, water level gauges, flow meters, soil moisture sensors, etc. Technologies such as wireless sensor networks (WSN), narrowband Internet of Things (NB-IoT), and 5G are used to achieve stable transmission of sensor data. The collected data is cleaned, denoised, and normalized, and fused with satellite remote sensing data, land classification data, urban pipe network data, etc. Big data technology is used to perform real-time analysis on multi-source data, extract key features, and provide input data for the flood risk prediction model. The input data of the flood risk prediction model is updated in real time to perform rainfall-runoff, surface runoff, and inundation simulations. According to real-time data and risk grading criteria, the flood risk levels of different regions and facilities are dynamically evaluated. Flood risk warning information is released in real time through various channels such as the warning system interface, mobile applications, text messages, and emails to ensure that relevant departments and the public can obtain information in a timely manner. According to the risk warning information, targeted protection measures are taken, with a focus on high-risk regions and facilities, and emergency response and disaster prevention and mitigation work are carried out in a timely manner.

[0096] The solution of this application can improve the accuracy of risk prediction. Specifically, by combining physical process simulation and machine learning optimization, the accuracy of the prediction model is improved. Physical process simulation can describe in detail the transformation process of rainfall-runoff, while the machine learning model can extract features from a large amount of historical data, optimize and correct the prediction results, and improve the prediction accuracy of sudden heavy rainfall events. By combining multi-source meteorological data (such as ground observations, radar, satellite remote sensing, etc.), a meteorological data set with high spatio-temporal resolution is formed, so as to more accurately predict rainfall events. The model parameters are updated in real time, and the model is adjusted by combining the latest observation data and prediction results to improve the prediction accuracy of sudden heavy rainfall events.

[0097] In addition, the underlying surface parameters are refined: in the flood risk prediction model, different parameters are set for simulation according to different underlying surface types (such as impervious surfaces, forests, grasslands, farmlands, etc.). Specifically, different underlying surface types can be accurately identified through high-resolution land use data (such as the classification results of remote sensing images), and their different hydrological response characteristics are reflected in the model. The processes of rainfall interception, infiltration, and evaporation are simulated in detail, considering the interception capacity, infiltration rate, and evaporation rate of different underlying surface types. For example, forests have a higher interception capacity and evaporation volume, while impervious surfaces hardly infiltrate, resulting in more runoff generation. The distributed hydrological model can simulate hydrological processes on a finer spatial scale and reflect the impact of underlying surface heterogeneity on flood risk.

[0098] The solution of this application uses multiple data sources, including satellite remote sensing data, land classification data, topographic and geomorphic data, urban pipe network data, rainfall forecast data, power grid GIS data, and historical flood event data. Through multi-source data fusion technology, comprehensive and accurate environmental information is provided. This multi-source data fusion can more accurately reflect the characteristics of the surface and underlying surface, improving the accuracy of flood risk prediction. For different underlying surface types such as impervious surfaces, forest land, and bare soil, specific parameters and models are set to consider the effects of different underlying surface types on rainfall infiltration, evaporation, and runoff. This refined underlying surface classification and processing method makes the flood risk prediction model more targeted and accurate.

[0099] At the same time, classic hydrological models such as the SCS-CN model and the HEC-HMS model are adopted, and a digital elevation model (DEM) is used for terrain analysis. Combined with two-dimensional hydrodynamic models (SWMM, Flood Modeller) for inundation simulation. Through these advanced physical process simulation technologies, the process of rainfall being converted into runoff and inundation can be accurately simulated, improving the scientificity and accuracy of flood risk prediction. Machine learning algorithms such as random forest, gradient boosting decision tree, and support vector machine are introduced to optimize the flood risk prediction model. Through methods such as feature extraction, model training, and model integration, the accuracy and real-time performance of flood risk prediction are further improved.

[0100] Moreover, a risk assessment and grading method combining comprehensive scoring and the analytic hierarchy process (AHP) is proposed. Considering factors such as inundation depth, underlying surface type, transmission line voltage level, and drainage system, dynamic flood risk assessment and grading are carried out. This method can more scientifically evaluate the flood risks of different regions and facilities, providing more accurate risk grading results. By deploying devices such as rain gauges, water level gauges, flow meters, and soil moisture sensors, a real-time sensor network is constructed. Combined with multi-source data fusion technology, real-time monitoring and analysis of environmental data are achieved. By updating the flood risk prediction model in real time, dynamically evaluating the risk level, and timely issuing early warning information, it is ensured that relevant departments and the public can take emergency response measures in a timely manner to reduce the impact of flood disasters on transmission lines. It is not only applicable to urban impervious surface areas but also to forest land and bare soil areas, with a wide range of application scenarios. Through the comprehensive processing and analysis of different types of underlying surfaces, the generality and applicability of the flood risk prediction method are improved. It can effectively prevent the impact of flood disasters on transmission lines, improve the safety and stability of the power system through accurate flood risk prediction and early warning, and ensure the continuity and reliability of power supply.

[0101] According to the embodiments of this application, an embodiment of a power grid flood risk warning device is also provided. Figure 6 It is a structural schematic diagram of a power grid flood risk warning device provided according to the embodiments of this application. AsFigure 6 As shown, the device includes:

[0102] A physical process simulation module 60, configured to obtain multi-source data and establish a physical process simulation model based on the multi-source data, where the physical process simulation model is used to characterize the hydrological characteristics of regions with different underlying surface types under different rainfall amounts;

[0103] A machine learning optimization module 62, configured to obtain a machine learning model, where the machine learning model is trained based on historical flood event data corresponding to the monitoring area and is used to predict the inundation depth of the monitoring area under different land and precipitation conditions;

[0104] A prediction model integration module 64, configured to integrate the physical process simulation model and the machine learning model to obtain a flood risk prediction model;

[0105] A flood risk warning module 66, configured to use the flood risk prediction model to determine the grid flood risk level of the monitoring area based on the weather forecast data, underlying surface type, and voltage level of the grid equipment corresponding to the monitoring area, where the grid flood risk level is used to characterize the degree of influence of the monitoring area and grid equipment by floods.

[0106] Optionally, the multi-source data includes: satellite remote sensing data, land classification data, topographic and geomorphic data, urban pipe network data, rainfall data, and grid geographic information system data, where the satellite remote sensing data is used to characterize the distribution of surface water bodies and the distribution of rainfall areas, the land classification data is used to characterize the underlying surface types of the surfaces in different regions; the topographic and geomorphic data is used to characterize the slope and aspect of different regions; the urban pipe network data is used to characterize the distribution and drainage capacity of the drainage pipe network, the rainfall data is used to characterize the temporal and spatial distribution of rainfall, and the grid geographic information system data is used to characterize the location distribution and voltage level of grid equipment.

[0107] Optionally, before establishing the physical process simulation model based on the multi-source data, the grid flood risk warning device is further configured to: perform data preprocessing on the multi-source data and perform normalization processing on the multi-source data after data preprocessing, where the data preprocessing includes: outlier removal and missing value filling; perform data fusion operations on the multi-source data after normalization processing, where the data fusion operations include: spatial fusion, temporal fusion, and attribute fusion, spatial fusion is used to superimpose spatial data from different sources in the multi-source data onto the same coordinate system, temporal fusion is used to match and integrate data with different time scales in the multi-source data, and attribute fusion is used to merge attribute information from different sources in the multi-source data.

[0108] Optionally, the physical process simulation model includes at least one of the following: rainfall-runoff model, surface runoff model, inundation simulation model; establishing the physical process simulation model based on multi-source data includes: determining the rainfall-runoff model based on satellite remote sensing data, land classification data, and rainfall data, where the rainfall-runoff model is used to simulate the process of rainfall being converted into surface runoff. In the rainfall-runoff model, the runoff curve numbers and infiltration parameters corresponding to different underlying surface types are different, and the underlying surface types include at least one of the following: impervious ground type, forest land ground type, bare land ground type; determining the surface runoff model based on land classification data, topographic and geomorphic data, and urban pipe network data, where the surface runoff model is used to simulate the flow process of surface water in areas with different underlying surface types; determining the inundation simulation model based on rainfall data, topographic and geomorphic data, and the surface runoff model, where the inundation simulation model is used to simulate the diffusion process of floods in areas with different underlying surface types and predict the inundation range and depth.

[0109] Optionally, obtaining the machine learning model includes: obtaining historical flood event data, where the historical flood event data includes: historical rainfall data of the monitoring area, topographic data of the monitoring area, the underlying surface type of the monitoring area, and historical inundation depth; using machine learning algorithms to extract data features from the historical flood event data and training the initial model based on the data features to obtain the machine learning model, where the machine learning algorithms include at least one of the following: random forest algorithm, gradient boosting decision tree algorithm, support vector machine algorithm.

[0110] Optionally, using the flood risk prediction model to determine the grid flood risk level of the monitoring area based on the weather forecast data corresponding to the monitoring area, the underlying surface type, and the voltage level of the grid equipment includes: obtaining the environmental data collected by the sensor devices in the monitoring area, where the sensor devices include at least one of the following: rain gauge, water level gauge, flow meter, soil moisture sensor; using the flood risk prediction model to predict the inundation depth and drainage system drainage capacity corresponding to the monitoring area based on the environmental data and the weather forecast data; using the analytic hierarchy process to determine the weight coefficients corresponding to the inundation depth, underlying surface type, voltage level of the grid equipment, and drainage system drainage capacity of the monitoring area respectively; determining the grid flood risk level of the monitoring area based on the inundation depth, underlying surface type, voltage level of the grid equipment, drainage system drainage capacity, and the corresponding weight coefficients of the monitoring area.

[0111] Optionally, after determining the grid flood risk level of the monitoring area, the grid flood risk warning device is further used to: generate a warning message based on the grid flood risk level and send the warning message to the mobile terminal device located in the monitoring area.

[0112] It should be noted that each module in the above power grid flood risk warning device can be a program module (for example, a set of program instructions that implement a specific function), or a hardware module. For the latter, it can be presented in the following forms, but not limited to: the manifestation form of each of the above modules is a processor, or the functions of each of the above modules are implemented by a processor.

[0113] It should be noted that the power grid flood risk warning device provided in this embodiment can be used to execute Figure 2 the power grid flood risk warning method shown. Therefore, the relevant explanations of the above power grid flood risk warning method also apply to the embodiments of this application and will not be elaborated here.

[0114] The embodiments of this application also provide a non-volatile storage medium. The non-volatile storage medium includes a stored computer program. Among them, the device where the non-volatile storage medium is located executes the following power grid flood risk warning method by running the computer program: obtaining multi-source data, and based on the multi-source data, establishing a physical process simulation model, where the physical process simulation model is used to characterize the hydrological characteristics of regions with different underlying surface types under different rainfall amounts; obtaining a machine learning model, where the machine learning model is trained based on historical flood event data corresponding to the monitoring area and is used to predict the inundation depth of the monitoring area under different land and precipitation conditions; integrating the physical process simulation model and the machine learning model to obtain a flood risk prediction model; using the flood risk prediction model, based on the weather forecast data, underlying surface type, and voltage level of the power grid equipment corresponding to the monitoring area, determining the power grid flood risk level of the monitoring area, where the power grid flood risk level is used to characterize the degree of influence of the monitoring area and the power grid equipment by floods.

[0115] The embodiments of this application also provide a computer program product, including a computer program. When the computer program is executed by a processor, it implements the steps of the power grid flood risk warning method described in each embodiment of this application: obtaining multi-source data, and based on the multi-source data, establishing a physical process simulation model, where the physical process simulation model is used to characterize the hydrological characteristics of regions with different underlying surface types under different rainfall amounts; obtaining a machine learning model, where the machine learning model is trained based on historical flood event data corresponding to the monitoring area and is used to predict the inundation depth of the monitoring area under different land and precipitation conditions; integrating the physical process simulation model and the machine learning model to obtain a flood risk prediction model; using the flood risk prediction model, based on the weather forecast data, underlying surface type, and voltage level of the power grid equipment corresponding to the monitoring area, determining the power grid flood risk level of the monitoring area, where the power grid flood risk level is used to characterize the degree of influence of the monitoring area and the power grid equipment by floods.

[0116] The serial numbers of the embodiments of the present application above are only for description and do not represent the superiority or inferiority of the embodiments.

[0117] In the above embodiments of the present application, the descriptions of the respective embodiments have their own emphases. For the parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0118] In the several embodiments provided by the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of the units can be a logical function division. In actual implementation, there can be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of units or modules can be in an electrical or other form.

[0119] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0120] In addition, the functional units in the various embodiments of the present application can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0121] If the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. And the aforementioned storage medium includes: USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs and other various media that can store program codes.

[0122] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present application.

Claims

1. A power grid flood risk early warning method, characterized in that: include: Acquire multi-source data, and establish a physical process simulation model based on the multi-source data, wherein the physical process simulation model is used to characterize the hydrological characteristics of areas with different underlying surface types under different rainfall amounts; Obtaining a machine learning model, wherein the machine learning model is trained based on historical flood event data corresponding to the monitoring area, and is used to predict the flooding depth of the monitoring area under different land and precipitation conditions; Integrating the physical process simulation model and the machine learning model to obtain a flood risk prediction model; The flood risk prediction model is used to determine the power grid flood risk level of the monitoring area based on the weather forecast data corresponding to the monitoring area, the underlying surface type, and the voltage level of the power grid equipment, wherein the power grid flood risk level is used to characterize the extent to which the monitoring area and power grid equipment are affected by floods.

2. The power grid flood risk early warning method according to claim 1 is characterized in that: The multi-source data includes: satellite remote sensing data, land classification data, topographic data, urban pipe network data, rainfall data, and power grid geographic information system data, wherein the satellite remote sensing data is used to characterize the distribution of surface water bodies and the distribution of rainfall areas, and the land classification data is used to characterize the underlying surface type of the surface in different areas; the topographic data is used to characterize the slope and slope direction in different areas; the urban pipe network data is used to characterize the distribution and drainage capacity of the drainage network, the rainfall data is used to characterize the temporal and spatial distribution of rainfall, and the power grid geographic information system data is used to characterize the location distribution and voltage level of power grid equipment.

3. The power grid flood risk early warning method according to claim 2 is characterized in that: Before establishing a physical process simulation model based on the multi-source data, the method further includes: Performing data preprocessing on the multi-source data, and normalizing the multi-source data after the data preprocessing, wherein the data preprocessing includes: outlier removal and missing value filling; A data fusion operation is performed on the multi-source data after the normalization processing, wherein the data fusion operation includes: spatial fusion, temporal fusion, and attribute fusion. The spatial fusion is used to superimpose spatial data from different sources in the multi-source data into the same coordinate system, the temporal fusion is used to match and integrate data of different time scales in the multi-source data, and the attribute fusion is used to merge attribute information from different sources in the multi-source data.

4. The power grid flood risk early warning method according to claim 2, characterized in that: The physical process simulation model includes at least one of the following: a rainfall runoff model, a surface runoff model, and a flooding simulation model; establishing the physical process simulation model based on the multi-source data includes: Determine the rainfall runoff model based on the satellite remote sensing data, the land classification data, and the rainfall data, wherein the rainfall runoff model is used to simulate the process of converting rainfall into surface runoff, and in the rainfall runoff model, different underlying surface types correspond to different runoff curve numbers and infiltration parameters, and the underlying surface type includes at least one of the following: impermeable ground type, forest ground type, and bare soil ground type; Determining the surface runoff model based on the land classification data, the topographic data, and the urban pipe network data, wherein the surface runoff model is used to simulate the flow process of surface water in areas under different underlying surface types; The flood simulation model is determined based on the rainfall data, the topographic data, and the surface runoff model, wherein the flood simulation model is used to simulate the diffusion process of floods in areas with different underlying surface types and predict the flood range and depth.

5. The power grid flood risk early warning method according to claim 1, characterized in that: The obtaining of the machine learning model comprises: Acquiring the historical flood event data, wherein the historical flood event data includes: historical rainfall data of the monitoring area, terrain data of the monitoring area, underlying surface type of the monitoring area, and historical flooding depth; A machine learning algorithm is used to extract data features from the historical flood event data, and an initial model is trained based on the data features to obtain the machine learning model, wherein the machine learning algorithm includes at least one of the following: a random forest algorithm, an extraction boosting decision tree algorithm, and a support vector machine algorithm.

6. The power grid flood risk early warning method according to claim 1, characterized in that: Using the flood risk prediction model, according to the weather forecast data corresponding to the monitoring area, the underlying surface type, and the voltage level of the power grid equipment, determining the flood risk level of the power grid in the monitoring area includes: Acquire environmental data collected by sensor equipment in the monitoring area, wherein the sensor equipment includes at least one of the following: a rain gauge, a water level meter, a flow meter, and a soil moisture sensor; Using the flood risk prediction model, based on the environmental data and the weather forecast data, predict the flood depth and drainage capacity of the drainage system corresponding to the monitoring area; Using the analytic hierarchy process, determine the weight coefficients corresponding to the flooding depth, the underlying surface type, the voltage level of the power grid equipment, and the drainage capacity of the drainage system corresponding to the monitoring area; The grid flood risk level of the monitoring area is determined based on the flood depth corresponding to the monitoring area, the underlying surface type, the voltage level of the grid equipment, the drainage capacity of the drainage system, and the corresponding weight coefficient.

7. The power grid flood risk early warning method according to claim 6, characterized in that: After determining the flood risk level of the power grid in the monitoring area, the method further includes: According to the flood risk level of the power grid, early warning information is generated, and the early warning information is sent to a mobile terminal device located in the monitoring area.

8. A power grid flood risk warning device, characterized in that: include: A physical process simulation module, used to obtain multi-source data and establish a physical process simulation model based on the multi-source data, wherein the physical process simulation model is used to characterize the hydrological characteristics of areas with different underlying surface types under different rainfall amounts; A machine learning optimization module, used to obtain a machine learning model, wherein the machine learning model is trained based on historical flood event data corresponding to the monitoring area, and is used to predict the flooding depth of the monitoring area under different land and precipitation conditions; A prediction model integration module, used to integrate the physical process simulation model and the machine learning model to obtain a flood risk prediction model; The flood risk warning module is used to adopt the flood risk prediction model to determine the power grid flood risk level of the monitoring area according to the weather forecast data corresponding to the monitoring area, the underlying surface type, and the voltage level of the power grid equipment, wherein the power grid flood risk level is used to characterize the degree to which the monitoring area and power grid equipment are affected by floods.

9. A non-volatile storage medium, characterized in that: The non-volatile storage medium includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the power grid flood risk warning method as described in any one of claims 1 to 7 by running the computer program.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the power grid flood risk warning method described in any one of claims 1 to 7 are implemented.

Images