A flood disaster emergency simulation method and system based on digital twin

Abstract

The present invention relates to the technical field of flood disaster management, and specifically to a flood disaster emergency simulation method and system based on digital twins, comprising the following steps: S1, data collection and preprocessing: real-time collection of environmental data related to flood disasters; S2, establishment of digital twin models: establishment of digital twin models of flood disaster areas; S3, simulation calculation: prediction of flood propagation paths, water depths and flow rates; S4, flood propagation path optimization: identification of patterns and anomalies; S5, emergency plan formulation and real-time monitoring: formulation of emergency plans in combination with historical flood disaster response experience; S6, evaluation and improvement: after the emergency response to flood disasters is completed, the execution effect of the emergency plan is evaluated, lessons learned are summarized, and the digital twin model and emergency plan are optimized. The present invention realizes accurate prediction and efficient response to flood disasters.

Description

Technical Field

[0001] The present invention relates to the technical field of flood disaster management, and in particular to a flood disaster emergency simulation method and system based on digital twins. Background Art

[0002] Flood disasters not only affect people's daily lives, but also cause huge losses to urban infrastructure, agricultural production and the ecological environment. Therefore, how to efficiently and accurately predict flood disasters and formulate scientific emergency plans has become an important research topic in the current field of disaster prevention and mitigation.

[0003] Existing flood disaster emergency management methods usually rely on historical data and expert experience for prediction and decision-making, lacking the support of real-time data and accurate simulation models. In addition, traditional methods have problems such as poor information transmission, inflexible command and dispatch, and difficulty in data integration during the emergency response process, resulting in unsatisfactory implementation of emergency plans and inability to effectively respond to complex and changeable flood disaster scenarios, making it difficult to control disaster losses.

[0004] The present invention aims to address the deficiencies in the prior art and provide a flood disaster emergency simulation method based on digital twins to enhance the ability to respond to flood disasters and reduce the threat of disasters to life and property. Summary of the invention

[0005] Based on the above objectives, the present invention provides a flood disaster emergency simulation method and system based on digital twins.

[0006] A flood disaster emergency simulation method based on digital twins includes the following steps:

[0007] S1, data collection and preprocessing: collect environmental data related to flood disasters in real time through a variety of sensors and satellite remote sensing technology, including rainfall, water level, flow rate and terrain data, and preprocess the environmental data;

[0008] S2, build a digital twin model: Based on the collected environmental data, use the geographic information platform (GIS), computer simulation and modeling technology to build a digital twin model of the flood disaster area, including topography, hydrological characteristics and building distribution;

[0009] S3, simulation calculation: Based on the digital twin model, the flood propagation path prediction model is used to simulate different flood disaster scenarios and predict the flood propagation path, water depth and flow rate;

[0010] S4, Flood propagation path optimization: Integrate meteorological forecast data, historical flood data and environmental data to optimize the prediction of flood propagation paths and identify patterns and anomalies, including:

[0011] S41, multi-source heterogeneous data fusion: fusion of meteorological forecast data, historical flood data and environmental data into multi-source heterogeneous data;

[0012] S42, Optimal prediction of flood propagation paths: Using artificial neural network (ANN) model to optimize flood propagation paths based on fused multi-source heterogeneous data;

[0013] S43, Pattern and Anomaly Recognition: Identify patterns and anomalies in multi-source heterogeneous data and adjust artificial neural network (ANN) models in real time;

[0014] S5, emergency plan formulation and real-time monitoring: According to the results of simulation calculation and flood propagation path optimization, combined with historical flood disaster response experience, emergency plans are formulated, and the latest environmental data of the flood disaster site is monitored in real time. The emergency plan is dynamically adjusted according to the real-time environmental data and simulation results;

[0015] S6, Evaluation and Improvement: After the emergency response to flood disasters is completed, the implementation effect of the emergency plan is evaluated, lessons learned are summarized, and the digital twin model and emergency plan are optimized.

[0016] Optionally, the data collection and preprocessing in S1 includes:

[0017] S11, data collection: real-time collection of environmental data related to flood disasters, including:

[0018] Rainfall collection: real-time measurement of rainfall through rain gauge sensors;

[0019] Water level collection: Measure water level changes through ultrasonic water level sensors;

[0020] Flow velocity collection: water flow velocity is measured by Doppler flowmeter;

[0021] Topographic data collection: Obtain terrain data through satellite remote sensing technology and use digital elevation model (DEM) to process terrain height information;

[0022] S12, data preprocessing: preprocessing the environmental data collected in real time, specifically including:

[0023] Data cleaning: Use median filtering algorithm to remove noise;

[0024] Format conversion: Use linear interpolation method to convert environmental data in different formats into a standard format;

[0025] Missing data imputation: Lagrange interpolation method is used to fill missing data.

[0026] Optionally, establishing the digital twin model in S2 includes:

[0027] S21, data input and registration: import the pre-processed environmental data into the geographic information platform and perform spatial registration and projection conversion;

[0028] S22, spatial analysis: using the spatial analysis function of the geographic information platform to extract topographic features, analyze hydrological features (such as watershed division and river network generation), and overlay building distribution data;

[0029] S23, Topography modeling: Construct topography based on digital elevation model (DEM), and use triangulated irregular network (TIN) method to represent the terrain surface, which is expressed as:

[0030] ;

[0031] in, Represents the terrain height, and Represents geographic coordinates;

[0032] S24, hydrological characteristics modeling: The river water level and flow rate are simulated using hydrological simulation software (such as HEC-RAS), and the Saint-Venant equation (shallow water equation) is used for numerical calculation, which is expressed as:

[0033] ;

[0034] ;

[0035] ;

[0036] in, For water depth, and The water flow and The velocity component in the direction, is the acceleration due to gravity, is the terrain height, and is the bottom friction, is the source term (such as rainfall or runoff);

[0037] S25, building distribution modeling: Using terrain data (water system, landform, slope) and building distribution data (building location, height, shape), the building is modeled as an impermeable area, which affects the flow path of water and the water accumulation area, expressed as:

[0038] ;

[0039] in, The coordinates of the vertices representing the building boundary.

[0040] Optionally, the simulation calculation in S3 includes:

[0041] S31, initial condition setting: according to the environmental data in the digital twin model, the initial conditions for flood propagation are set, including initial water depth, flow velocity and terrain elevation;

[0042] S32, boundary condition setting: according to the actual flood disaster scenario, the boundary conditions of the simulation calculation are set, including rainfall intensity, rainfall duration and water inflow flow;

[0043] S33, finite volume method solution: The finite volume method (FVM) is used for simulation calculation to predict the propagation path, water depth and flow rate of the flood, which can be expressed as:

[0044] Conservation equations: ;

[0045] in, is the conserved variable vector, is the flux vector, is the source term vector;

[0046] Discretization: Divide the computational domain into several small finite volume units, integrate and discretize each unit, and express it as:

[0047] ;

[0048] in, Yes Unit The conserved variable, is the unit volume, is the unit boundary The flux at is the boundary area, is the source term of the unit;

[0049] Flux calculation: The flux at the cell boundary is calculated using the Riemann solver , update the conserved variables of each unit , expressed as:

[0050] ;

[0051] in, is the flux at the cell boundary, and are the fluxes of the left and right states, respectively, and are the conserved variables on the left and right sides, is the maximum characteristic velocity (wave speed).

[0052] Optionally, the multi-source heterogeneous data fusion in S41 includes:

[0053] S411, data normalization: normalize weather forecast data, historical flood data, and environmental data;

[0054] S412, data alignment and registration: align and register weather forecast data, historical flood data, and environmental data using timestamps and spatial coordinates;

[0055] S413, data fusion: Use the weighted average method to fuse multi-source data, combine the weights of different data sources, and calculate the comprehensive value, expressed as:

[0056] ;

[0057] in, is the fused data, For data source The weight of For data source The data value of is the number of data sources;

[0058] S414, verification and correction of fusion results: Verify and correct the fused multi-source heterogeneous data, and correct errors by comparing and analyzing them with the actual observed data.

[0059] Optionally, the artificial neural network (ANN) model in S42 includes:

[0060] S421, multi-source data feature fusion: feature fusion of weather forecast data, historical flood data and environmental data, design feature fusion layer, expressed as:

[0061] ;

[0062] in, , and are the feature vectors of weather forecast data, historical flood data, and environmental data, respectively;

[0063] S422, time series feature processing: extract time series features through recurrent neural network (RNN), expressed as:

[0064] ;

[0065] in, For the moment The hidden state of For the moment The fusion characteristics of For the moment The hidden state of and is the weight matrix, is the bias, is the ReLU activation function;

[0066] S423, spatial feature extraction: Use convolutional neural network (CNN) to extract spatial features, expressed as:

[0067] ;

[0068] in, is the output feature matrix Middle position The value of is the convolution kernel Middle position The value of is the input feature matrix The value of the corresponding position in ;

[0069] S424, fully connected layer: The temporal features and spatial features are fused through the fully connected layer and the prediction results are output, which are expressed as:

[0070] ;

[0071] in, To output the prediction results, and are the weight matrices of temporal features and spatial features respectively, is the bias, is the extracted time series feature, is the extracted spatial features;

[0072] S425, optimized loss function: A weighted loss function is used to weight the importance of different data sources, expressed as:

[0073] ;

[0074] in, and is the weight coefficient, is the predicted value, is the true value, is the number of samples.

[0075] Optionally, the pattern and anomaly identification in S43 includes:

[0076] S431, Feature Extraction: Use principal component analysis (PCA) to extract features from multi-source heterogeneous data and identify patterns in the data, expressed as:

[0077] ;

[0078] in, is the original data matrix, and is an orthogonal matrix, is a diagonal matrix;

[0079] S432, anomaly detection: Use a distance-based method to calculate the distance between each data point and its nearest neighbor to detect outliers in the data, expressed as:

[0080] ;

[0081] in, For data points The distance For data points In the The values ​​in the dimensions, For the The average value of the dimensions;

[0082] S433, real-time model adjustment: using the online gradient descent method, the parameters of the artificial neural network (ANN) model are adjusted in real time according to the detected abnormal values, which is expressed as:

[0083] ;

[0084] in, For the moment The model parameters, is the learning rate, is the loss function.

[0085] Optionally, the emergency plan formulation and real-time monitoring in S5 include:

[0086] S51, emergency plan formulation: Based on the results of simulation calculations and flood propagation path optimization, combined with historical flood disaster response experience, formulate emergency plans, including personnel evacuation routes, material allocation plans, and rescue force deployment;

[0087] S52, real-time monitoring: through sensors and satellite remote sensing technology, real-time monitoring of the latest environmental data at the flood disaster site;

[0088] S53, dynamic adjustment: according to the real-time environmental data and simulation results, the emergency plan is dynamically adjusted using an adaptive adjustment algorithm, which is expressed as:

[0089] ;

[0090] in, For the moment The plan, For the moment The plan, is the adjustment factor, To monitor the results in real time.

[0091] Optionally, the evaluation and improvement in S6 includes:

[0092] S61, Data Collection: After the emergency response to flood disasters is completed, collect the effectiveness data of the emergency plan implementation process, including personnel evacuation time, material deployment efficiency, and rescue force response speed;

[0093] S62, Implementation effect evaluation: Use performance indicators (KPIs) to evaluate the implementation effect of the emergency plan;

[0094] S63, Summarize lessons learned: Based on the results of the implementation effect evaluation, summarize the successful experiences and shortcomings in the emergency plan, and identify bottlenecks and problems in the emergency response;

[0095] S64, Optimize the digital twin model: Based on the lessons learned, optimize the digital twin model, expressed as:

[0096] ;

[0097] in, For the moment The model parameters, is the learning rate, Based on input data and target data The loss function of

[0098] S65, Optimize emergency plans: Based on the optimized digital twin model, re-formulate emergency plans, including adjusting evacuation routes, optimizing material allocation plans, updating rescue force deployment, improving information transmission processes, strengthening emergency drills and training, updating key steps and processes, increasing the flexibility of resource allocation, and strengthening coordination and cooperation with relevant departments.

[0099] A flood disaster emergency simulation system based on digital twins is used to implement the above-mentioned flood disaster emergency simulation method based on digital twins, and includes the following modules:

[0100] Data collection and preprocessing module: collects environmental data related to flood disasters in real time through a variety of sensors and satellite remote sensing technology, including rainfall, water level, flow rate and terrain data, and preprocesses the environmental data;

[0101] Digital twin model building module: Based on the collected environmental data, using geographic information platform, computer simulation and modeling technology, a digital twin model of the flood disaster area is established, including topography, hydrological characteristics and building distribution;

[0102] Simulation calculation module: Based on the digital twin model, the flood propagation path prediction model is used to simulate and calculate different flood disaster scenarios, and predict the flood propagation path, water depth and flow rate;

[0103] Flood propagation path optimization module: integrates meteorological forecast data, historical flood data and environmental data to optimize the prediction of flood propagation paths, identify patterns and anomalies, and adjust the flood propagation path prediction model in real time;

[0104] Emergency plan formulation and real-time monitoring module: Based on the results of simulation calculation and flood propagation path optimization, combined with historical flood disaster response experience, emergency plans are formulated, and the latest environmental data of the flood disaster site is monitored in real time. The emergency plan is dynamically adjusted based on real-time environmental data and simulation results;

[0105] Evaluation and Improvement Module: After the emergency response to flood disasters is completed, the implementation effect of the emergency plan will be evaluated, lessons learned will be summarized, and the digital twin model and emergency plan will be optimized.

[0106] Beneficial effects of the present invention:

[0107] The present invention realizes accurate prediction and efficient response to flood disasters through systematic data collection and preprocessing, establishment of digital twin models, simulation calculations and optimization of flood propagation paths, collects environmental data related to flood disasters in real time, and ensures data accuracy and completeness through data cleaning, format conversion and missing data filling, thereby providing a reliable data foundation for the establishment of digital twin models and simulation calculations. Through the establishment of digital twin models, the topography, hydrological characteristics and building distribution can be accurately reflected, providing precise model support for simulation calculations.

[0108] The present invention integrates meteorological forecast data, historical flood data and environmental data, uses an artificial neural network model to optimize the prediction of flood propagation paths, identifies patterns and anomalies in the data in real time, and dynamically adjusts model parameters, thereby improving the accuracy and timeliness of the prediction. At the same time, based on simulation calculations and optimization results, a detailed emergency plan is formulated, and environmental data is monitored in real time through sensors and satellite remote sensing technology. The emergency plan is dynamically adjusted according to real-time data, which significantly improves the efficiency and accuracy of emergency response.

[0109] The present invention, through systematic evaluation and improvement, uses performance indicators to evaluate the effectiveness of plan execution, summarizes successful experiences and shortcomings, optimizes digital twin models and emergency plans, and improves future emergency response capabilities. By collecting data, summarizing experiences and lessons, and optimizing digital twin models and emergency plans, the present invention realizes scientific and intelligent emergency management, provides valuable guidance and reference for responding to future flood disasters, and reduces the threat of disasters to life and property. BRIEF DESCRIPTION OF THE DRAWINGS

[0110] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0111] Figure 1 A schematic diagram of a simulation method flow chart of an embodiment of the present invention;

[0112] Figure 2 Schematic diagram of system function modules according to an embodiment of the present invention. DETAILED DESCRIPTION

[0113] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0114] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiments", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0115] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0116] like Figure 1 As shown, a flood disaster emergency simulation method based on digital twins includes the following steps:

[0117] S1, data collection and preprocessing: collect environmental data related to flood disasters in real time through a variety of sensors and satellite remote sensing technology, including rainfall, water level, flow rate and terrain data, and preprocess the environmental data to ensure the accuracy and integrity of the data;

[0118] S2, build a digital twin model: Based on the collected environmental data, use the geographic information platform (GIS), computer simulation and modeling technology to build a digital twin model of the flood disaster area, including topography, hydrological characteristics and building distribution;

[0119] S3, simulation calculation: Based on the digital twin model, the flood propagation path prediction model is used to simulate different flood disaster scenarios and predict the flood propagation path, water depth and flow rate;

[0120] S4, Flood propagation path optimization: Integrate meteorological forecast data, historical flood data and environmental data to optimize the prediction of flood propagation paths, identify patterns and anomalies, and improve the accuracy and timeliness of predictions, including:

[0121] S41, multi-source heterogeneous data fusion: fusion of meteorological forecast data, historical flood data and environmental data into multi-source heterogeneous data;

[0122] S42, Optimal prediction of flood propagation paths: Using artificial neural network (ANN) model to optimize flood propagation paths based on fused multi-source heterogeneous data;

[0123] S43, Pattern and Anomaly Recognition: Identify patterns and anomalies in multi-source heterogeneous data and adjust artificial neural network (ANN) models in real time;

[0124] S5, emergency plan formulation and real-time monitoring: According to the results of simulation calculation and flood propagation path optimization, combined with historical flood disaster response experience, emergency plans are formulated, and the latest environmental data of the flood disaster site is monitored in real time. According to the real-time environmental data and simulation results, the emergency plan is dynamically adjusted to improve the effectiveness of emergency response;

[0125] S6, Evaluation and Improvement: After the emergency response to flood disasters is completed, the implementation effect of the emergency plan will be evaluated, lessons learned will be summarized, and the digital twin model and emergency plan will be optimized to enhance the ability of future emergency response;

[0126] Through the above steps, accurate prediction and efficient response to flood disasters have been achieved, which has significantly improved the accuracy, real-time and effectiveness of emergency response, reduced the threat of disasters to life and property, and improved the scientific and intelligent level of disaster management.

[0127] Data collection and preprocessing in S1 include:

[0128] S11, data collection: real-time collection of environmental data related to flood disasters, including:

[0129] Rainfall collection: real-time measurement of rainfall through rain gauge sensors;

[0130] Water level collection: Measure water level changes through ultrasonic water level sensors;

[0131] Flow velocity collection: water flow velocity is measured by Doppler flowmeter;

[0132] Topographic data collection: Obtain terrain data through satellite remote sensing technology and use digital elevation model (DEM) to process terrain height information;

[0133] S12, data preprocessing: preprocessing the environmental data collected in real time, specifically including:

[0134] Data cleaning: Use the median filter algorithm to remove noise, expressed as:

[0135] y i = median ( x [ i - k ], … , x i , … , x i + k ) ;

[0136] in, For the cleaned data, is the original data, is the filter window size;

[0137] Format conversion: The environmental data in different formats are converted into a standard format using linear interpolation method, which is expressed as:

[0138] ;

[0139] in, is the current time point, and is the time of adjacent data points, and is the value of the adjacent data point;

[0140] Missing data filling: Use Lagrange interpolation to fill missing data, expressed as:

[0141] ;

[0142] in, is the interpolation function, and is a known data point;

[0143] Through the above content, the accuracy and completeness of the data are guaranteed, thus providing a reliable data foundation for the subsequent digital twin model establishment and simulation calculations.

[0144] Digital elevation model (DEM) includes:

[0145] Interpolation algorithm: The Kriging interpolation algorithm is used to generate a continuous terrain surface, expressed as:

[0146] ;

[0147] in, is the weight coefficient, obtained by solving the Kriging equations, is the elevation value of the known point;

[0148] Data filtering: Gaussian filtering algorithm is used to eliminate noise and outliers, expressed as:

[0149] ;

[0150] in, is the elevation value after filtering, is the Gaussian kernel function, is the window size;

[0151] ;

[0152] in, is the standard deviation, controlling the degree of smoothing.

[0153] The digital twin model building in S2 includes:

[0154] S21, data input and registration: import the pre-processed environmental data into the geographic information platform, and perform spatial registration and projection conversion to ensure that the data is analyzed in the same coordinate system;

[0155] S22, spatial analysis: using the spatial analysis function of the geographic information platform to extract topographic features, analyze hydrological features (such as watershed division and river network generation), and overlay building distribution data;

[0156] S23, Topography modeling: Construct topography based on digital elevation model (DEM), and use triangulated irregular network (TIN) method to represent the terrain surface, which is expressed as:

[0157] ;

[0158] in, Represents the terrain height, and Represents geographic coordinates;

[0159] S24, hydrological characteristics modeling: The river water level and flow rate are simulated using hydrological simulation software (such as HEC-RAS), and the Saint-Venant equation (shallow water equation) is used for numerical calculation, which is expressed as:

[0160] ;

[0161] ;

[0162] ;

[0163] in, For water depth, and The water flow and The velocity component in the direction, is the acceleration due to gravity, is the terrain height, and is the bottom friction, is the source term (such as rainfall or runoff);

[0164] S25, building distribution modeling: Using terrain data (water system, landform, slope) and building distribution data (building location, height, shape), the building is modeled as an impermeable area, which affects the flow path of water and the water accumulation area, expressed as:

[0165] ;

[0166] in, The coordinates of the vertices representing the building boundary;

[0167] Through the above content, it is possible to accurately reflect environmental changes, optimize flood propagation path predictions, and provide a scientific basis for the formulation and dynamic adjustment of emergency plans, thereby significantly improving the efficiency and accuracy of emergency response and reducing the threat of disasters to life and property.

[0168] The simulation calculations in S3 include:

[0169] S31, initial condition setting: according to the environmental data in the digital twin model, the initial conditions for flood propagation are set, including initial water depth, flow velocity and terrain elevation;

[0170] S32, boundary condition setting: according to the actual flood disaster scenario, the boundary conditions of the simulation calculation are set, including rainfall intensity, rainfall duration and water inflow flow;

[0171] S33, finite volume method solution: The finite volume method (FVM) is used for simulation calculation to predict the propagation path, water depth and flow rate of the flood, which can be expressed as:

[0172] Conservation equations: ;

[0173] in, is the conserved variable vector, is the flux vector, is the source term vector;

[0174] Discretization: Divide the computational domain into several small finite volume units, integrate and discretize each unit, and express it as:

[0175] ;

[0176] in, Yes Unit The conserved variable, is the unit volume, is the unit boundary The flux at is the boundary area, is the source term of the unit;

[0177] Flux calculation: The flux at the cell boundary is calculated using the Riemann solver , update the conserved variables of each unit , expressed as:

[0178] ;

[0179] in, is the flux at the cell boundary, and are the fluxes of the left and right states, respectively, and are the conserved variables on the left and right sides, is the maximum characteristic velocity (wave speed), usually the maximum value between the left and right states;

[0180] Through the above steps, not only the accuracy and stability of the calculation are improved, but also complex boundary conditions and initial conditions can be handled, providing a reliable scientific basis for the formulation and dynamic adjustment of emergency plans, thereby significantly improving the efficiency and effectiveness of emergency response and reducing the threat of disasters to life and property.

[0181] Multi-source heterogeneous data fusion in S41 includes:

[0182] S411, data normalization: normalize the weather forecast data, historical flood data and environmental data so that they can be compared and analyzed on the same scale, expressed as:

[0183] ;

[0184] in, is the original data, is the normalized data, and are the minimum and maximum values ​​of the data respectively;

[0185] S412, data alignment and registration: Use timestamps and spatial coordinates to align and register weather forecast data, historical flood data, and environmental data to ensure the consistency of data from different sources in time and space, expressed as:

[0186] ;

[0187] in, is the current time point, and is the time of adjacent data points, and is the value of the adjacent data point;

[0188] S413, data fusion: Use the weighted average method to fuse multi-source data, combine the weights of different data sources, and calculate the comprehensive value, expressed as:

[0189] ;

[0190] in, is the fused data, For data source The weight of For data source The data value of is the number of data sources;

[0191] S414, verification and correction of fusion results: Verify and correct the fused multi-source heterogeneous data, compare and analyze with the actual observation data, correct errors, and improve the accuracy of data fusion, which is expressed as:

[0192] ;

[0193] ;

[0194] in, is the fused data, is the actual observation data, is the number of data points, is the corrected data, is the correction factor;

[0195] Through the above steps, the fusion results can be effectively evaluated and corrected, the accuracy and reliability of the data can be improved, and a more comprehensive, accurate and reliable data basis can be provided for flood propagation path optimization and emergency response, thereby significantly improving the scientificity and effectiveness of emergency decision-making.

[0196] The artificial neural network (ANN) models in S42 include:

[0197] S421, multi-source data feature fusion: feature fusion of weather forecast data, historical flood data and environmental data, design feature fusion layer, expressed as:

[0198] ;

[0199] in, , and are the feature vectors of weather forecast data, historical flood data, and environmental data, respectively;

[0200] S422, time series feature processing: extract time series features through recurrent neural network (RNN), expressed as:

[0201] ;

[0202] in, For the moment The hidden state of For the moment The fusion characteristics of For the moment The hidden state of and is the weight matrix, is the bias, is the ReLU activation function;

[0203] S423, spatial feature extraction: Use convolutional neural network (CNN) to extract spatial features to improve the model's understanding of spatial data such as terrain and landforms, expressed as:

[0204] ;

[0205] in, is the output feature matrix Middle position The value of is the convolution kernel Middle position The value of is the input feature matrix The value of the corresponding position in ;

[0206] S424, fully connected layer: The temporal features and spatial features are fused through the fully connected layer and the prediction results are output, which are expressed as:

[0207] ;

[0208] in, To output the prediction results, and are the weight matrices of temporal features and spatial features respectively, is the bias, is the extracted time series feature, is the extracted spatial features;

[0209] S425, optimized loss function: A weighted loss function is used to weight the importance of different data sources to improve the accuracy and robustness of the model, which is expressed as:

[0210] ;

[0211] in, and is the weight coefficient, is the predicted value, is the true value, is the number of samples;

[0212] Through the above content, we can better integrate and understand complex spatiotemporal characteristics, improve the accuracy and robustness of predictions, and thus provide a reliable scientific basis for the formulation and dynamic adjustment of emergency plans, significantly improve the efficiency and effectiveness of emergency response, and reduce the threat of disasters to life and property.

[0213] Pattern and anomaly recognition in S43 includes:

[0214] S431, Feature Extraction: Use principal component analysis (PCA) to extract features from multi-source heterogeneous data and identify patterns in the data, expressed as:

[0215] ;

[0216] in, is the original data matrix, and is an orthogonal matrix, is a diagonal matrix, representing the main components of the data;

[0217] S432, anomaly detection: Use a distance-based method to calculate the distance between each data point and its nearest neighbor to detect outliers in the data, expressed as:

[0218] ;

[0219] in, For data points The distance For data points In the The values ​​in the dimensions, For the The average value of the dimensions;

[0220] S433, real-time model adjustment: using the online gradient descent method, the parameters of the artificial neural network (ANN) model are adjusted in real time according to the detected abnormal values, which is expressed as:

[0221] ;

[0222] in, For the moment The model parameters, is the learning rate, is the loss function;

[0223] Through the above steps, it is ensured that the model can respond to data changes in a timely manner, improve the accuracy and reliability of the prediction, provide a more scientific and dynamic adjustment mechanism for the optimization prediction of flood propagation paths, and enhance the effectiveness of emergency response.

[0224] The emergency plan formulation and real-time monitoring in S5 include:

[0225] S51, emergency plan formulation: Based on the results of simulation calculations and flood propagation path optimization, combined with historical flood disaster response experience, formulate emergency plans, including personnel evacuation routes, material allocation plans, and rescue force deployment;

[0226] The Dijkstra algorithm is used to calculate the optimal evacuation route for personnel, which is expressed as:

[0227] ;

[0228] in, For Node The shortest path distance, For Node The neighbor nodes of For edge The weight of (distance or time);

[0229] The material allocation plan uses a linear programming model to optimize the material allocation plan, which is expressed as:

[0230] ;

[0231] The constraints are: ;

[0232] in, For supplies The cost, For supplies The amount of distribution, For supplies Meeting the needs The coefficient of For demand Minimum demand;

[0233] The deployment of rescue forces is optimized using an integer linear programming model, which is expressed as:

[0234] ;

[0235] The constraints are: ;

[0236] in, For rescue forces Deployed at location The benefits, A decision variable of 0 or 1, indicating whether to deploy. For rescue forces Total quantity limit;

[0237] S52, real-time monitoring: through sensors and satellite remote sensing technology, real-time monitoring of the latest environmental data at the flood disaster site;

[0238] S53, dynamic adjustment: according to the real-time environmental data and simulation results, the emergency plan is dynamically adjusted using an adaptive adjustment algorithm, which is expressed as:

[0239] ;

[0240] in, For the moment The plan, For the moment The plan, is the adjustment factor, To monitor the results in real time;

[0241] Through the above content, the intelligent and real-time optimization of emergency plans has been realized. This method improves the efficiency and accuracy of emergency response, effectively reduces the threat of flood disasters to life and property, and enhances the scientificity and reliability of emergency management.

[0242] Evaluation and improvement in S6 include:

[0243] S61, Data Collection: After the emergency response to flood disasters is completed, collect the effectiveness data of the emergency plan implementation process, including personnel evacuation time, material deployment efficiency, and rescue force response speed;

[0244] S62, Implementation effect evaluation: Use performance indicators (KPI) to evaluate the implementation effect of the emergency plan, expressed as:

[0245] ;

[0246] ;

[0247] ;

[0248] in, For the number of rescue missions, For the The response time of each task;

[0249] S63, Summarize lessons learned: Based on the results of the implementation effect evaluation, summarize the successful experiences and shortcomings in the emergency plan, and identify bottlenecks and problems in the emergency response;

[0250] The lessons learned include:

[0251] S631, Analyze successful experiences: Identify and record the good effects of measures in the emergency plan, including efficient evacuation routes and rapid material deployment;

[0252] S632, Identify deficiencies: Identify difficulties and problems encountered during the emergency response process, including congestion on some evacuation routes and untimely deployment of some supplies;

[0253] S633, Bottleneck problems: Identify bottleneck problems in emergency response, including poor information transmission and inflexible command and dispatch;

[0254] S634, Improvement Suggestions: Based on the analysis results, make improvement suggestions, including adjusting the content of the plan, optimizing the execution process, and strengthening personnel training;

[0255] S64, Optimize the digital twin model: Based on the lessons learned, optimize the digital twin model, expressed as:

[0256] ;

[0257] in, For the moment The model parameters, is the learning rate, Based on input data and target data The loss function of

[0258] S65, Optimize emergency plans: Based on the optimized digital twin model, re-formulate emergency plans, including adjusting evacuation routes, optimizing material allocation plans, updating rescue force deployment, improving information transmission processes, strengthening emergency drills and training, updating key steps and processes, increasing the flexibility of resource allocation, and strengthening coordination and cooperation with relevant departments to ensure that it is more scientific and efficient;

[0259] Through the above steps, not only the effectiveness of the emergency plan is improved, but also valuable guidance and reference are provided for future emergency management, thereby significantly improving the ability to respond to flood disasters and reducing the threat of disasters to life and property.

[0260] like Figure 2 As shown, a flood disaster emergency simulation system based on digital twins is used to implement the above-mentioned flood disaster emergency simulation method based on digital twins, including the following modules:

[0261] Data collection and preprocessing module: collects environmental data related to flood disasters in real time through a variety of sensors and satellite remote sensing technology, including rainfall, water level, flow rate and terrain data, and preprocesses the environmental data;

[0262] Digital twin model building module: Based on the collected environmental data, using geographic information platform, computer simulation and modeling technology, a digital twin model of the flood disaster area is established, including topography, hydrological characteristics and building distribution;

[0263] Simulation calculation module: Based on the digital twin model, the flood propagation path prediction model is used to simulate and calculate different flood disaster scenarios, and predict the flood propagation path, water depth and flow rate;

[0264] Flood propagation path optimization module: integrates meteorological forecast data, historical flood data and environmental data to optimize the prediction of flood propagation paths, identify patterns and anomalies, and adjust the flood propagation path prediction model in real time;

[0265] Emergency plan formulation and real-time monitoring module: Based on the results of simulation calculation and flood propagation path optimization, combined with historical flood disaster response experience, emergency plans are formulated, and the latest environmental data of the flood disaster site is monitored in real time. The emergency plan is dynamically adjusted based on real-time environmental data and simulation results;

[0266] Evaluation and Improvement Module: After the emergency response to flood disasters is completed, the implementation effect of the emergency plan will be evaluated, lessons learned will be summarized, and the digital twin model and emergency plan will be optimized.

[0267] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0268] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A flood disaster emergency simulation method based on digital twins, characterized in that: The following steps are involved: S1, data collection and preprocessing: collect environmental data related to flood disasters in real time through a variety of sensors and satellite remote sensing technology, including rainfall, water level, flow rate and terrain data, and preprocess the environmental data; S2, establish a digital twin model: Based on the collected environmental data, use the geographic information platform, computer simulation and modeling technology to establish a digital twin model of the flood disaster area, including topography, hydrological characteristics and building distribution. The establishment of the digital twin model specifically includes: S21, data input and registration: import the pre-processed environmental data into the geographic information platform and perform spatial registration and projection conversion; S22, spatial analysis: using the spatial analysis function of the geographic information platform to extract topographic features, analyze hydrological characteristics, and overlay building distribution data; S23, topographic modeling: constructing topography based on digital elevation model, using triangular irregular network method to represent the terrain surface, expressed as: ； in, Represents the terrain height, and Represents geographic coordinates; S24, hydrological characteristics modeling: The river water level and flow rate were simulated using hydrological simulation software, and the Saint-Venant equation was used for numerical calculation, which is expressed as: ； ； ； in, For water depth, and The water flow and The velocity component in the direction, is the acceleration due to gravity, is the terrain height, and is the bottom friction, is the source term; S25, building distribution modeling: Using terrain data and building distribution data, buildings are modeled as impermeable areas, which affect the flow path of water and the water accumulation area, expressed as: ； in, The coordinates of the vertices representing the building boundary; S3, simulation calculation: Based on the digital twin model, a flood propagation path prediction model is used to simulate different flood disaster scenarios and predict the flood propagation path, water depth and flow rate. The simulation calculation specifically includes: S31, initial condition setting: according to the environmental data in the digital twin model, the initial conditions for flood propagation are set, including initial water depth, flow velocity and terrain elevation; S32, boundary condition setting: according to the actual flood disaster scenario, the boundary conditions of the simulation calculation are set, including rainfall intensity, rainfall duration and water inflow flow; S33, finite volume method solution: The finite volume method is used for simulation calculation to predict the propagation path, water depth and flow rate of the flood, which can be expressed as: Conservation equations: ; in, is the conserved variable vector, is the flux vector, is the source term vector; Discretization: Divide the computational domain into several small finite volume units, integrate and discretize each unit, and express it as: ； in, Yes Unit The conserved variable, is the unit volume, is the unit boundary The flux at is the boundary area, is the source term of the unit; Flux calculation: The flux at the cell boundary is calculated using the Riemann solver , update the conserved variables of each unit , expressed as: ； in, is the flux at the cell boundary, and are the fluxes of the left and right states, respectively, and are the conserved variables on the left and right sides, is the maximum characteristic velocity; S4, Flood propagation path optimization: Integrate meteorological forecast data, historical flood data and environmental data to optimize the prediction of flood propagation paths and identify patterns and anomalies, including: S41, multi-source heterogeneous data fusion: fusion of meteorological forecast data, historical flood data and environmental data into multi-source heterogeneous data; S42, Optimal prediction of flood propagation paths: Using artificial neural network models to optimize the prediction of flood propagation paths based on the fused multi-source heterogeneous data; S43, Pattern and Anomaly Recognition: Identify patterns and anomalies in multi-source heterogeneous data and adjust artificial neural network models in real time; S5, emergency plan formulation and real-time monitoring: According to the results of simulation calculation and flood propagation path optimization, combined with historical flood disaster response experience, emergency plans are formulated, and the latest environmental data of the flood disaster site is monitored in real time. The emergency plan is dynamically adjusted according to the real-time environmental data and simulation results; S6, Evaluation and Improvement: After the emergency response to flood disasters is completed, the implementation effect of the emergency plan is evaluated and the digital twin model and emergency plan are optimized.

2. The flood disaster emergency simulation method based on digital twin according to claim 1 is characterized in that: The data collection and preprocessing in S1 include: S11, data collection: real-time collection of environmental data related to flood disasters, including: Rainfall collection: real-time measurement of rainfall through rain gauge sensors; Water level collection: Measure water level changes through ultrasonic water level sensors; Flow velocity collection: water flow velocity is measured by Doppler flowmeter; Terrain data collection: obtain terrain data through satellite remote sensing technology and process terrain height information using digital elevation models; S12, data preprocessing: preprocessing the environmental data collected in real time, specifically including: Data cleaning: Use median filtering algorithm to remove noise; Format conversion: Use linear interpolation method to convert environmental data in different formats into a standard format; Missing data imputation: Lagrange interpolation method is used to fill missing data.

3. The flood disaster emergency simulation method based on digital twin according to claim 1 is characterized in that: The multi-source heterogeneous data fusion in S41 includes: S411, data normalization: normalize weather forecast data, historical flood data, and environmental data; S412, data alignment and registration: align and register weather forecast data, historical flood data, and environmental data using timestamps and spatial coordinates; S413, data fusion: Use the weighted average method to fuse multi-source data, combine the weights of different data sources, and calculate the comprehensive value, expressed as: ； in, is the fused data, For data source The weight of For data source The data value of is the number of data sources; S414, verification and correction of fusion results: Verify and correct the fused multi-source heterogeneous data, and correct errors by comparing and analyzing them with the actual observed data.

4. The flood disaster emergency simulation method based on digital twin according to claim 3 is characterized in that: The artificial neural network model in S42 includes: S421, multi-source data feature fusion: feature fusion of weather forecast data, historical flood data and environmental data, design feature fusion layer, expressed as: ； in, , and are the feature vectors of weather forecast data, historical flood data, and environmental data, respectively; S422, time series feature processing: extracting time series features through a recursive neural network, expressed as: ； in, For the moment The hidden state of For the moment The fusion characteristics of For the moment The hidden state of and is the weight matrix, is the bias, is the ReLU activation function; S423, spatial feature extraction: Use convolutional neural network to extract spatial features, expressed as: ； in, is the output feature matrix Middle position The value of is the convolution kernel Middle position The value of is the input feature matrix The value of the corresponding position in ; S424, fully connected layer: The temporal features and spatial features are fused through the fully connected layer and the prediction results are output, which are expressed as: ； in, To output the prediction results, and are the weight matrices of temporal features and spatial features respectively, is the bias, is the extracted time series feature, is the extracted spatial features; S425, optimized loss function: A weighted loss function is used to weight the importance of different data sources, expressed as: ； in, and is the weight coefficient, is the predicted value, is the true value, is the number of samples.

5. The flood disaster emergency simulation method based on digital twin according to claim 4 is characterized in that: The pattern and anomaly identification in S43 includes: S431, Feature Extraction: Use principal component analysis to extract features from multi-source heterogeneous data and identify patterns in the data, expressed as: ； in, is the original data matrix, and is an orthogonal matrix, is a diagonal matrix; S432, anomaly detection: Use a distance-based method to calculate the distance between each data point and its nearest neighbor to detect outliers in the data, expressed as: ； in, For data points The distance For data points In the The values ​​in the dimensions, For the The average value of the dimensions; S433, real-time model adjustment: using the online gradient descent method, according to the detected abnormal values, the parameters of the artificial neural network model are adjusted in real time, which is expressed as: ； in, For the moment The model parameters, is the learning rate, is the loss function.

6. The flood disaster emergency simulation method based on digital twin according to claim 1 is characterized in that: The emergency plan formulation and real-time monitoring in S5 include: S51, emergency plan formulation: Based on the results of simulation calculations and flood propagation path optimization, combined with historical flood disaster response experience, formulate emergency plans, including personnel evacuation routes, material allocation plans, and rescue force deployment; S52, real-time monitoring: through sensors and satellite remote sensing technology, real-time monitoring of the latest environmental data at the flood disaster site; S53, dynamic adjustment: according to the real-time environmental data and simulation results, the emergency plan is dynamically adjusted using an adaptive adjustment algorithm, which is expressed as: ； in, For the moment The plan, For the moment The plan, is the adjustment factor, To monitor the results in real time.

7. The flood disaster emergency simulation method based on digital twin according to claim 6 is characterized in that: The evaluation and improvement in S6 include: S61, Data Collection: After the emergency response to flood disasters is completed, collect the effectiveness data of the emergency plan implementation process, including personnel evacuation time, material deployment efficiency, and rescue force response speed; S62, Implementation effect evaluation: Use performance indicators to evaluate the implementation effect of the emergency plan; S63, Problem Identification: Identify problems in emergency response based on the results of the execution effectiveness evaluation; S64, optimize the digital twin model: According to the identified emergency response problems, the digital twin model is optimized, which is expressed as: ； in, For the moment The model parameters, is the learning rate, Based on input data and target data The loss function of S65, Optimize emergency plans: Based on the optimized digital twin model, re-formulate emergency plans, including adjusting evacuation routes, optimizing material allocation plans, updating rescue force deployment, improving information transmission processes, strengthening emergency drills and training, updating key steps and processes, and increasing the flexibility of resource allocation.

8. A flood disaster emergency simulation system based on digital twins, used to implement a flood disaster emergency simulation method based on digital twins as described in any one of claims 1 to 7, characterized in that: Includes the following modules: Data collection and preprocessing module: collects environmental data related to flood disasters in real time through a variety of sensors and satellite remote sensing technology, including rainfall, water level, flow rate and terrain data, and preprocesses the environmental data; Digital twin model building module: Based on the collected environmental data, using geographic information platform, computer simulation and modeling technology, a digital twin model of the flood disaster area is established, including topography, hydrological characteristics and building distribution; Simulation calculation module: Based on the digital twin model, the flood propagation path prediction model is used to simulate and calculate different flood disaster scenarios, and predict the flood propagation path, water depth and flow rate; Flood propagation path optimization module: integrates meteorological forecast data, historical flood data and environmental data to optimize the prediction of flood propagation paths, identify patterns and anomalies, and adjust the flood propagation path prediction model in real time; Emergency plan formulation and real-time monitoring module: Based on the results of simulation calculation and flood propagation path optimization, combined with historical flood disaster response experience, emergency plans are formulated, and the latest environmental data of the flood disaster site is monitored in real time. The emergency plan is dynamically adjusted based on real-time environmental data and simulation results; Evaluation and improvement module: After the emergency response to flood disasters is completed, the implementation effect of the emergency plan will be evaluated, and the digital twin model and emergency plan will be optimized.

Images