A flood control safety evaluation system for levee projects

Abstract

The present invention discloses a flood control safety evaluation system for dike projects, which relates to the technical field of flood control safety. The initial position and boundary conditions of dike breaches are set through a simulation scenario construction module, and the MIKE 21 model is used to simulate the process of dike failure and breach formation; the data acquisition module records the breach boundary and water flow dynamic data in real time to provide support for numerical stability analysis; the numerical stability analysis module evaluates the stability of the model based on the dynamic data and identifies unstable areas; the instability adjustment module adopts an adaptive grid refinement technique to optimize the model calculation by dynamically adjusting the grid division; finally, the optimized model conducts a risk assessment of dike breaches through a risk analysis report generation module to generate an accurate risk analysis report. This technology improves the prediction accuracy and numerical stability in the process of dike breach simulation, and significantly enhances the reliability of risk assessment.

Description

Technical Field

[0001] The present invention relates to the technical field of flood control safety, and particularly relates to a flood control safety evaluation system for levee projects. Background Art

[0002] The flood control safety evaluation of levee projects refers to the systematic assessment of the flood control functions of levee projects to ensure that they can effectively prevent floods in the face of extreme weather conditions such as floods and heavy rains, and protect the safety of people's lives and property and the integrity of the ecological environment. The flood control safety evaluation usually involves comprehensive considerations of aspects such as the structural strength, stability, and anti-seepage performance of the levees, and evaluates their ability to resist external hydrological conditions such as specific water levels and flow velocities.

[0003] In the prior art, the flood control safety evaluation is mainly carried out through methods such as hydrological and hydraulic model analysis, numerical simulation, and geological exploration. For example, common techniques include simulating the bearing capacity of levees through finite element analysis (FEM), or using hydrodynamic simulation to evaluate the deformation and failure risks of levees under different flow conditions. These methods can perform accurate safety assessments by inputting the geological information, structural parameters, and historical flood data of the levees. For example, using the MIKE 21 hydrodynamic model to analyze the breach risk of levees to help decision-makers formulate flood control measures.

[0004] The prior art has the following deficiencies:

[0005] When a levee breach occurs, the free surface of the water flow shows a high degree of instability, and the expansion of the breach and the dynamic changes of the water flow are extremely complex. Although MIKE 21 provides a relatively powerful water flow simulation function, when facing dynamic free surface flow, especially when the levee breach begins to expand, it may face numerical stability problems. In addition, the rapid changes in the water flow and the irregular expansion of the breach boundary will bring a highly nonlinear dynamic process to the model, which may lead to numerical oscillations or non-convergence of the solution, thus affecting the accuracy of the simulation results and may cause serious deviations in risk assessment. Summary of the Invention

[0006] The purpose of the present invention is to provide a flood control safety evaluation system for levee projects to solve the deficiencies in the background art.

[0007] To achieve the above purpose, the present invention provides the following technical solution: A flood control safety evaluation system for levee projects, including a simulation scenario construction module, a data acquisition module, a numerical stability analysis module, an instability adjustment module, and a risk analysis report generation module;

[0008] Simulation scenario construction module: In the MIKE 21 model, set the initial position and boundary conditions of the breach occurrence, and use the model to simulate the process of levee failure and the preliminary formation of the breach;

[0009] Data acquisition module: During the simulation process, use the time stepping in MIKE 21 to record the changes of the breach boundary at each time step, and extract the expansion dynamic data of the boundary shape during the breach expansion process; set the free surface boundary of the water flow in the model and output the water level fluctuation data during the flood overflow process;

[0010] Numerical stability analysis module: According to the extracted expansion dynamic data of the boundary shape during the breach expansion process and the water level fluctuation data during the flood overflow process, evaluate the numerical stability during the MIKE 21 model simulation process;

[0011] Instability adjustment module: Based on the evaluation results, identify the areas with numerical instability, and adopt the dynamic grid refinement technology. Through the adaptive grid division method, dynamically adjust the calculation grid according to the expansion speed of the breach;

[0012] Risk analysis report generation module: After optimizing the MIKE 21 model, judge the prediction accuracy of the MIKE 21 model, and use the MIKE 21 model with high accuracy prediction for the risk assessment of levee breaches and generate the corresponding risk analysis report.

[0013] Preferably, in the data acquisition module, after analyzing the dynamic rate of breach expansion, generate a breach expansion rate anomaly index. The acquisition method of the breach expansion rate anomaly index is as follows:

[0014] Collect the time series data of the breach expansion rate , where represents the breach expansion rate at the nth time step, and measure its density in the local area by calculating the relative density of the data point and its neighbors. For a certain data point , its local reachability density The calculation expression is: ; where, reachability distance refers to the distance to the nearest point among the k nearest neighbors of the data point ; The distance of the nearest point to

[0015] Compare the density of the data point with its neighbors to judge whether it is an outlier , The calculation formula of the value is: ; where: is the set of k nearest neighbors of the data point , is the local reachability density of the point , calculate the breach expansion rate anomaly index, and the expression is: ; In the formula, is the abnormal index of the breach expansion rate, and respectively represent the minimum and maximum values among the local outlier values of all data points X.

[0016] Preferably, in the data acquisition module, after analyzing the water level rise and fall trends at different positions and time points, a water level rise and fall fluctuation index is generated. The method for obtaining the water level rise and fall fluctuation index is as follows:

[0017] Obtain the water level data W(t) of the time series, where t represents the time point and W(t) represents the water level at time t. Divide the water level time series into several time windows to obtain several subsequences , and each time point corresponds to a network node, and the weight of the node is set to the water level value ; Calculate the similarity of the water level between two time points. The expression is: ; where respectively represent the water level data at time points and , represents the similarity between time points and . According to the similarity between time points, construct an adjacency matrix A, where each element represents the connection strength between time points and . The calculation expression of the degree is: ; The clustering coefficient measures the connection density between the neighbors of node p and represents the local fluctuation trend of the water level change. The calculation formula of the clustering coefficient is: ; In the formula, is the degree of node p, represents the number of triangles formed between the neighbors of node p; After weighted average summation calculation of the calculated similarity and clustering coefficient, the water level rise and fall fluctuation index is obtained.

[0018] Preferably, in the numerical stability analysis module, the abnormal index of the breach expansion rate and the water level rise and fall fluctuation index are normalized so that they are both within [0,1]. After weighted average calculation of the normalized abnormal index of the breach expansion rate and the water level rise and fall fluctuation index, the numerical stability coefficient in the MIKE 21 model simulation process is obtained.

[0019] Preferably, in the instability adjustment module, the numerical stability coefficient obtained during the simulation process of the MIKE 21 model is compared with the pre-set numerical stability reference threshold based on experimental data. If the numerical stability coefficient during the simulation process of the MIKE 21 model is greater than or equal to the pre-set numerical stability reference threshold, it indicates that the numerical stability during the simulation process of the MIKE 21 model is high, and the area with high numerical stability is divided into the numerical stability area without the need to adjust the grid; if the numerical stability coefficient during the simulation process of the MIKE 21 model is less than the pre-set numerical stability reference threshold, it indicates that the numerical stability during the simulation process of the MIKE 21 model is low, and the area with low numerical stability is divided into the numerical instability area where grid refinement is required.

[0020] Preferably, in the risk analysis report generation module, after optimizing the MIKE 21 model, the prediction accuracy of the MIKE 21 model is judged. When judging the prediction accuracy of the MIKE 21 model, multiple error analysis methods are used, including the root mean square error RMSE and the coefficient of determination , and the calculation formula for the root mean square error RMSE is: ; where: is the value of the model prediction result at the b-th time point, is the value of the actual observed data at the b-th time point, b represents each specific time point in the calculation process, and E is the total number of time points; the coefficient of determination is used to measure the goodness of fit of the model, and its calculation formula is: ; where: is the mean value of the actual observed data.

[0021] Preferably, according to the numerical stability coefficient of the optimized MIKE 21 model and the improved grid refinement technology, the prediction accuracy of the model is judged through the calculated error index, specifically:

[0022] If RMSE < 0.1 and > 0.9, it is considered that the prediction accuracy is high, and the MIKE 21 model with high-accuracy prediction is used for the risk assessment of levee breaches and to generate the corresponding risk analysis report; if RMSE ≥ 0.1 and < 0.9, it is considered that the prediction accuracy is low, and the MIKE 21 model is re-optimized.

[0023] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0024] 1. The present invention effectively improves the prediction accuracy of the model in the face of dynamic water flow and complex breach expansion processes by introducing an optimized process with multiple modules, including simulation scenario construction, data acquisition, numerical stability analysis, instability adjustment, and risk analysis report generation. Especially in the data acquisition module, by calculating the breach expansion rate anomaly index and the water level rise and fall fluctuation index, and combining with the adaptive grid refinement technology, the stability of the model is optimized, avoiding numerical oscillation problems caused by the irregularity of breach expansion and the dynamic changes of water flow, and ensuring the accuracy of risk assessment.

[0025] 2. After normalization, the present invention performs a weighted average on the breach expansion rate anomaly index and the water level rise and fall fluctuation index to calculate the numerical stability coefficient, and performs grid refinement on the unstable area based on this coefficient, thereby improving the numerical stability of the model. By using error analysis methods (such as root mean square error RMSE and coefficient of determination), the prediction accuracy of the MIKE 21 model is evaluated, further improving the reliability of the simulation results. If the prediction accuracy reaches a high standard, an accurate risk analysis report can be generated to assist in the risk assessment of levee breaches and ensure the scientificity and rationality of flood control decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0027] Figure 1 It is a system module diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0029] Embodiment, please refer to Figure 1 As shown, a flood control safety evaluation system for a levee project in this embodiment includes a simulation scenario construction module, a data acquisition module, a numerical stability analysis module, an instability adjustment module, and a risk analysis report generation module;

[0030] Simulation scenario construction module: In the MIKE 21 model, set the initial position and boundary conditions of the breach occurrence, and use the model to simulate the process of dike failure and the initial formation of the breach;

[0031] Data acquisition module: During the simulation process, use the time stepping in MIKE 21 to record the changes of the breach boundary at each time step, and extract the dynamic data of the boundary shape expansion during the breach expansion process; Set the free surface boundary of the water flow in the model and output the water level fluctuation data during the flood overflow process;

[0032] Numerical stability analysis module: According to the extracted dynamic data of the boundary shape expansion during the breach expansion process and the water level fluctuation data during the flood overflow process, evaluate the numerical stability during the MIKE 21 model simulation process;

[0033] Instability adjustment module: Based on the evaluation results, identify the areas with numerical instability, and adopt the dynamic grid refinement technology. Through the adaptive grid division method, dynamically adjust the calculation grid according to the expansion speed of the breach;

[0034] Risk analysis report generation module: After optimizing the MIKE 21 model, judge the prediction accuracy of the MIKE 21 model, and use the MIKE 21 model with high-accuracy prediction for the risk assessment of dike breaches and generate the corresponding risk analysis report.

[0035] In the MIKE 21 model, set the initial position and boundary conditions of the dike breach occurrence, and simulate the process of dike failure and the initial formation of the breach. First, it is necessary to obtain the topographic data around the dike, including the shape of the dike, the surrounding river channels, spillways, etc. The topographic data can usually be obtained through digital elevation models (DEMs) or ground survey data. In MIKE 21, the topographic data is imported in the form of grids or irregular grids. The structural parameters of the dike (such as the elevation, width, material type, etc. of the dike) also need to be input. The soil parameters of the dike (such as permeability, shear strength, etc.) have an important impact on the formation of the breach.

[0036] In MIKE 21, the simulation area can be discretized using a regular grid (traditional rectangular grid) or an irregular grid (Flexible Mesh). Depending on the scale and complexity of the levee, grids with different resolutions can be selected to capture details. For the area where the levee breach occurs, the grid needs to be appropriately refined to accurately capture the process of breach expansion. In the levee model, boundary conditions such as inflow, outflow, and water level need to be set. Common boundary conditions include: Inflow boundary: Simulate the water flow input conditions of rivers or other water sources to the levee, such as water level or flow velocity. Outflow boundary: Set the outflow conditions of the simulation area, which can be fixed water level or flow rate conditions. Water level change: During the simulation, the water level change of flood events can be set, especially the water level rise process before the breach occurs.

[0037] Initial water level of the water body: Set the initial water level of the simulation area, usually the normal water level of the river or water body. When the breach occurs, the water level will change drastically, and the initial condition is the benchmark for simulating the water flow change. Initial state of the levee material and structure: The initial state of the levee structure and soil (such as dry or wet conditions, strength of soil layers, etc.) also needs to be set in the model. The material properties of the levee determine its durability under water flow scouring, affecting the formation speed and scale of the breach.

[0038] In MIKE 21, the levee failure can be simulated by setting the erosion characteristics and stress conditions of the levee material. Common failure mechanisms include: Scour erosion: The scouring action of the water flow will cause the loss of the soil on the surface of the levee, and then form a breach. In MIKE 21, the hydrodynamic module (such as MIKE 21 Flow Model) is usually used to simulate the scouring of the water flow on the levee surface. When the water level rises continuously, the permeability of the levee may increase, and the water flow will penetrate through the bottom or side of the levee, resulting in the weakening of the levee structure and further promoting the formation of the levee breach. Stress-strain model: The strength of the levee material can be described by the stress-strain model. In this model, the deformation and crack propagation on the levee surface under the action of the water flow are dynamic, and the finite element analysis (FEA) module may be needed to handle the elastoplastic behavior of the soil.

[0039] Once the levee begins to be scoured by the water flow, the erosive force of the water flow on the levee will increase, gradually weakening the stability of the levee. When simulating the initial formation of the levee breach, the hydrodynamic module can be used to simulate the velocity distribution of the water flow on the levee surface to determine which areas may be damaged first.

[0040] At the beginning of the levee failure, the breach boundary usually expands in an irregular manner. The initial formation of the breach may be local, but with the continuous scouring of the water flow, the breach will gradually increase. At this time, the breach expansion model or erosion model can be used to simulate the failure process of the levee and the expansion of the breach.

[0041] In the MIKE 21 model, the role of the data acquisition module is to extract and record important data such as breach expansion, boundary changes, and water flow fluctuations during the simulation process. Specifically, the time-stepping function of the model can be used to record and extract this data at each time step for subsequent analysis and evaluation. The following are the specific steps of the data acquisition module:

[0042] In MIKE 21, the time-stepping function is used to gradually simulate the development process of the levee breach. At each time step, the model calculates the evolution of the breach area based on water flow, soil properties, and external conditions, and records the changes in the breach boundary.

[0043] In the MIKE 21 model, the setting of the time step is very important, which determines the accuracy and stability of the simulation. The time step needs to be set according to the changes in water flow, the speed of the breach development process, etc. The time step can be set as a fixed value through the model or automatically adjusted according to the simulation process to better capture data during rapid changes such as breach expansion.

[0044] MIKE 21 will regularly record the data of the breach boundary during the simulation process, including: Boundary morphology: the geometric shape, position, expansion direction, etc. of the breach. Boundary size: the width, height, etc. of the breach. Breach expansion speed: the expansion rate and progress of the breach. This data will be saved at each time step for subsequent analysis of the breach expansion trend. The model provides an output function that can save these boundary change data in the form of time series data or spatial datasets. Users can extract the breach boundary data at specific time points according to their needs for further analysis.

[0045] The MIKE 21 model can generate dynamic data of the breach boundary changes, which is convenient for analyzing the specific evolution of the breach expansion process. The breach expansion data includes the changes in the damaged area of the levee at different time steps, helping to judge the evolution trend and potential risks of the breach. At each time step, the model calculates and outputs the morphological and position data of the breach area. Users can extract this data and display the breach expansion process in the form of graphs or data tables. These data can help analyze: The path of breach expansion: how the breach expands along the levee boundary and whether there is a specific expansion direction. The rate of breach expansion: the rate of breach expansion at different time steps, which may be related to factors such as water flow velocity and levee structure. The model can also output the boundary change rate, that is, the dynamic rate of breach expansion, to help evaluate the speed of breach progress.

[0046] MIKE 21 provides visualization tools that can display the breach boundary data in the form of dynamic graphs. Through these graphs, users can intuitively see the expansion and changes of the breach boundary. Different time-step data during the breach development process can be observed through methods such as layer overlay or time series graphs.

[0047] After analyzing the dynamic rate of breach expansion, a breach expansion rate anomaly index is generated. The method for obtaining the breach expansion rate anomaly index is as follows:

[0048] Collect the time series data of the breach expansion rate, which represents the expansion rate at different time steps (or spatial positions). Assume these data are represented as , where represents the breach expansion rate at the nth time step. The density of this point in the local area is measured by calculating the relative density of the data point and its neighbors. For a certain data point , its local reachability density The calculation expression is: ; where reachabilitydistance refers to the distance to the point closest to among the k nearest neighbors of the data point .

[0049] Compare the density of the data point with that of its neighbors to determine whether this point is an outlier. If the density of a point is significantly lower than that of its neighbors, then this point is considered an outlier , The value of is relatively large. The calculation formula for the value of is: is the set of k nearest neighbors of the data point , is the local reachability density of the point . If ≈1, it means that the density of is similar to that of its neighbors and belongs to a normal point. If >1, it means that the density of is significantly lower than that of its neighbors and belongs to an abnormal point. If <1, it means that the density of is significantly higher than that of its neighbors and may belong to the center of the cluster. Calculate the breach expansion rate anomaly index, and the expression is: ; In the formula, is the breach expansion rate anomaly index, which represents the degree of anomaly of the breach expansion rate at a specific time step (or position). and respectively represent the minimum and maximum values of the local outlier values of all data points X.

[0050] The closer the value is to 1, the more abnormal the point is, that is, there is a significant difference between the change in the breach expansion rate and the surrounding data points. When the value is close to 0, it indicates that the point is relatively normal, and the change in the breach expansion rate conforms to the normal pattern of the surrounding area.

[0051] In MIKE 21, setting the free surface boundary of the water flow is an important step in simulating the water flow fluctuations during the breach process. The free surface of the water flow refers to the boundary where the water level changes over time. Especially during the flood overflow process, the water level fluctuations are crucial for the dike failure and the formation of the breach.

[0052] In MIKE 21, the free surface boundary of the water flow can be defined through boundary condition settings. Usually, it includes: water level boundary conditions: setting the water level conditions at the inflow and outflow ends of the simulation area, especially for the water level fluctuations at the breach during the occurrence and development of the breach. Water flow rate boundary conditions: In the breach area, the water flow velocity and direction also have an impact on the dike failure. By setting the flow velocity boundary, the model can simulate the scouring effect of the water flow on the dike. These boundary conditions will affect the formation of the dike breach, and according to the different water flows, the water level fluctuations at the boundary will change.

[0053] During the simulation process, MIKE 21 will continuously monitor and output the water level fluctuation data. These data include: water level change curve: the water level data of each measuring point within each time step, recording the water level fluctuations during the flood overflow process. Water level fluctuation trend: recording the rising and falling trends of the water level at different positions and time points, helping to analyze the impact of the flood on the dike. In addition to the water level fluctuation data, MIKE 21 can also output dynamic change data such as flow velocity and water depth. These data are crucial for evaluating the damage degree of the dike during the flood overflow process and the development of the breach. These data can be collected by setting the sensor positions and time intervals, so as to understand the water flow characteristics in detail during the flood process.

[0054] After the MIKE 21 model outputs the data, users can further process and analyze it through the built-in data analysis tools or external data analysis software (such as MATLAB, Excel, etc.): Time series analysis: Analyze the time series data of the breach boundary change, expansion rate, and water level fluctuation. Spatial data analysis: Through the spatial analysis tool, observe the spatial changes of the breach boundary and evaluate the breach risks in different regions.

[0055] After analyzing the rising and falling trends of the water level at different positions and time points, a water level rising and falling fluctuation index is generated. The method for obtaining the water level rising and falling fluctuation index is as follows:

[0056] Obtain the water level data W(t) of the time series, where t represents the time point and W(t) represents the water level at time t. Divide the water level time series into several time windows to obtain several subsequences. , each time point corresponds to a network node, and the weight of the node is set to the water level value. ; Calculate the similarity of the water level between two time points. The expression is: ; where respectively represent the water level data at time points and , represents the similarity between time points and . Construct an adjacency matrix A according to the similarity between time points. Each element represents the connection strength between time points and . The calculation expression of the degree is: ; The clustering coefficient measures the connection density between the neighbors of node p and represents the local fluctuation trend of the water level change. The calculation formula of the clustering coefficient is: ; In the formula, is the degree of node p, represents the number of triangles formed between the neighbors of node p; After calculating the weighted average sum of the obtained similarity and clustering coefficient, the water level rise and fall fluctuation index is obtained.

[0057] In the numerical stability analysis module, normalize the breach expansion rate anomaly index and the water level rise and fall fluctuation index so that they are both in the range of [0, 1]. After weighted average calculation of the normalized breach expansion rate anomaly index and the water level rise and fall fluctuation index, the numerical stability coefficient in the MIKE 21 model simulation process is obtained.

[0058] For example, the present invention can calculate the numerical stability coefficient in the MIKE 21 model simulation process using the following formula. The calculation expression is: ; In the formula, is the numerical stability coefficient in the MIKE 21 model simulation process, is the breach expansion rate anomaly index, is the water level rise and fall fluctuation index, is the weight coefficient of the breach expansion rate anomaly index and the water level rise and fall fluctuation index (which can be optimized according to experimental experience or machine learning), and are both greater than 0.

[0059] Instability Adjustment Module: Based on the evaluation results, identify the regions with numerical instability, and adopt dynamic grid refinement technology. Through the adaptive grid division method, dynamically adjust the computational grid according to the expansion speed of the breach.

[0060] Compare the numerical stability coefficient obtained during the simulation of the MIKE 21 model with the pre-set numerical stability reference threshold based on experimental data. If the numerical stability coefficient during the simulation of the MIKE 21 model is greater than or equal to the pre-set numerical stability reference threshold, it indicates that the numerical stability during the simulation of the MIKE 21 model is high, and divide the region with high numerical stability into the numerical stability region without grid adjustment; if the numerical stability coefficient during the simulation of the MIKE 21 model is less than the pre-set numerical stability reference threshold, it indicates that the numerical stability during the simulation of the MIKE 21 model is low, and divide the region with low numerical stability into the numerical instability region, where grid refinement is required.

[0061] For the numerical instability region, adopt dynamic grid refinement technology to improve the numerical accuracy and stability of the simulation. The steps are as follows:

[0062] By observing the initial formation and expansion process of the breach, determine the expansion speed of the breach. The expansion speed is calculated based on the time step and the spatial change in the breach boundary: ; where is the change in the breach boundary within the time step Δt.

[0063] Based on the breach expansion speed , dynamically adjust the grid division:

[0064] Compare the calculated breach expansion speed with the pre-set breach expansion speed threshold. If the breach expansion speed is greater than or equal to the pre-set breach expansion speed threshold, divide it into the high expansion speed region; if the breach expansion speed is less than the pre-set breach expansion speed threshold, divide it into the low expansion speed region.

[0065] High expansion speed region: In the region with a high breach expansion speed, the grid should be more refined to ensure the accuracy of the simulation results. This region requires smaller computational units. Low expansion speed region: In the region with a low breach expansion speed, larger computational units can be used to reduce the computational amount. Through the adaptive grid technology, the grid will automatically adjust according to the breach expansion speed, enabling higher computational accuracy in the region with higher numerical instability.

[0066] Define a grid resolution function for each time step , this function adjusts the grid refinement degree according to the breach expansion speed, and the expression is: ; among which, is an adaptability function. When is relatively high, the grid will be refined; when is relatively low, the grid will be coarsened.

[0067] At each time step of the model, by adjusting the distribution of grid nodes, according to the current breach expansion speed and numerical stability coefficient, grid refinement is carried out on the numerically unstable area. For the high-expansion speed area, refine the grid cells to increase the calculation accuracy; for the stable area, appropriately coarsen the grid cells to reduce the consumption of calculation resources.

[0068] Risk analysis report generation module: After optimizing the MIKE 21 model, judge the prediction accuracy of the MIKE 21 model, and use the MIKE 21 model with high prediction accuracy for the risk assessment of levee breaches and generate the corresponding risk analysis report.

[0069] When judging the prediction accuracy of the MIKE 21 model, multiple error analysis methods can be used. Commonly used ones include root mean square error (RMSE) and coefficient of determination . Among them, RMSE is widely used in the evaluation of simulation prediction accuracy because it can quantify the difference between the simulation results and the actual observed data.

[0070] The calculation formula of root mean square error RMSE is: ; among which: is the value of the model prediction result at the b-th time point, is the value of the actual observed data at the b-th time point, b represents each specific time point in the calculation process, and E is the total number of time points; the coefficient of determination is used to measure the goodness of fit of the model, and its calculation formula is: ; among which: is the mean value of the actual observed data. The closer the value is to 1, the higher the prediction accuracy of the model.

[0071] According to the numerical stability coefficient of the optimized MIKE 21 model and the improved grid refinement technology, by calculating the RMSE and indexes, to quantify the prediction accuracy of the model. Usually, a lower RMSE and a higher value indicate that the optimized model has a higher prediction accuracy.

[0072] According to the calculated error indexes, judge the prediction accuracy of the model, specifically:

[0073] If RMSE < 0.1, and If RMSE > 0.9, it is considered that the prediction accuracy is high. If RMSE ≥ 0.1, and < 0.9, it is considered that the prediction accuracy is low, and the MIKE 21 model needs to be re - optimized.

[0074] After determining that the optimized MIKE 21 model has high prediction accuracy, the model can be used for the risk assessment of levee breaches. The following are the specific steps for the risk assessment of levee breaches:

[0075] Use the optimized MIKE 21 model to simulate different scenarios of levee breaches. For example: simulate the process of levee breaches under different water levels and flow rates. Evaluate the bearing capacity of the levee structure through data such as the breach expansion rate and the fluctuation of flood overflow water levels.

[0076] According to the simulation results, combine risk analysis methods to classify the risks of levee breaches: High - risk area: The breach expands rapidly and the flood fluctuates greatly, which may lead to the destruction of the levee. Medium - risk area: The breach expands relatively slowly, the water level fluctuates moderately, but still needs to be vigilant. Low - risk area: The breach expands relatively slowly, the water level fluctuates little, and the levee stability is relatively high.

[0077] Based on the above assessment results, generate a risk analysis report. The content of the report includes: The risk level of the levee breach. The numerical stability coefficient, breach expansion rate, and flood fluctuation conditions of each area. Suggestions for unstable areas (such as strengthening levee reinforcement, adjusting the breach boundary, etc.). The content of the report should include the following aspects:

[0078] Model overview: Introduce the optimized MIKE 21 model, the simulation technology used, data sources, and analysis methods. Prediction accuracy assessment: Based on RMSE and , provide a detailed analysis of the model prediction accuracy. Risk assessment results: Based on the optimized simulation results, analyze and classify the risks of levee breaches. Suggestions and measures: According to the risk assessment results, propose targeted disaster prevention and mitigation measures, such as levee reinforcement, early warning system optimization, etc.

[0079] In this embodiment, first, in the simulated scenario construction module, the initial position and boundary conditions of the breach are set, and the MIKE 21 model is used to simulate the process of the dike failure and the initial formation of the breach. Then, the data acquisition module records the changes in the breach boundary through time stepping and extracts the extended dynamic data, while monitoring the water level fluctuations. Based on these data, the numerical stability analysis module evaluates the numerical stability of the model and determines whether there are unstable regions. If instability is found, the instability adjustment module uses dynamic grid refinement technology to adaptively adjust the computational grid, thereby optimizing the model. Finally, the risk analysis report generation module evaluates the prediction accuracy of the model, conducts a risk assessment of the dike breach based on the optimized high-accuracy model, and finally generates a detailed risk analysis report to provide a basis for decision-making.

[0080] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. A flood control safety evaluation system for dike projects, characterized in that: It includes a simulation scenario construction module, a data acquisition module, a numerical stability analysis module, an instability adjustment module, and a risk analysis report generation module; Simulation scenario construction module: In the MIKE 21 model, set the initial position and boundary conditions of the breach occurrence, and use the model to simulate the process of dike failure and the initial formation of the breach; Data acquisition module: During the simulation process, use the time stepping in MIKE 21 to record the changes of the breach boundary at each time step, and extract the expansion dynamic data of the boundary shape during the breach expansion process; Set the free surface boundary of the water flow in the model and output the water level fluctuation data during the flood overflow process; After analyzing the dynamic rate of breach expansion, generate a breach expansion rate anomaly index. The method for obtaining the breach expansion rate anomaly index is: Collect time - series data on the breach expansion rate , where represents the breach expansion rate at the n - th time step, and its density within the local area is measured by calculating the relative density of the data point and its neighbors. For a certain data point , its local reachability density is calculated by the following expression: ; where the reachability distance refers to the distance to the point closest to among the k nearest neighbors of the data point . Compare data points with the density of its neighbors to determine whether it is an outlier , The calculation formula for the value is: ; where: is the set of k nearest neighbors of the data point , is the local reachability density of the point Calculate the abnormal index of the breach expansion rate, and the expression is: ; In the formula, is the abnormal index of the breach expansion rate, and respectively represent the minimum and maximum values of the local outlier values of all data points X; After analyzing the rising and falling trends of the water level at different positions and time points, generate a water level rising and falling fluctuation index. The method for obtaining the water level rising and falling fluctuation index is: Obtain the water level data of the time series , where t represents the time point, represents the water level at time t. Divide the water level time series into several time windows to obtain several subsequences , and each time point corresponds to a network node, and the weight of the node is set to the water level value ; Calculate the similarity of the water level between two time points. The expression is: ; where, respectively represent the water level data at time points and ; represents the similarity between time points and . According to the similarity between time points, construct an adjacency matrix A, where each element represents the connection strength between time points and . The calculation expression of the degree is: ; The clustering coefficient measures the connection density between the neighbors of node p and represents the local fluctuation trend of the water level change. The calculation formula of the clustering coefficient is: ; In the formula, is the degree of node p, represents the number of triangles formed between the neighbors of node p; After calculating the weighted average sum of the obtained similarity and clustering coefficient, the water level rise and fall fluctuation index is obtained; Numerical stability analysis module: Based on the extracted expansion dynamic data of the boundary shape during the breach expansion process and the water level fluctuation data during the flood overflow process, evaluate the numerical stability during the MIKE 21 model simulation; Instability adjustment module: Based on the evaluation results, identify the areas with numerical instability, and adopt the dynamic grid refinement technology. Through the adaptive grid division method, dynamically adjust the calculation grid according to the expansion speed of the breach; Risk analysis report generation module: After optimizing the MIKE 21 model, judge the prediction accuracy of the MIKE 21 model, and use the MIKE 21 model with high prediction accuracy for the risk assessment of dike breaches and generate the corresponding risk analysis report.

2. The flood control safety evaluation system for a dike project according to claim 1, wherein: In the numerical stability analysis module, normalize the breach expansion rate anomaly index and the water level rising and falling fluctuation index so that they are both within [0,1]. After weighted average calculation of the normalized breach expansion rate anomaly index and the water level rising and falling fluctuation index, obtain the numerical stability coefficient during the MIKE 21 model simulation.

3. The flood control safety evaluation system for a dike project according to claim 2, characterized in that: In the instability adjustment module, compare the obtained numerical stability coefficient during the MIKE 21 model simulation with the pre-set numerical stability reference threshold according to the experimental data. If the numerical stability coefficient during the MIKE 21 model simulation is greater than or equal to the pre-set numerical stability reference threshold, it indicates that the numerical stability during the MIKE 21 model simulation is high, and divide the area with high numerical stability into the numerical stability area without grid adjustment; If the numerical stability coefficient during the MIKE 21 model simulation is less than the pre-set numerical stability reference threshold, it indicates that the numerical stability during the MIKE 21 model simulation is low, and divide the area with low numerical stability into the numerical instability area, where grid refinement is required.

4. The flood control safety evaluation system for a dike project according to claim 3, characterized in that: In the risk analysis report generation module, after optimizing the MIKE 21 model, the prediction accuracy of the MIKE 21 model is judged. When judging the prediction accuracy of the MIKE 21 model, multiple error analysis methods are used, including root mean square error (RMSE) and coefficient of determination , and the calculation formula for the root mean square error (RMSE) is: ; where: is the value of the model prediction result at the b-th time point, is the value of the actual observed data at the b-th time point, b represents each specific time point in the calculation process, and E is the total number of time points; the coefficient of determination is used to measure the goodness of fit of the model, and its calculation formula is: ; where: is the mean value of the actual observed data.

5. The flood control safety evaluation system for a dike project according to claim 4, characterized in that: Based on the numerical stability coefficient of the optimized MIKE 21 model and the improved grid refinement technology, judge the prediction accuracy of the model through the calculated error index, specifically: If and , the prediction accuracy is considered high, and the MIKE 21 model with high-accuracy prediction is used for the risk assessment of levee breaches and to generate the corresponding risk analysis report; if and , the prediction accuracy is considered low, and the MIKE 21 model is re-optimized.

Images