Typhoon disaster scene generation method capable of improving accuracy and efficiency

Abstract

The invention discloses a typhoon disaster scene generation method capable of improving accuracy and high efficiency, and aims to solve the problem of low simulation precision of an existing typhoon disaster simulator. The typhoon disaster scene generation method comprises the following steps: 1, selecting all typhoon samples passing through a target area from a database, and extracting typhoon information from the typhoon samples; 2, calculating a gradient wind speed, superposing a moving wind field, and converting a superposed total wind speed field into a ground wind speed; 3, discretizing the target area into a plurality of grid units, and screening out the maximum ground wind speed when a typhoon sample passes each time, so as to form a maximum wind speed intensity measurement map of the target area; 4, reducing the number of the maximum wind speed intensity measurement diagrams of all typhoon samples by using a disaster quantification method; and 5, outputting NHQ optimized quanta and weights of the optimized quanta. According to the method, the damage condition of the typhoon disaster to the target infrastructure system can be evaluated more accurately, and meanwhile, the analysis efficiency is improved.

Description

Technical Field

[0001] The present invention belongs to the field of typhoon disaster simulation methods for disaster prevention and mitigation, and specifically relates to a typhoon disaster scene generation method that takes both accuracy and efficiency into consideration, and is suitable for improving the accuracy of typhoon simulation. Background Art

[0002] Typhoons are a frequent natural disaster that can cause serious structural damage to infrastructure systems such as water networks and power grids, resulting in huge economic losses. Based on this, modeling typhoon disasters, analyzing the damage state and functional loss of infrastructure systems under disasters, and conducting resilience assessments and considering resilience improvement have become research hotspots, which are of great significance for effectively preventing typhoon disasters.

[0003] There are few existing studies on damage analysis of infrastructure systems under typhoon disasters, and many of them are based on single historical typhoon events. This method cannot effectively reproduce the typhoon disaster situation in the area where the infrastructure system is located. The accuracy of disaster analysis requires a large number of typhoon simulation scenarios, but this method also reduces efficiency. How to balance the accuracy and efficiency of typhoon disaster scene generation is a problem worth studying. Summary of the invention

[0004] The purpose of the present invention is to solve the problem of low simulation accuracy of existing typhoon disaster simulators and to provide a method for generating typhoon disaster scenes to improve the efficiency of disaster simulation and achieve a balance between accuracy and efficiency.

[0005] The typhoon disaster scene generation method of the present invention with improved accuracy and efficiency is implemented by the following steps:

[0006] Step 1: Select the target area (i.e. the area where the infrastructure system is located), select all typhoon samples passing through the target area from the simulated typhoon database, and extract typhoon information from the selected typhoon samples. The typhoon information includes longitude, latitude, typhoon center pressure difference, maximum wind speed and translation speed;

[0007] Step 2. Step 2.1, calculate the parameters A and B in the gradient wind field model;

[0008] The calculation formula of parameter A is as follows:

[0009]

[0010] The calculation formula of parameter B is as follows:

[0011]

[0012] Among them, C and C 0They represent correction parameters, f is the Coriolis parameter, r is the distance between the wind speed calculation point and the typhoon center, ΔP is the pressure difference at the typhoon center, e is a natural constant, V m is the maximum wind speed, ρ a is the air density;

[0013] Step 2.2: Calculate the maximum wind speed radius R max , calculate the maximum wind speed radius R through the XGBoost model max , maximum wind speed radius R max The model is as follows:

[0014]

[0015] in, is the latitude of the typhoon, ΔP is the pressure difference at the center of the typhoon, ε Rmax is the error term;

[0016] Step 2.3: Set parameters A, B and maximum wind speed radius R max Substitute it into the gradient wind field model and calculate the gradient wind speed

[0017]

[0018] Step 2.4: Superimpose the transitional wind field and Based on the typhoon's migration wind field V t , and the total wind speed field V after superposition is obtained S :

[0019] V S =V G +V t

[0020] Step 2.5: The superimposed total wind speed field V S Converted to ground wind speed V H (t), the conversion formula is as follows:

[0021]

[0022] Where H represents the target height, that is, the height from the ground, z 0 Indicates the surface roughness length;

[0023] Step 3: Discretize the target area into multiple grid cells, record the longitude and latitude coordinates of the center point of each grid cell, and calculate the total wind speed field V after superposition based on the improved wind field model in step 2. S, thus calculating the ground wind speed values ​​of the typhoon center at different times at the center of each grid unit when (a certain) typhoon sample passes by, and for each grid unit center point, screening out the maximum ground wind speed when each typhoon sample passes by, thereby forming a maximum wind speed intensity measurement map of the target area, and each typhoon sample obtains a maximum wind speed intensity measurement map;

[0024] Step 4: Use the disaster quantification method to reduce the number of maximum wind speed intensity measurement maps for all typhoon samples. The process of reducing the number of maximum wind speed intensity measurement maps is as follows:

[0025] Step 4.1: Define the sample as the maximum wind speed intensity measurement diagram obtained from all typhoon samples. The maximum wind speed intensity measurement diagram of each sample is represented by V i (lon,lat) means that the total number of samples is N events , define quantum as the reduced maximum wind speed intensity measurement map, and set the number of quanta to N HQ ; First, randomly select N from the sample HQ samples as quanta;

[0026] Step 4.2: Calculate the nth quantum With all samples V i Wind speed field distance (lon,lat) Represents the square of the maximum wind speed difference between the i-th sample and the n-th quantum at the center point (lon, lat) of the corresponding grid unit, the wind speed field distance The calculation formula is as follows:

[0027]

[0028] In the formula, lon represents the longitude of the location, lat represents the latitude of the location, and lat min and lat max Respectively represent the minimum and maximum latitude of the wind speed field area, lon min and lon max They represent the minimum and maximum latitudes of the wind speed field area, respectively;

[0029] Step 4.3: Take the maximum wind speed intensity measurement graph V in the sample i (lon,lat) is assigned to the jth quantum Q that is closest to it j ,Right now where d ij represents the wind speed field distance between the i-th sample and quantum j, d ik Represents the wind speed field distance between the i-th sample and quantum k, thereby obtaining the maximum wind speed intensity measurement map set assigned to quantum j The number of samples in the maximum wind speed intensity measurement map set assigned to quantum j is N j ;

[0030] Step 4.4: Take the average of the samples assigned to each quantum to obtain the updated quantum

[0031]

[0032] Step 4.5: Repeat steps 4.2 to 4.4 to iterate the quantum until convergence, and obtain N HQ An optimized post-quantum;

[0033] Step 4.6: Calculate the optimized quantum weight P j , P j The calculation formula is:

[0034]

[0035] Among them, N j is the number of samples assigned to the jth optimized quantum, N events is the total number of samples;

[0036] Step 5: Output N from step 4 HQ The optimized post-quantum and the weight of the optimized post-quantum, that is, N HQ The maximum wind speed intensity measurement diagram and the weight P of the optimized quantum j , thus generating a typhoon disaster scenario.

[0037] The present invention uses a disaster quantification method to reduce the number of maximum wind speed intensity measurement maps, improve the efficiency of typhoon disaster simulation, and input the reduced maximum wind speed intensity measurement maps into the infrastructure system to be studied to conduct damage analysis, recovery analysis, resilience assessment, etc.

[0038] The present invention realizes the generation of typhoon disaster scenarios that takes both accuracy and efficiency into consideration. It can more accurately assess the damage caused by typhoon disasters to target infrastructure systems while improving analysis efficiency. It is of great significance for damage analysis, function analysis, resilience assessment and improvement of infrastructure under a large number of typhoon simulation scenarios.

[0039] The present invention integrates two methods, namely, a machine learning method and a disaster quantification method, thereby effectively improving the accuracy of typhoon disaster simulation and improving the efficiency of typhoon disaster simulation.

[0040] Compared with the prior art, the typhoon disaster scene generation method of the present invention has the following advantages with improved accuracy and efficiency:

[0041] 1. The wind field model has been improved, and the accuracy of typhoon disaster simulation has been improved.

[0042] 2. Applying disaster quantification methods to the field of typhoon disaster modeling improves the efficiency of typhoon disaster scenario generation without affecting accuracy.

[0043] 3. A typhoon disaster modeling framework that combines both accuracy and efficiency is proposed. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a schematic diagram of discretization of the target area in the embodiment;

[0045] Figure 2 Schematic diagram of wind speed in the target area in the embodiment;

[0046] Figure 3 Schematic diagram of wind direction in the target area in the embodiment;

[0047] Figure 4 It is a comparison diagram of the probability of exceeding the extreme wind speed before and after the reduction of HQ disaster in the embodiment;

[0048] Figure 5 : A comparison diagram of the probability of exceeding the number of transmission tower damages before and after the reduction of the HQ disaster in the embodiment;

[0049] Figure 6 The following is a comparison chart of the maximum wind speed simulated and the actual record in different implementation modes in the example. DETAILED DESCRIPTION

[0050] Specific implementation method 1: The typhoon disaster scene generation method of this implementation method with improved accuracy and efficiency is implemented according to the following steps:

[0051] Step 1: Select the target area (i.e. the area where the infrastructure system is located), select all typhoon samples passing through the target area from the simulated typhoon database, and extract typhoon information from the selected typhoon samples. The typhoon information includes longitude, latitude, typhoon center pressure difference, maximum wind speed and translation speed;

[0052] Step 2. Step 2.1, calculate the parameters A and B in the gradient wind field model;

[0053] The calculation formula of parameter A is as follows:

[0054]

[0055] The calculation formula of parameter B is as follows:

[0056]

[0057] Among them, C and C 0They represent correction parameters, f is the Coriolis parameter, r is the distance between the wind speed calculation point and the typhoon center, ΔP is the pressure difference at the typhoon center, e is a natural constant, V m is the maximum wind speed, ρ a is the air density;

[0058] Step 2.2: Calculate the maximum wind speed radius R max , calculate the maximum wind speed radius R through the XGBoost model max , maximum wind speed radius R max The model is as follows:

[0059]

[0060] in, is the latitude of the typhoon, ΔP is the pressure difference at the center of the typhoon, ε Rmax is the error term;

[0061] Step 2.3: Set parameters A, B and maximum wind speed radius R max Substitute it into the gradient wind field model and calculate the gradient wind speed

[0062]

[0063] Step 2.4: Superimpose the transitional wind field and Based on the typhoon's migration wind field V t , and the total wind speed field V after superposition is obtained S :

[0064] V S =V G +V t

[0065] Step 2.5: The superimposed total wind speed field V S Converted to ground wind speed V H (t), the conversion formula is as follows:

[0066]

[0067] Where H represents the target height, that is, the height from the ground, z 0 Indicates the surface roughness length;

[0068] Step 3: Discretize the target area into multiple grid cells, record the longitude and latitude coordinates of the center point of each grid cell, and calculate the total wind speed field V after superposition based on the improved wind field model in step 2. S, thus calculating the ground wind speed values ​​of the typhoon center at different times at the center of each grid unit when (a certain) typhoon sample passes by, and for each grid unit center point, screening out the maximum ground wind speed when each typhoon sample passes by, thereby forming a maximum wind speed intensity measurement map of the target area, and each typhoon sample obtains a maximum wind speed intensity measurement map;

[0069] Step 4: Use the disaster quantification method to reduce the number of maximum wind speed intensity measurement maps for all typhoon samples. The process of reducing the number of maximum wind speed intensity measurement maps is as follows:

[0070] Step 4.1: Define the sample as the maximum wind speed intensity measurement diagram obtained from all typhoon samples. The maximum wind speed intensity measurement diagram of each sample is represented by V i (lon,lat) means that the total number of samples is N events , define quantum as the reduced maximum wind speed intensity measurement map, and set the number of quanta to N HQ ; First, randomly select N from the sample HQ samples as quanta;

[0071] Step 4.2: Calculate the nth quantum With all samples V i Wind speed field distance (lon,lat) Represents the square of the maximum wind speed difference between the i-th sample and the n-th quantum at the center point (lon, lat) of the corresponding grid unit, the wind speed field distance The calculation formula is as follows:

[0072]

[0073] In the formula, lon represents the longitude of the location, lat represents the latitude of the location, and lat min and lat max Respectively represent the minimum and maximum latitude of the wind speed field area, lon min and lon max They represent the minimum and maximum latitudes of the wind speed field area, respectively;

[0074] Step 4.3: Take the maximum wind speed intensity measurement graph V in the sample i (lon,lat) is assigned to the jth quantum Q that is closest to it j ,Right now where d ij represents the wind speed field distance between the i-th sample and quantum j, d ik Represents the wind speed field distance between the i-th sample and quantum k, thereby obtaining the maximum wind speed intensity measurement map set assigned to quantum j The number of samples in the maximum wind speed intensity measurement map set assigned to quantum j is N j ;

[0075] Step 4.4: Take the average of the samples assigned to each quantum to obtain the updated quantum

[0076]

[0077] Step 4.5: Repeat steps 4.2 to 4.4 to iterate the quantum until convergence, and obtain N HQ An optimized post-quantum;

[0078] Step 4.6: Calculate the optimized quantum weight P j , P j The calculation formula is:

[0079]

[0080] Among them, N j is the number of samples assigned to the jth optimized quantum, N events is the total number of samples;

[0081] Step 5: Output N from step 4 HQ The optimized post-quantum and the weight of the optimized post-quantum, that is, N HQ The maximum wind speed intensity measurement diagram and the weight P of the optimized quantum j , thus generating a typhoon disaster scenario.

[0082] In step 4 of this embodiment, the maximum wind speed intensity measurement diagram of each sample is converted into V i (lon, lat) indicates that it is a sequence of the maximum wind speeds calculated at each grid in the target area, which are arranged in sequence according to the division number of each position (lon, lat) in the target area.

[0083] Specific implementation method 2: The difference between this implementation method and specific implementation method 1 is that the simulated typhoon database in step 1 is a CMA database, and the number of typhoon samples is greater than 2000.

[0084] Specific implementation method three: This implementation method is different from specific implementation methods one or two in that the calculation formula of the correction parameter C in the formula of step 2.1 is as follows:

[0085]

[0086] This embodiment C 0 The value of is 1.0628, ρ a is the air density, taken as 1.15*10 -3 kg / m 3.

[0087] Specific implementation method 4: This implementation method is different from the specific implementation methods 1 to 3 in that the maximum wind speed radius R is calculated by the XGBoost model in step 2.2. max The process is to input the typhoon latitude and typhoon center pressure difference in all typhoon samples into the XGBoost model for training, and then substitute the typhoon latitude and typhoon center pressure difference into the trained XGBoost model to predict the maximum wind speed radius R max .

[0088] Specific implementation method 5: This implementation method is different from the specific implementation method 4 in that the maximum wind speed radius R is calculated by the XGBoost model in step 2.2 max The method is replaced by using the empirical formula to calculate the maximum wind speed radius R max , the empirical formula is as follows:

[0089]

[0090] Specific implementation method 6: This implementation method is different from the specific implementation methods 1 to 5 in that the wind field V in step 2.4 is transferred t The calculation formula is as follows:

[0091] V t =c·e -r / 500

[0092] Where c represents the translation speed of the typhoon point.

[0093] Specific implementation method 7: The difference between this implementation method and any one of specific implementation methods 1 to 6 is that in step 3, the target area is discretized into a plurality of grid units, and the interval between the longitude and latitude of each grid unit is 0.01 degree.

[0094] Specific implementation eight: This implementation differs from specific implementations one to seven in that the quantum number N in step 4.1 HQ It accounts for 4% to 8% of the total sample.

[0095] Specific embodiment 9: This embodiment is different from any one of the specific embodiments 1 to 8 in that in step 4.5, the distortion metric Δ is calculated. n+1 To determine whether the iteration has converged, the distortion metric formula is as follows:

[0096]

[0097] Specific implementation method 10: This implementation method is different from specific implementation method 9 in that the distortion metric meets the predefined convergence standard, that is, Δ n+1 and the Δ at the previous moment n The relative deviation (Δn+1- Δ n ) / Δ n When it is less than 0.001, the algorithm converges.

[0098] Embodiment: The typhoon disaster scene generation method of this embodiment with improved accuracy and efficiency is implemented according to the following steps:

[0099] Step 1: Select the target area (i.e., the area where the infrastructure system is located). In this embodiment, the target area is a circular area with (116.5°N, 23.5°E) as the center and a radius of 250km. Select all typhoon samples passing through the target area from the simulated typhoon database, and extract typhoon information from the selected typhoon samples. The typhoon information includes longitude, latitude, typhoon center pressure difference, and maximum wind speed;

[0100] Step 2. Step 2.1, calculate the parameters A and B in the gradient wind field model;

[0101] The calculation formula of parameter A is as follows:

[0102]

[0103] The calculation formula of parameter B is as follows:

[0104]

[0105] Among them, C 0 represents the correction parameter, the value is 1.0628, and C is calculated according to the following formula:

[0106]

[0107] f is the Coriolis parameter, r is the distance from the point where the wind speed is to be calculated to the center of the typhoon, ΔP is the pressure difference at the center of the typhoon, e is a natural constant, V m is the maximum wind speed, ρ a is the air density, taken as 1.15*10 -3 kg / m 3 ;

[0108] Step 2.2: Calculate the maximum wind speed radius R max , calculate the maximum wind speed radius R through the XGBoost model max , maximum wind speed radius R max The model is as follows:

[0109]

[0110] in, is the latitude of the typhoon, ΔP is the pressure difference at the center of the typhoon, ε Rmax is the error term;

[0111] Step 2.3: Set parameters A, B and maximum wind speed radius R max Substitute it into the gradient wind field model and calculate the gradient wind speed

[0112]

[0113] Step 2.4: Superimpose the transitional wind field and Based on the typhoon's migration wind field V t , and the total wind speed field V after superposition is obtained S :

[0114] V S =V G +V t

[0115] Step 2.5: The superimposed total wind speed field V S Converted to ground wind speed V H (t), the conversion formula is as follows:

[0116]

[0117] Where H represents the height from the ground, which is 10m; 0 It represents the length of the surface roughness. According to the land use type map of the target area, the ground type of the area is obtained, and the value is determined by referring to the following table. In this embodiment, the reference type is a suburb with sparse buildings, and the value is 0.3.

[0118] Table 1 z of different ground roughness 0 value

[0119]

[0120] Step 3: Discretize the target area into multiple grid cells and record the longitude and latitude coordinates of the center point of each grid cell, such as Figure 1 As shown, the total wind speed field V after superposition is calculated based on the improved wind field model in step 2. S , obtain the ground wind speed value of the typhoon center point at each grid unit center point at different times when (a certain) typhoon sample passes by. The wind speed diagram and wind direction diagram when a certain typhoon passes by in this embodiment are as follows Figure 2 and Figure 3 As shown;

[0121] For each grid unit center point, the maximum ground wind speed when each typhoon sample passes through is screened out, thereby forming a maximum wind speed intensity measurement map of the target area. Each typhoon sample obtains a maximum wind speed intensity measurement map. This embodiment removes the maximum wind speed intensity measurement maps with too small wind speeds, and finally forms 4383 maximum wind speed intensity measurement maps;

[0122] Step 4: Use the disaster quantification method to reduce the number of maximum wind speed intensity measurement maps for all typhoon samples. The process of reducing the number of maximum wind speed intensity measurement maps is as follows:

[0123] Step 4.1: Define the sample as the maximum wind speed intensity measurement map obtained from all typhoon samples, and the total number of samples is N. events , define quantum as the reduced maximum wind speed intensity measurement map, and set the number of quanta to N HQ =200, that is, the maximum wind speed intensity measurement graph is reduced to 200; first, randomly select N from the sample HQ samples as quanta;

[0124] Step 4.2: Calculate the nth quantum All samples V i Wind speed field distance (lon,lat) Wind speed field distance The calculation formula is as follows:

[0125]

[0126] In the formula, lon represents the longitude of the location, lat represents the latitude of the location, and lat min and lat max Respectively represent the minimum and maximum latitude of the wind speed field area, lon min and lon max They represent the minimum and maximum latitudes of the wind speed field area respectively, and then according to all the grid center points in the area, the integral of the square of the difference between the sample and quantum wind speeds corresponding to each position (lon, lat) is calculated.

[0127] Step 4.3: Take the maximum wind speed intensity measurement graph V in the sample i (lon,lat) is assigned to the jth quantum Q that is closest to it j ,Right now where d ij represents the wind speed field distance between the i-th sample and quantum j, d ik Represents the wind speed field distance between the i-th sample and quantum k, thereby obtaining the maximum wind speed intensity measurement map set assigned to quantum j The number of samples in the maximum wind speed intensity measurement map set assigned to quantum j is N j ;

[0128] Step 4.4: Take the average of the samples assigned to each quantum to obtain the updated quantum

[0129]

[0130] Step 4.5: Repeat steps 4.2 to 4.4 to iterate the quantum until convergence, and obtain N HQ An optimized post-quantum;

[0131] In this embodiment, the variation of distortion with the number of iterations is shown in Table 2 below:

[0132] Table 2 Distortion changes with the number of iterations

[0133]

[0134] Step 4.6: Calculate the optimized quantum weight P j , P j The calculation formula is:

[0135]

[0136] Among them, N j is the number of samples assigned to the jth optimized quantum, N events is the total number of samples;

[0137] Step 5: Output N from step 4 HQ The optimized post-quantum and the weight of the optimized post-quantum, that is, N HQ The maximum wind speed intensity measurement diagram and the weight P of the optimized quantum j , thus generating a typhoon disaster scenario.

[0138] Embodiment 2: This embodiment is different from Embodiment 1 in that the maximum wind speed radius R is calculated by the XGBoost model in step 2.2. max The method is replaced by using the empirical formula to calculate the maximum wind speed radius R max , the empirical formula is as follows:

[0139]

[0140] Example 1 Step 4 Output is a HQ The maximum wind speed intensity measurement diagram (quantum) and its corresponding weight p j The quantizer provides an optimal representation of regional wind speed hazards based on the minimum mean square error, which can be used for subsequent wind disaster risk analysis or loss prediction assessment. In this embodiment, the wind speed exceedance probability calculated from the maximum wind speed intensity measurement map of each grid point in the study area before and after the disaster reduction is first calculated, such as Figure 4 shown.

[0141] It can be seen that after reducing a large number of wind speed samples, the accuracy still maintains a high level and the efficiency is greatly improved. Then, an actual power grid is analyzed to obtain the number of transmission tower damages before the HQ typhoon disaster scene diagram is reduced and the number of transmission tower damages under the typhoon disaster scene diagram after the HQ method of the present invention is reduced, and the exceedance probability curve is drawn and compared, as shown in FIG. Figure 5 shown.

[0142] The accuracy of the damage analysis of the power grid before and after the disaster was reduced also remained basically unchanged. The number of Monte Carlo simulations of power grid damage and recovery in this embodiment is 50,000. If the simulation is based on the previous 4,383 typhoon events, the calculation amount reaches 4,383*50,000. After the disaster was reduced using the HQ method, the efficiency was increased by about 20 times, which is of great significance for the damage analysis, resilience assessment and resilience improvement of infrastructure systems (such as power grids) under a large number of artificially simulated disaster scenarios.

[0143] In addition, this embodiment also makes a comparison. That is, compared with R in Example 2 max The estimated wind speed of a typhoon passing through several real recording stations and the maximum wind speed of the simulation were calculated and compared using empirical formulas. Figure 6 It can be seen that the improvement effect of the method proposed in the present invention is more obvious, and the typhoon model is more accurate.

Claims

1. A method for generating typhoon disaster scenes with improved accuracy and efficiency, characterized in that The typhoon disaster scene generation method is implemented according to the following steps: Step 1: Select the target area, select all typhoon samples passing through the target area from the simulated typhoon database, and extract typhoon information from the selected typhoon samples. The typhoon information includes longitude, latitude, typhoon center pressure difference, maximum wind speed and translation speed; Step 2. Step 2.1, calculate the parameters A and B in the gradient wind field model; The calculation formula of parameter A is as follows: The calculation formula of parameter B is as follows: Among them, C and C0 represent correction parameters, f is the Coriolis parameter, r is the distance between the wind speed calculation point and the typhoon center, ΔP is the pressure difference in the typhoon center, e is a natural constant, and V m is the maximum wind speed, ρ a is the air density; Step 2.2: Calculate the maximum wind speed radius R max , calculate the maximum wind speed radius R through the XGBoost model max , maximum wind speed radius R max The model is as follows: in, is the latitude of the typhoon, ΔP is the pressure difference at the center of the typhoon, ε Rmax is the error term; Step 2.3: Set parameters A, B and maximum wind speed radius R max Substitute it into the gradient wind field model and calculate the gradient wind speed Step 2.4: Superimpose the transitional wind field and Based on the typhoon's migration wind field V t , and the total wind speed field V after superposition is obtained S : V s =V G +V t Step 2.5: The superimposed total wind speed field V S Converted to ground wind speed V H (t), the conversion formula is as follows: Among them, H represents the target height, that is, the height from the ground, and z0 represents the surface roughness length; Step 3: Discretize the target area into multiple grid cells, record the longitude and latitude coordinates of the center point of each grid cell, and calculate the total wind speed field V after superposition based on the improved wind field model in step 2. S , thus calculating the ground wind speed value of the typhoon center at different times at the center of each grid unit when the typhoon sample passes by. For each grid unit center, the maximum ground wind speed when each typhoon sample passes by is screened out, thereby forming a maximum wind speed intensity measurement map of the target area. Each typhoon sample obtains a maximum wind speed intensity measurement map; Step 4: Use the disaster quantification method to reduce the number of maximum wind speed intensity measurement maps for all typhoon samples. The process of reducing the number of maximum wind speed intensity measurement maps is as follows: Step 4.1: Define the sample as the maximum wind speed intensity measurement diagram obtained from all typhoon samples. The maximum wind speed intensity measurement diagram of each sample is represented by V i (lon,lat) means that the total number of samples is N events , define quantum as the reduced maximum wind speed intensity measurement map, and set the number of quanta to N HQ ; First, randomly select N from the sample HQ samples as quanta; Step 4.2: Calculate the nth quantum With all samples V i Wind speed field distance (lon,lat) Represents the square of the maximum wind speed difference between the i-th sample and the n-th quantum at the center point (lon, lat) of the corresponding grid unit, the wind speed field distance The calculation formula is as follows: In the formula, lon represents the longitude of the location, lat represents the latitude of the location, and lat min and lat max Respectively represent the minimum and maximum latitude of the wind speed field area, lon min and lon max They represent the minimum and maximum latitudes of the wind speed field area, respectively; Step 4.3: Take the maximum wind speed intensity measurement graph V in the sample i (lon,lat) is assigned to the jth quantum Q that is closest to it j ,Right now , where d ij represents the wind speed field distance between the i-th sample and quantum j, d ik Represents the wind speed field distance between the i-th sample and quantum k, thereby obtaining the maximum wind speed intensity measurement map set assigned to quantum j The number of samples in the maximum wind speed intensity measurement map set assigned to quantum j is N j ; Step 4.4: Take the average of the samples assigned to each quantum to obtain the updated quantum Step 4.5: Repeat steps 4.2 to 4.4 to iterate the quantum until convergence, and obtain N HQ Optimized quantum; Step 4.6, calculate the weight P of the optimized quantum j , P j The calculation formula is: Among them, N j is the number of samples assigned to the jth optimized quantum, N events is the total number of samples; Step 5: Output N from step 4 HQ The optimized post-quantum and the weight of the optimized post-quantum, that is, N HQ The maximum wind speed intensity measurement diagram and the weight P of the optimized quantum j , thus generating a typhoon disaster scenario.

2. The method for generating a typhoon disaster scene with improved accuracy and efficiency according to claim 1, characterized in that The simulated typhoon database in step 1 is the CMA database, and the number of typhoon samples is greater than 2000.

3. The method for generating a typhoon disaster scene with improved accuracy and efficiency according to claim 1, characterized in that The calculation formula for the correction parameter C in step 2.1 is as follows:

4. The method for generating a typhoon disaster scene with improved accuracy and efficiency according to claim 1, characterized in that In step 2.2, the maximum wind speed radius R is calculated by the XGBoost model max The process is to input the typhoon latitude and typhoon center pressure difference in all typhoon samples into the XGBoost model for training, and then substitute the typhoon latitude and typhoon center pressure difference into the trained XGBoost model to predict the maximum wind speed radius R max .

5. The method for generating a typhoon disaster scene with improved accuracy and efficiency according to claim 4, characterized in that The maximum wind speed radius R calculated by the XGBoost model in step 2.2 max The method is replaced by using the empirical formula to calculate the maximum wind speed radius R max , the empirical formula is as follows:

6. The method for generating a typhoon disaster scene with improved accuracy and efficiency according to claim 1, characterized in that Transition wind field V in step 2.4 t The calculation formula is as follows: V t =c·e -r / 500 Where c represents the translation speed of the typhoon point.

7. The method for generating a typhoon disaster scene with improved accuracy and efficiency according to claim 1, characterized in that In step 3, the target area is discretized into multiple grid cells, and the interval between longitude and latitude of each grid cell is 0.01 degree.

8. The method for generating a typhoon disaster scene with improved accuracy and efficiency according to claim 1, characterized in that Step 4.1: The quantum number N HQ It accounts for 4% to 8% of the total sample.

9. The method for generating a typhoon disaster scene with improved accuracy and efficiency according to claim 1, characterized in that In step 4.5, the distortion metric Δ is calculated n+1 To determine whether the iteration has converged, the distortion metric formula is as follows:

10. The method for generating a typhoon disaster scene with improved accuracy and efficiency according to claim 9, characterized in that The distortion metric meets the predefined convergence criteria, that is, Δ n+1 and the Δ at the previous moment n The relative deviation (Δ n+1- Δ n ) / Δ n When it is less than 0.001, the algorithm converges.

Images