A general wind speed prediction method and device for real-time typhoon rapid warning

Abstract

The application provides a general wind speed prediction method and device for real-time typhoon rapid warning, and relates to the technical field of natural disaster warning. The method comprises the following steps: based on the simulation circle method, typhoon data analysis is performed according to coastal city information and typhoon historical information to obtain typhoon basic information; based on a mathematical statistical inference method, a model is constructed according to the typhoon basic information to obtain a parameter probability distribution model; according to the parameter probability distribution model, random sample generation is performed using the Monte Carlo method to obtain a first training data set; based on a dynamic elite reservation strategy, the first training data set and typhoon real-time information are used to optimize and train a general wind speed prediction model to be trained to obtain an optimized general wind speed prediction model. The application is an efficient and accurate general wind speed prediction method based on a GA-BP neural network and a dynamic elite reservation strategy.

Description

Technical Field

[0001] This invention relates to the field of natural disaster early warning technology, and in particular to a general wind speed prediction method and device for real-time rapid typhoon early warning. Background Technology

[0002] Among numerous natural disasters, typhoons, with their powerful destructive force and wide-ranging impact, have become one of the main threats to the southeastern coastal areas of China. Typhoons not only directly destroy coastal infrastructure such as ports, seawalls, and buildings, but can also trigger secondary disasters such as torrential rains, storm surges, and floods.

[0003] Against this backdrop, accurately assessing typhoon hazards becomes particularly crucial. This not only helps in taking effective preventative measures in advance to reduce disaster losses, but also provides a scientific basis for the design of engineering structures in coastal areas, ensuring their safety and stability under extreme weather conditions such as typhoons. However, traditional typhoon wind speed forecasting methods have many limitations. Significant differences in meteorological conditions and geographical environments across different regions lead to deviations in the application of empirical coefficients in various locations, thus affecting the accuracy of forecast results. Furthermore, traditional methods often fail to fully consider the complexity and variability of typhoons, as well as various constraints in actual manufacturing and construction processes, such as material properties, construction techniques, and cost factors, further limiting their effectiveness in practical applications.

[0004] Therefore, there is an urgent need to develop a new, more accurate, and reliable method for typhoon hazard assessment. This method should comprehensively consider various characteristics of typhoons, including their intensity, path, speed of movement, and impact range, and combine modern information technology and data analysis methods to improve the accuracy and timeliness of forecasts. Simultaneously, the method should also possess good adaptability and flexibility, capable of being adjusted and optimized according to the specific circumstances and needs of different regions, providing strong support for disaster prevention and mitigation efforts in coastal areas.

[0005] In the existing technology, there is a lack of an efficient and accurate general wind speed prediction method based on GA-BP neural network and dynamic elite retention strategy. Summary of the Invention

[0006] To address the technical problems of scarce measured typhoon data and inaccurate predictions by traditional empirical models due to variations in regional parameters in existing technologies, this invention provides a general wind speed prediction method and apparatus for real-time rapid typhoon early warning. The technical solution is as follows:

[0007] On the one hand, a general wind speed forecasting method for real-time rapid typhoon early warning is provided. This method is implemented by a general wind speed forecasting device and includes:

[0008] Acquire information on coastal cities and their corresponding historical typhoon information; based on the simulated circle method, conduct typhoon data analysis according to the coastal city information and historical typhoon information to obtain basic typhoon information;

[0009] Based on mathematical statistical inference methods, a model is constructed according to basic typhoon information to obtain a parameter probability distribution model;

[0010] Based on the parametric probability distribution model, random samples are generated using the Monte Carlo method to obtain the first training dataset.

[0011] A general wind speed prediction model to be trained is constructed based on genetic algorithm and BP neural network.

[0012] Obtain real-time typhoon information; based on the dynamic elite retention strategy, use the first training dataset and real-time typhoon information to optimize the training of the general wind speed prediction model to obtain the optimized general wind speed prediction model.

[0013] Obtain current typhoon information; based on the current typhoon information, use an optimized general wind speed prediction model to predict wind speed.

[0014] On the other hand, a general wind speed prediction device for real-time rapid typhoon warning is provided. This device is applied to a general wind speed prediction method for real-time rapid typhoon warning. The device includes:

[0015] The information acquisition module is used to acquire information about coastal cities and their corresponding historical typhoon information; based on the simulated circle method, it performs typhoon data analysis based on the coastal city information and historical typhoon information to obtain basic typhoon information;

[0016] The parameter modeling module is used to construct models based on basic typhoon information using mathematical statistical inference methods, and to obtain parameter probability distribution models.

[0017] The training set generation module is used to generate random samples using the Monte Carlo method based on the parameter probability distribution model to obtain the first training dataset;

[0018] The model building module is used to build a general wind speed prediction model to be trained based on genetic algorithms and BP neural networks.

[0019] The model training module is used to obtain real-time typhoon information. Based on the dynamic elite retention strategy, the first training dataset and real-time typhoon information are used to optimize and train the general wind speed prediction model to be trained, and obtain the optimized general wind speed prediction model.

[0020] The wind speed prediction module is used to obtain current typhoon information; based on the current typhoon information, it uses an optimized general wind speed prediction model to predict wind speed.

[0021] On the other hand, a general-purpose wind speed forecasting device is provided, the general-purpose wind speed forecasting device comprising: a processor; a memory storing computer-readable instructions, wherein when the computer-readable instructions are executed by the processor, any one of the general-purpose wind speed forecasting methods for real-time rapid typhoon warning described above is implemented.

[0022] On the other hand, a computer-readable storage medium is provided, wherein at least one instruction is stored therein, the at least one instruction being loaded and executed by a processor to implement any of the above-described general wind speed prediction methods for real-time rapid typhoon warning.

[0023] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0024] This invention proposes a general wind speed prediction method for real-time rapid typhoon early warning. Typhoon parameters are generated through probabilistic modeling and Monte Carlo simulation. A GA-BP neural network and a dynamic elite-preservation strategy are used to optimize the typhoon boundary layer empirical model, predicting typhoon wind speeds with different return periods. This model can provide real-time and rapid typhoon wind speed predictions, effectively avoiding the problems of scarce measured typhoon data and inaccurate predictions caused by regional parameter variations in traditional empirical models, significantly improving prediction accuracy and generalization ability. This invention is an efficient and accurate general wind speed prediction method based on a GA-BP neural network and a dynamic elite-preservation strategy. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a flowchart of a general wind speed prediction method for real-time rapid typhoon early warning provided by an embodiment of the present invention;

[0027] Figure 2 This is a block diagram of a general wind speed prediction device for real-time rapid typhoon early warning provided by an embodiment of the present invention;

[0028] Figure 3 This is a schematic diagram of the structure of a general wind speed prediction device provided in an embodiment of the present invention. Detailed Implementation

[0029] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0030] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0031] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent.

[0032] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0033] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0034] This invention provides a general wind speed forecasting method for real-time rapid typhoon early warning. This method can be implemented using a general wind speed forecasting device, which can be a terminal or a server. Figure 1 The flowchart shown is for a general wind speed prediction method for real-time rapid typhoon warning. The processing flow of this method may include the following steps:

[0035] S1. Obtain information on coastal cities and their corresponding historical typhoon information; based on the simulated circle method, perform typhoon data analysis according to the information on coastal cities and historical typhoon information to obtain basic typhoon information.

[0036] Optionally, based on the simulated circle method, typhoon data analysis is performed using information from coastal cities and historical typhoon data to obtain basic typhoon information, including:

[0037] Based on a preset radius and information about coastal cities, a simulated circle method is used to draw a simulated typhoon circle.

[0038] Extract typhoon path information based on historical typhoon data;

[0039] Based on the typhoon path information, select the historical typhoon information corresponding to the typhoon path that passes through the simulated typhoon circle to obtain the basic typhoon information.

[0040] One feasible implementation involves collecting relevant data on coastal cities in China, including basic information such as city names and latitude and longitude coordinates, to determine the specific location of each coastal city, so that a simulated circle can be drawn with these cities as the center.

[0041] Using a specific coastal city as the center and setting a radius of 250km, a corresponding simulated circle is drawn on a map. The simulated circle can be automatically generated by inputting the city coordinates and radius parameters using MATLAB's `coast` and `circle` functions, ensuring the accuracy and precision of the simulated circle.

[0042] Visit the official website of the China Meteorological Administration or related data sharing platforms to find and download the optimal tropical cyclone track dataset from 1949 to 2024. Ensure the data is complete and accurate, including key information such as typhoon name, time, track coordinates, and intensity.

[0043] The simulated circle is compared and analyzed with the downloaded optimal tropical cyclone path data. MATLAB is used to determine whether the typhoon path intersects with or passes through the simulated circle. For typhoons that meet the criteria, detailed information such as typhoon name, duration of impact, wind speed, and wind direction is extracted as the basic information for typhoon wind speed prediction.

[0044] S2. Based on mathematical statistical inference methods, a model is constructed according to the basic information of the typhoon to obtain a parameter probability distribution model.

[0045] Optionally, based on mathematical statistical inference methods, a model is constructed according to the basic information of the typhoon to obtain a parameter probability distribution model, including:

[0046] Based on MATLAB, parameter information is extracted from basic typhoon information to obtain a key parameter dataset. The typhoon parameter types in the key parameter dataset include annual occurrence rate, central pressure difference, movement speed, movement direction, minimum distance between path and simulation point, and maximum wind speed radius.

[0047] Multiple probability distribution models are selected based on the data characteristics of the key parameter dataset; the type of the probability distribution model corresponds to the typhoon parameter type of the key parameter dataset.

[0048] Based on multiple probability distribution models and key parameter datasets, the least squares method is used to estimate parameters and obtain a set of parameter estimates.

[0049] Based on the use of χ 2 Test method and R 2 The test method evaluates the goodness of fit based on the key parameter dataset and the set of parameter estimates, and obtains a set of goodness of fit evaluation results.

[0050] Based on multiple probability distribution models, the probability distribution model corresponding to the highest value in the set of goodness-of-fit evaluation results is selected as the parameter probability distribution model.

[0051] In one feasible implementation, specific values ​​of key parameters for each typhoon are extracted from the basic typhoon information, including annual occurrence rate (λ), central pressure difference (Δp), moving speed (Vt), moving direction (θ), minimum distance between the path and the simulation point (Dmin), and maximum wind speed radius (Rmax), forming a complete dataset of key parameters to provide basic data support for subsequent probabilistic modeling. Assuming the typhoon attack event sequence follows a homogeneous Poisson process and the annual occurrence rate λ follows a Poisson distribution; Δp is the difference between the typhoon center pressure P0 and the peripheral pressure Pe, with Pe set to 1010.0 hPa, and the range of Δp is limited to 0–135 hPa; Vt and θ are calculated based on the coordinates of the typhoon center position at two consecutive time intervals (6 h) from the China Meteorological Administration (CMA) dataset, with θ taking due north as zero degrees, clockwise as positive, and counterclockwise as negative, and the range of Vt is limited to 2–65 km / h; Dmin is calculated using the latitude and longitude of the study point and the typhoon's moving center, with positive values ​​taken when the typhoon's moving direction line is to the left of the study point and negative values ​​taken when it is to the right, and the range of Dmin is limited to -250 km to 250 km; Rmax is determined based on the wind field model.

[0052] Based on the characteristics of the collected key parameter data and the research objectives, a probabilistic modeling method was chosen. Common probability distribution models, such as the normal distribution, Poisson distribution, and Weibull distribution, were selected to model different key parameters. The possible distribution functions and probability density functions are shown in Table 1 below (Distribution Probability Density Function Table):

[0053] Table 1

[0054]

[0055] For a selected probability distribution model, the model parameters are estimated using either the least squares method or the maximum likelihood estimation method. Taking the least squares method as an example, the model parameters are solved by constructing an objective function, which minimizes the sum of squares of the differences between the observed values ​​and the model's predicted values. The maximum likelihood estimation method estimates the parameters by maximizing the likelihood function; both can be implemented using MATLAB.

[0056] After obtaining the probability distribution models and parameter estimates for each key parameter, the χ² value was used. 2 Test method and R 2 The χ² test is used to evaluate the goodness of fit and scientific validity of the model. 2The test method determines the reasonableness of a model by comparing the observed frequencies with the theoretical frequencies (calculated from the probability distribution model). Calculate χ². 2 The statistic is compared with the critical value. If χ² is... 2 If the statistic is less than the critical value, the model is considered to fit well. R 2 R-squared is a metric used to measure the degree of variation in data explained by a model. 2 The closer the value is to 1, the stronger the model's ability to interpret the data. Calculate R. 2 The value is used to determine the accuracy of the model.

[0057] By using these two testing methods, the probability distribution models of each key parameter are comprehensively evaluated, the optimal probability distribution is determined, and the model is ensured to scientifically and accurately describe the distribution characteristics of the key parameters, providing a reliable theoretical basis for subsequent research such as typhoon wind speed prediction.

[0058] S3. Based on the parameter probability distribution model, use the Monte Carlo method to generate random samples to obtain the first training dataset.

[0059] Optionally, based on the parametric probability distribution model, random samples are generated using the Monte Carlo method to obtain the first training dataset, including:

[0060] Use the Monte Carlo method to generate a set of random numbers;

[0061] Based on the parameter probability distribution model, the inverse transformation method is used to calculate the parameters according to the random number set, and the set of key typhoon parameter values ​​is obtained.

[0062] The first training dataset is obtained by randomly combining typhoon parameters based on the set of key typhoon parameter values.

[0063] In one feasible implementation, the Monte Carlo method is used to generate a large number of random numbers uniformly distributed within the interval (0,1). These random numbers will serve as input for subsequently generating virtual typhoon samples based on a probability distribution model. The generated random numbers are then substituted into the probability distribution model of key typhoon parameters for each city, and the corresponding key typhoon parameter values ​​are calculated using methods such as inverse transformation.

[0064] The values ​​of each typhoon's key parameters are randomly combined, so that one set of typhoon key parameters represents a random typhoon in a Monte Carlo simulation. Multiple typhoons form a virtual typhoon sample dataset. This data will be used for subsequent wind speed prediction model training and validation, improving the model's generalization ability and enabling it to better adapt to typhoon wind speed prediction under different conditions.

[0065] S4. Construct a general wind speed prediction model to be trained based on genetic algorithm and BP neural network.

[0066] In one feasible implementation, the backpropagation neural network (BP) model includes an input layer, hidden layers, and an output layer. The input layer has 5 nodes, the output layer has 1 node, and the number of hidden layer nodes is determined by trial and error to be between 4 and 13. The training iterations are 10,000, the learning rate is 0.01, and the target minimum error is 1 × 10⁻⁶. -5 .

[0067] The Genetic Algorithm-Backpropagation Neural Network (GA-BP) model is based on the BP neural network. It uses a genetic algorithm to optimize the number of hidden layer nodes. The initial population size is 50, the number of generations is 1000, the crossover rate is 0.8, the mutation rate is 0.1, and the remaining parameters are the same as those of the BP neural network.

[0068] S5. Obtain real-time typhoon information; Based on the dynamic elite retention strategy, use the first training dataset and real-time typhoon information to optimize and train the general wind speed prediction model to be trained, and obtain the optimized general wind speed prediction model.

[0069] Optionally, based on a dynamic elite retention strategy, the general wind speed prediction model to be trained is optimized using the first training dataset and real-time typhoon information to obtain an optimized general wind speed prediction model, including:

[0070] Using the first training dataset, train the general wind speed prediction model to obtain the general wind speed prediction model.

[0071] Based on a dynamic elite retention strategy and a genetic algorithm, the general wind speed prediction model is optimized and trained according to real-time typhoon information to obtain an optimized general wind speed prediction model.

[0072] In one feasible implementation, the data undergoes preprocessing operations such as cleaning and standardization to make it suitable for neural network training. The preprocessed dataset is then divided into a training set and a test set, in a ratio of 7:3.

[0073] The data used in this invention is the 10-minute average wind speed at a height of 10m at the target location, obtained through the following method. The wind field model is the Holland wind field, and the mathematical expression of the wind field is as follows (1):

[0074]

[0075] Where p0 is the central pressure of the typhoon, and R... max It is the radius of maximum wind speed, and B is the Holland parameter.

[0076] Combining the above equation with the gradient wind balance equation, we can obtain the gradient wind speed of the typhoon as follows (2):

[0077]

[0078] There are two unknown parameters, namely R. max B and B are two important parameters in the typhoon wind field model. Their calculation formulas directly determine the magnitude of the wind speed calculated by the wind field model, and thus affect the prediction of extreme wind speeds. Their mathematical expressions are as follows: (3) and (4):

[0079]

[0080] Where, ε lnRmax For standard deviation, when Δp < 87 hPa, ε lnRmax =0.448; when 87hPa < Δp < 120hPa, ε lnRmax = 1.137 - 0.00792Δp; when Δp > 120 hPa, ε lnRmax =0.186; ψ is latitude; ε B =0.221. Finally, the gradient wind speed was converted to the actual wind speed at a height of 10m above the corresponding ocean surface using the reduction factor method.

[0081] The general wind speed prediction model includes a BP neural network component and a genetic algorithm component.

[0082] The BP neural network consists of an input layer, hidden layers, and an output layer; the input layer has 5 nodes; and the output layer has 1 node.

[0083] The genetic algorithm is used to optimize the number of nodes in the hidden layers of the BP neural network.

[0084] In one feasible implementation, the GA-BP neural network model, due to its use of a genetic algorithm to optimize initial weights and biases, converges faster and exhibits smaller prediction errors, demonstrating better predictive performance. This model can be used for typhoon wind speed prediction in other regions, requiring only appropriate adjustments and training based on local typhoon data.

[0085] Optionally, based on a dynamic elite retention strategy and a genetic algorithm, the general wind speed prediction model is optimized and trained according to real-time typhoon information to obtain an optimized general wind speed prediction model, including:

[0086] Based on the pre-set CMA dataset, typhoon parameters are completed according to real-time typhoon information to obtain the second training dataset;

[0087] Based on the general wind speed prediction model, a third training dataset is obtained by dynamically retaining elite individuals from the second training dataset; and a fourth training dataset is constructed using the individuals not retained in the second training dataset from the third training dataset.

[0088] Based on the general wind speed prediction model, a fifth training dataset is obtained by dynamically selecting non-elite individuals from the fourth training dataset.

[0089] Based on the fifth training dataset, a genetic algorithm is used to perform crossover and mutation to obtain the sixth training dataset; the seventh training dataset is constructed based on the third and sixth training datasets.

[0090] Using the seventh training dataset, the wind speed prediction model was optimized and trained to obtain an optimized general wind speed prediction model.

[0091] In one feasible implementation, since the real-time typhoon information only includes time-history information on wind speed and direction, it is necessary to match the time-history information based on the CMA dataset. For key typhoon parameters that are missing from the measured data, such as typhoon movement speed, typhoon movement direction, and central pressure difference, etc.

[0092] In the dynamic elite retention process, in the early stage (iterations ≤ 30% of the total): Actual typhoon individuals are forcibly designated as elites, with the number being the smaller of the number of actual samples and 20% of the population size. In the later stage: the elite ratio is adjusted to 10% of the total population, and elite individuals are selected based on fitness ranking. Non-elite individuals in the population are grouped into groups of three for competition, and the top 70% of individuals by fitness are retained.

[0093] Non-elite individuals are subjected to conventional genetic algorithm crossover and mutation operations to generate new offspring. Low-fitness new individuals are eliminated to maintain population stability. The updated population is then combined with the retained elite individuals to form a new population, which is then used to train the model.

[0094] S6. Obtain current typhoon information; based on the current typhoon information, use the optimized general wind speed prediction model to predict wind speed.

[0095] In one feasible implementation, the present invention calculates the model's performance indicators such as mean square error and mean absolute error based on the difference between the predicted results and the actual results. The comparison results are shown in Table 2 below (Performance Indicator Comparison Table).

[0096] Table 2

[0097]

[0098] In one feasible implementation, the typhoon early warning system collects parameters in real time, including the typhoon's central pressure difference, movement speed, direction of movement, minimum distance between its path and a simulated point, and the distance between the current trajectory point and the typhoon. This real-time data is input into a GA-BP neural network model to quickly generate wind speed predictions for the target location. Based on the predictions, typhoon early warning information is promptly issued, providing a scientific basis for disaster prevention and mitigation.

[0099] This invention uses Xiamen as a case study, aiming to estimate the 50-year and 100-year return periods of wind speed in various areas along the southeast coast, demonstrating the effectiveness and accuracy of the general wind speed prediction model for real-time typhoon rapid early warning. A GA-BP neural network model is used for prediction, outputting the 10-minute average wind speed at a height of 10m. The predicted wind speed corresponding to each virtual typhoon sample is recorded.

[0100] Statistical analysis was performed on the predicted wind speed data to calculate the extreme wind speeds for different return periods (e.g., 10 years, 50 years, 100 years, etc.). Extreme value type I, Weibull distribution, and Generalized Pareto Distribution (GPD) models were used to fit the distribution of the predicted wind speeds and infer the extreme wind speeds for different return periods. The calculation formula for this process is as follows (5):

[0101]

[0102] Where P is the exceedance probability of extreme wind speed x; T is the return period, T = 1 / P; u is the location parameter; σ is the scale parameter; ζ is the shape parameter. When ζ > 0, the Generalized Extreme Value Distribution (GEV) is a Type II extreme value distribution; when ζ < 0, the GEV distribution is a Type III extreme value distribution; and when ζ = 0, the GEV distribution is a Type I extreme value distribution, i.e., the Gumbel distribution. Given the parameters of the extreme value distribution model, the extreme wind speed under a specific return period can be expressed as the following equation (6):

[0103]

[0104] We collected return period data for typhoon wind speed from existing standards and typhoon wind speed predictions from other related studies using different methods (such as empirical path models and Monte Carlo methods). We then compared the wind speeds predicted by the GA-BP neural network model with existing standards and other research results.

[0105] In one feasible implementation, this invention presents a general wind speed prediction model for real-time typhoon rapid early warning, demonstrating significant advantages and effectiveness in practical applications. Typhoon wind speed data predicted by this model has an error of less than 2 m / s, allowing for direct application in typhoon early warning systems and engineering structure design in coastal areas. This not only significantly reduces design and early warning costs but also lessens reliance on the experience of specialized engineers, making early warning and design work more efficient and convenient. Although there are slight differences in the predicted typhoon wind speeds under different return periods (e.g., 50 years, 100 years), these differences precisely reflect the actual changing trends of typhoon wind speeds under different conditions, providing a more accurate and detailed design basis for engineering design. In summary, this numerical example fully verifies the feasibility and stability of the general wind speed prediction model described in this invention, demonstrating that it can provide strong technical support for typhoon early warning and typhoon-resistant design of engineering structures in coastal areas, effectively improving the scientific rigor and accuracy of early warning and design work, and possessing broad application prospects and significant practical importance.

[0106] This invention proposes a general wind speed prediction method for real-time rapid typhoon early warning. Typhoon parameters are generated through probabilistic modeling and Monte Carlo simulation. A GA-BP neural network and a dynamic elite-preservation strategy are used to optimize the typhoon boundary layer empirical model, predicting typhoon wind speeds with different return periods. This model can provide real-time and rapid typhoon wind speed predictions, effectively avoiding the problems of scarce measured typhoon data and inaccurate predictions caused by regional parameter variations in traditional empirical models, significantly improving prediction accuracy and generalization ability. This invention is an efficient and accurate general wind speed prediction method based on a GA-BP neural network and a dynamic elite-preservation strategy.

[0107] Figure 2 This is a block diagram of a general wind speed forecasting device for real-time rapid typhoon warning, according to an exemplary embodiment. The device is used in a general wind speed forecasting method for real-time rapid typhoon warning. (Refer to...) Figure 2 The device includes an information acquisition module 210, a parameter modeling module 220, a training set generation module 230, a model building module 240, a model training module 250, and a wind speed prediction module 260. Among them:

[0108] The information acquisition module 210 is used to acquire information about coastal cities and corresponding historical typhoon information; based on the simulated circle method, it performs typhoon data analysis based on the information about coastal cities and historical typhoon information to obtain basic typhoon information.

[0109] The parameter modeling module 220 is used to construct a model based on basic typhoon information using mathematical statistical inference methods to obtain a parameter probability distribution model.

[0110] The training set generation module 230 is used to generate random samples using the Monte Carlo method based on the parameter probability distribution model to obtain the first training dataset;

[0111] Model building module 240 is used to build a general wind speed prediction model to be trained based on genetic algorithm and BP neural network.

[0112] Model training module 250 is used to obtain real-time typhoon information; based on the dynamic elite retention strategy, the first training dataset and real-time typhoon information are used to optimize the training of the general wind speed prediction model to be trained, and obtain the optimized general wind speed prediction model.

[0113] The wind speed prediction module 260 is used to obtain current typhoon information; based on the current typhoon information, it uses an optimized general wind speed prediction model to predict wind speed.

[0114] Optionally, the information acquisition module 210 is further used for:

[0115] Based on a preset radius and information about coastal cities, a simulated circle method is used to draw a simulated typhoon circle.

[0116] Extract typhoon path information based on historical typhoon data;

[0117] Based on the typhoon path information, select the historical typhoon information corresponding to the typhoon path that passes through the simulated typhoon circle to obtain the basic typhoon information.

[0118] Optionally, the parametric modeling module 220 is further used for:

[0119] Based on MATLAB, parameter information is extracted from basic typhoon information to obtain a key parameter dataset. The typhoon parameter types in the key parameter dataset include annual occurrence rate, central pressure difference, movement speed, movement direction, minimum distance between path and simulation point, and maximum wind speed radius.

[0120] Multiple probability distribution models are selected based on the data characteristics of the key parameter dataset; the type of the probability distribution model corresponds to the typhoon parameter type of the key parameter dataset.

[0121] Based on multiple probability distribution models and key parameter datasets, the least squares method is used to estimate parameters and obtain a set of parameter estimates.

[0122] Based on the use of χ 2 Test method and R 2 The test method evaluates the goodness of fit based on the key parameter dataset and the set of parameter estimates, and obtains a set of goodness of fit evaluation results.

[0123] Based on multiple probability distribution models, the probability distribution model corresponding to the highest value in the set of goodness-of-fit evaluation results is selected as the parameter probability distribution model.

[0124] Optionally, the training set generation module 230 is further used for:

[0125] Use the Monte Carlo method to generate a set of random numbers;

[0126] Based on the parameter probability distribution model, the inverse transformation method is used to calculate the parameters according to the random number set, and the set of key typhoon parameter values ​​is obtained.

[0127] The first training dataset is obtained by randomly combining typhoon parameters based on the set of key typhoon parameter values.

[0128] Optionally, the model training module 250 is further used for:

[0129] Using the first training dataset, train the general wind speed prediction model to obtain the general wind speed prediction model.

[0130] Based on a dynamic elite retention strategy and a genetic algorithm, the general wind speed prediction model is optimized and trained according to real-time typhoon information to obtain an optimized general wind speed prediction model.

[0131] The general wind speed prediction model includes a BP neural network component and a genetic algorithm component.

[0132] The BP neural network consists of an input layer, hidden layers, and an output layer; the input layer has 5 nodes; and the output layer has 1 node.

[0133] The genetic algorithm is used to optimize the number of nodes in the hidden layers of the BP neural network.

[0134] Optionally, the model training module 250 is further used for:

[0135] Based on the pre-set CMA dataset, typhoon parameters are completed according to real-time typhoon information to obtain the second training dataset;

[0136] Based on the general wind speed prediction model, a third training dataset is obtained by dynamically retaining elite individuals from the second training dataset; and a fourth training dataset is constructed using the individuals not retained in the second training dataset from the third training dataset.

[0137] Based on the general wind speed prediction model, a fifth training dataset is obtained by dynamically selecting non-elite individuals from the fourth training dataset.

[0138] Based on the fifth training dataset, a genetic algorithm is used to perform crossover and mutation to obtain the sixth training dataset; the seventh training dataset is constructed based on the third and sixth training datasets.

[0139] Using the seventh training dataset, the wind speed prediction model was optimized and trained to obtain an optimized general wind speed prediction model.

[0140] This invention proposes a general wind speed prediction method for real-time rapid typhoon early warning. Typhoon parameters are generated through probabilistic modeling and Monte Carlo simulation. A GA-BP neural network and a dynamic elite-preservation strategy are used to optimize the typhoon boundary layer empirical model, predicting typhoon wind speeds with different return periods. This model can provide real-time and rapid typhoon wind speed predictions, effectively avoiding the problems of scarce measured typhoon data and inaccurate predictions caused by regional parameter variations in traditional empirical models, significantly improving prediction accuracy and generalization ability. This invention is an efficient and accurate general wind speed prediction method based on a GA-BP neural network and a dynamic elite-preservation strategy.

[0141] Figure 3 This is a schematic diagram of the structure of a general-purpose wind speed prediction device provided in an embodiment of the present invention, as shown below. Figure 3 As shown, a general-purpose wind speed forecasting device may include the above-mentioned Figure 2 The illustrated general-purpose wind speed forecasting device is designed for real-time rapid typhoon warning. Optionally, the general-purpose wind speed forecasting device 310 may include a first processor 2001.

[0142] Optionally, the general-purpose wind speed prediction device 310 may also include a memory 2002 and a transceiver 2003.

[0143] The first processor 2001, memory 2002, and transceiver 2003 can be connected via a communication bus.

[0144] The following is combined with Figure 3 A detailed introduction to each component of the general-purpose wind speed forecasting device 310 is provided below:

[0145] The first processor 2001 is the control center of the general-purpose wind speed forecasting device 310. It can be a single processor or a collective term for multiple processing elements. For example, the first processor 2001 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0146] Optionally, the first processor 2001 can perform various functions of the general wind speed prediction device 310 by running or executing software programs stored in the memory 2002 and calling data stored in the memory 2002.

[0147] In a specific implementation, as one example, the first processor 2001 may include one or more CPUs, for example... Figure 3 CPU0 and CPU1 are shown in the diagram.

[0148] In a specific implementation, as one example, the general-purpose wind speed forecasting device 310 may also include multiple processors, for example... Figure 3 The first processor 2001 and the second processor 2004 are shown in the diagram. Each of these processors can be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). Here, "processor" can refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0149] The memory 2002 is used to store the software program that executes the present invention, and is controlled by the first processor 2001 to execute it. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0150] Optionally, the memory 2002 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 2002 may be integrated with the first processor 2001 or may exist independently, and may be connected via the interface circuit of the general-purpose wind speed prediction device 310. Figure 3 (Not shown in the figure) is coupled to the first processor 2001, and the embodiments of the present invention do not specifically limit this.

[0151] The transceiver 2003 is used to communicate with network devices or with terminal devices.

[0152] Alternatively, transceiver 2003 may include a receiver and a transmitter. Figure 3 (Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the sending function.

[0153] Optionally, the transceiver 2003 can be integrated with the first processor 2001, or it can exist independently and be connected to the interface circuit of the general-purpose wind speed prediction device 310. Figure 3 (Not shown in the figure) is coupled to the first processor 2001, and the embodiments of the present invention do not specifically limit this.

[0154] It should be noted that, Figure 3 The structure of the general wind speed prediction device 310 shown does not constitute a limitation on this router. Actual knowledge structure identification devices may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0155] Furthermore, the technical effect of the general wind speed forecasting device 310 can be referred to the technical effect of the general wind speed forecasting method for real-time rapid typhoon early warning described in the above method embodiments, and will not be repeated here.

[0156] It should be understood that the first processor 2001 in this embodiment of the invention may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0157] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0158] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0159] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0160] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0161] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0162] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0163] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0164] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0165] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0166] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0167] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0168] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A general wind speed prediction method for real-time rapid typhoon early warning, characterized in that, The method includes: Acquire information on coastal cities and their corresponding historical typhoon information; based on the simulated circle method, conduct typhoon data analysis according to the coastal city information and historical typhoon information to obtain basic typhoon information; Based on mathematical statistical inference methods, a model is constructed according to basic typhoon information to obtain a parameter probability distribution model; Based on the parametric probability distribution model, random samples are generated using the Monte Carlo method to obtain the first training dataset. A general wind speed prediction model to be trained is constructed based on genetic algorithm and BP neural network. Obtain real-time typhoon information; based on the dynamic elite retention strategy, use the first training dataset and real-time typhoon information to optimize the training of the general wind speed prediction model to obtain the optimized general wind speed prediction model. Obtain current typhoon information; based on the current typhoon information, use an optimized general wind speed prediction model to predict wind speed.

2. The general wind speed prediction method for real-time rapid typhoon early warning as described in claim 1, characterized in that, The method based on simulated circles analyzes typhoon data using information from coastal cities and historical typhoon data to obtain basic typhoon information, including: Based on a preset radius and information about coastal cities, a simulated circle method is used to draw a simulated typhoon circle. Extract typhoon path information based on historical typhoon data; Based on the typhoon path information, select the historical typhoon information corresponding to the typhoon path that passes through the simulated typhoon circle to obtain the basic typhoon information.

3. The general wind speed prediction method for real-time rapid typhoon early warning as described in claim 1, characterized in that, The mathematical statistical inference method, based on typhoon basic information, constructs a model to obtain a parameter probability distribution model, including: Based on MATLAB, parameter information is extracted from basic typhoon information to obtain a key parameter dataset. The typhoon parameter types in the key parameter dataset include annual occurrence rate, central pressure difference, movement speed, movement direction, minimum distance between the path and the simulation point, and maximum wind speed radius. Multiple probability distribution models are selected based on the data characteristics of the key parameter dataset; the type of the probability distribution model corresponds to the typhoon parameter type of the key parameter dataset. Based on multiple probability distribution models and key parameter datasets, the least squares method is used to estimate parameters and obtain a set of parameter estimates. Based on the use of χ 2 Test method and R 2 The test method evaluates the goodness of fit based on the key parameter dataset and the set of parameter estimates, and obtains a set of goodness of fit evaluation results. Based on multiple probability distribution models, the probability distribution model corresponding to the highest value in the set of goodness-of-fit evaluation results is selected as the parameter probability distribution model.

4. The general wind speed prediction method for real-time rapid typhoon early warning according to claim 1, characterized in that, The first training dataset is obtained by generating random samples using the Monte Carlo method based on the parametric probability distribution model, including: Use the Monte Carlo method to generate a set of random numbers; Based on the parameter probability distribution model, the inverse transformation method is used to calculate the parameters according to the random number set, and the set of key typhoon parameter values ​​is obtained. The first training dataset is obtained by randomly combining typhoon parameters based on the set of key typhoon parameter values.

5. The general wind speed prediction method for real-time rapid typhoon early warning according to claim 1, characterized in that, The method based on a dynamic elite retention strategy, using a first training dataset and real-time typhoon information, optimizes the training of the general wind speed prediction model to obtain an optimized general wind speed prediction model, including: Using the first training dataset, train the general wind speed prediction model to obtain the general wind speed prediction model. Based on a dynamic elite retention strategy and a genetic algorithm, the general wind speed prediction model is optimized and trained according to real-time typhoon information to obtain an optimized general wind speed prediction model.

6. The general wind speed prediction method for real-time rapid typhoon early warning according to claim 5, characterized in that, The general wind speed prediction model includes a BP neural network component and a genetic algorithm component. The BP neural network consists of an input layer, a hidden layer, and an output layer; the input layer has 5 nodes; and the output layer has 1 node. The genetic algorithm is used to optimize the number of nodes in the hidden layer of the BP neural network.

7. The general wind speed prediction method for real-time rapid typhoon early warning according to claim 1, characterized in that, The method, based on a dynamic elite retention strategy and a genetic algorithm, optimizes and trains a general wind speed prediction model according to real-time typhoon information to obtain an optimized general wind speed prediction model, including: Based on the pre-set CMA dataset, typhoon parameters are completed according to real-time typhoon information to obtain the second training dataset; Based on the general wind speed prediction model, a third training dataset is obtained by dynamically retaining elite individuals from the second training dataset; and a fourth training dataset is constructed using the individuals not retained in the second training dataset from the third training dataset. Based on the general wind speed prediction model, a fifth training dataset is obtained by dynamically selecting non-elite individuals from the fourth training dataset. Based on the fifth training dataset, a genetic algorithm is used to perform crossover and mutation to obtain the sixth training dataset; the seventh training dataset is constructed based on the third and sixth training datasets. Using the seventh training dataset, the wind speed prediction model was optimized and trained to obtain an optimized general wind speed prediction model.

8. A general-purpose wind speed prediction device for real-time rapid typhoon warning, wherein the general-purpose wind speed prediction device for real-time rapid typhoon warning is used to implement the general-purpose wind speed prediction method for real-time rapid typhoon warning as described in any one of claims 1-7, characterized in that, The device includes: The information acquisition module is used to acquire information about coastal cities and their corresponding historical typhoon information; based on the simulated circle method, it performs typhoon data analysis based on the coastal city information and historical typhoon information to obtain basic typhoon information; The parameter modeling module is used to construct models based on basic typhoon information using mathematical statistical inference methods, and to obtain parameter probability distribution models. The training set generation module is used to generate random samples using the Monte Carlo method based on the parameter probability distribution model to obtain the first training dataset; The model building module is used to build a general wind speed prediction model to be trained based on genetic algorithms and BP neural networks. The model training module is used to obtain real-time typhoon information. Based on the dynamic elite retention strategy, the first training dataset and real-time typhoon information are used to optimize and train the general wind speed prediction model to be trained, and obtain the optimized general wind speed prediction model. The wind speed prediction module is used to obtain current typhoon information; based on the current typhoon information, it uses an optimized general wind speed prediction model to predict wind speed.

9. A general-purpose wind speed prediction device, characterized in that, The general-purpose wind speed prediction device includes: processor; A memory storing computer-readable instructions that, when executed by the processor, implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains program code that can be invoked by a processor to execute the method as described in any one of claims 1 to 7.

Images