Strong wind disaster weather index prediction system based on long short-term memory network model

Abstract

The invention provides a gale disaster weather index prediction system based on a long short-term memory network model, and relates to the technical field of weather monitoring, and the system specifically comprises the steps: collecting historical meteorological data of a to-be-evaluated region, and the affected area and duration of a gale disaster event occurring at a corresponding time; establishing an index prediction model based on a long short-term memory network, inputting historical meteorological data as a training set into the model, and training the model by taking the affected area and duration of the corresponding strong wind disaster as labels; collecting meteorological data of the to-be-evaluated region in real time, and obtaining a predicted disaster area and a predicted duration; and collecting terrain height and gradient data to correct the predicted disaster area and the predicted duration, comparing with a preset threshold value, and dividing risk grades according to a result. The application of the short-term memory network model can learn deep association between wind speed, humidity, temperature and other factors and strong wind disaster events from historical meteorological data, and the prediction accuracy is improved.

Description

[0001] This application is a divisional application of the application with the application number 202410923710.5, the application date of July 10, 2024, and the invention name of "A Prediction System and Method for Severe Wind Disaster Weather Index Based on Deep Learning" at the time of application. Technical Field

[0002] The present invention relates to the technical field of weather monitoring, and specifically to a prediction system for severe wind disaster weather index based on a long short-term memory network model. Background Technique

[0003] Severe wind disasters pose a serious threat to human society and economy. Therefore, accurately predicting the impact of severe wind disasters is crucial. Traditional meteorological forecasting methods mainly rely on numerical weather prediction models and empirical formulas, and are unable to fully capture the impact of complex meteorological elements and terrain on severe wind disasters. In recent years, deep learning techniques, especially long short-term memory networks, have shown excellent performance in processing time series data and can effectively capture the time dependence and non-linear characteristics in meteorological data. Therefore, applying LSTM to the prediction of severe wind disaster weather index can improve the prediction accuracy.

[0004] With climate change, the frequency and intensity of extreme weather events have increased, posing higher requirements for the accurate prediction of severe wind disasters. However, the prediction ability of traditional prediction methods is limited when facing multi-variable, multi-scale, and non-linear meteorological data. Especially in areas with complex terrain, the wind force distribution is significantly affected by factors such as terrain height and slope. Traditional models often have difficulty accurately reflecting these effects, resulting in a large deviation between the prediction results and the actual disaster situation. By using the rasterization method to collect terrain height and slope data of the area to be evaluated and through the correction of terrain data, the regulatory effect of terrain on severe wind disasters is fully considered, making the prediction results more practical. In addition, with the development of sensor technology and data acquisition technology, more abundant and detailed meteorological data can be obtained. However, how to effectively utilize these data and improve the performance of the prediction model is an important challenge faced by the current meteorological prediction field.

[0005] The above information disclosed in the background technique section is only used to strengthen the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present invention is to provide a prediction system for severe wind disaster weather index based on a long short-term memory network model to solve the problems raised in the above background technique.

[0007] To achieve the above purpose, the present invention provides the following technical solutions:

[0008] A gale disaster weather index prediction system based on a long short-term memory network model, specifically including:

[0009] A data acquisition module, used to collect the affected area and duration of gale disaster events that occurred in the historical records of the area to be evaluated, and collect historical meteorological data for one week before the gale disaster events that occurred in the historical records of the area to be evaluated. The meteorological data includes wind speed, humidity, temperature, air pressure, precipitation, and cloud cover;

[0010] A model construction module, which establishes an index prediction model based on a long short-term memory network, inputs the historical meteorological data as a training set into the index prediction model, and uses the corresponding affected area and duration of gale disasters as labels to train the model to obtain a trained index prediction model;

[0011] A prediction module, which collects the meteorological data of the area to be evaluated for the most recent week, inputs it into the trained index prediction model, obtains the predicted affected area and predicted duration of gale disaster weather, and uses a rasterization method to collect the terrain height and slope data of the area to be evaluated;

[0012] A comprehensive evaluation module, used to correct the predicted affected area and predicted duration of gale disaster weather using the collected terrain height and slope data of the area to be evaluated, obtain a comprehensive weather index, compare the comprehensive weather index with a preset evaluation threshold, and divide the risk level according to the comparison result.

[0013] Further, the specific logic for collecting the wind speed, humidity, temperature, and air pressure in the week before the gale disaster events that occurred in the historical records of the area to be evaluated is as follows:

[0014] Obtain the wind speed, humidity, temperature, and air pressure in the week before the gale disaster events that occurred in the historical records of the area to be evaluated through the weather data recorded by the meteorological bureau;

[0015] Select n measurement points around the area to be evaluated, covering the four directions of east, west, south, and north, and select measurement points at equal intervals and equal distances in each direction;

[0016] The specific logic for collecting the meteorological data of the area to be evaluated for the most recent week is as follows:

[0017] Install an anemometer, a hygrometer, a thermometer, and a barometer at the measurement points;

[0018] Record the wind speed data, humidity data, temperature data, and air pressure data at different measurement points to obtain the average wind speed, average humidity, average temperature, and average air pressure of the area to be evaluated;

[0019] The formula for obtaining the average wind speed is:

[0020]

[0021] Among them, is the average wind speed, v i represents the wind speed measured at the i-th measurement point;

[0022] The formula for obtaining the average humidity is:

[0023]

[0024] Among them, is the average humidity, RH i represents the humidity measured at the i-th measurement point;

[0025] The formula for obtaining the average temperature is:

[0026]

[0027] Among them, is the average temperature, Temp i represents the temperature measured at the i-th measurement point;

[0028] The formula for obtaining the average air pressure is:

[0029]

[0030] Among them, is the average air pressure, PRES i represents the air pressure measured at the i-th measurement point.

[0031] Furthermore, the specific logic for collecting the terrain height H and slope data C of the area to be evaluated by using the rasterization method is as follows:

[0032] The area to be evaluated is divided into regular small grids according to the area size, and then the average height and slope are calculated within each grid cell to generate a continuous terrain surface model. Each cell represents an average value, and the heights and slopes of all grid cells are weighted and averaged again, and the finally obtained value is used as the terrain height H and slope C of the area to be evaluated.

[0033] Furthermore, the specific logic for collecting the precipitation M and cloud cover N in the week before the occurrence of a strong wind disaster event in the area to be evaluated is as follows:

[0034] The daily average precipitation M and daily average cloud cover N in the week before the occurrence of a strong wind disaster event in the area to be evaluated are obtained through the historical weather data recorded by the meteorological bureau;

[0035] The collected cloud cover N is the proportion of the sky covered by clouds in the disaster area during the occurrence of a strong wind disaster, expressed as a percentage, ranging from 0% to 100%;

[0036] The specific logic for collecting precipitation and cloud cover data in the area to be evaluated in the recent week is as follows:

[0037] Collect the precipitation and cloud cover in the area to be evaluated in the recent week through the historical weather data recorded by the meteorological bureau and the prediction results of weather forecasts.

[0038] Furthermore, the process of establishing an exponential prediction model based on a long short-term memory network specifically includes:

[0039] Construct multiple groups of time series data from historical meteorological data according to types as the training set, and use the affected area and duration of the occurring gale disasters as labels to build a deep learning model based on a long short-term memory network. Input the training set and labels into the deep learning model to train the model and obtain a trained exponential prediction model. The historical meteorological data includes the wind speed, humidity, temperature, air pressure, precipitation, and cloud cover in the area to be evaluated.

[0040] Furthermore, the specific logic for obtaining the predicted affected area S and predicted duration T of gale disaster weather is as follows:

[0041] Collect the meteorological data in the area to be evaluated in the recent week, including the wind speed, humidity, temperature, air pressure, precipitation, and cloud cover in the area to be evaluated, and input it into the trained exponential prediction model to output the predicted affected area S and predicted duration T of gale disaster weather.

[0042] Furthermore, the formula for obtaining the comprehensive weather index by correcting the predicted affected area and predicted duration of gale weather using the collected terrain height and slope data of the area to be evaluated is as follows:

[0043]

[0044] Among them, QS is the comprehensive weather index, H is the terrain height of the area to be evaluated, α is its preset proportional coefficient, C is the terrain slope of the area to be evaluated, β is its preset proportional coefficient, S is the predicted affected area, T is the predicted duration, ρ is the preset proportional coefficient of (S 2 +T 3 )), and C1 is a constant correction index. Both α, β, and ρ are greater than zero.

[0045] Furthermore, the process of comparing the comprehensive weather index QS with a preset evaluation threshold and dividing the risk level according to the comparison result specifically includes: setting a series of thresholds corresponding to different gale disaster risk levels, comparing the obtained comprehensive weather index with the preset thresholds, and dividing the risk level according to the comparison result:

[0046] When QS < 30, it is predicted as low risk, and the public can carry out outdoor activities normally, but they need to pay attention to weather changes;

[0047] When 30 ≤ QS < 60, it is predicted as medium risk. It is recommended that the public reduce unnecessary outdoor activities and pay attention to fixing loose items to prevent them from being blown away by the wind;

[0048] When 60 ≤ QS < 80, it is predicted as high risk. It is recommended that the public avoid outdoor activities. It is recommended that relevant departments strengthen inspections to ensure the stability of public safety facilities. Residents should ensure that doors and windows are closed tightly and outdoor items are fixed;

[0049] When QS ≥ 80, it is predicted as extremely high risk. It is recommended that the public stay indoors and avoid all outdoor activities. Relevant departments should initiate emergency response measures, such as evacuating residents in high-risk areas and closing public facilities that may be affected.

[0050] Among them, QS is the comprehensive weather index.

[0051] The present invention also provides a method for predicting the gale disaster weather index based on deep learning. The method for predicting the gale disaster weather index is executed by the above-mentioned gale disaster weather index prediction system, and the specific steps include:

[0052] Step 1: Collect the affected area and duration of the historical gale disaster events that occurred in the area to be evaluated, and collect the historical meteorological data one week before the historical gale disaster events that occurred in the area to be evaluated. The meteorological data includes wind speed, humidity, temperature, air pressure, precipitation, and cloud cover;

[0053] Step 2: Establish an index prediction model based on the long short-term memory network. Input the historical meteorological data as the training set into the index prediction model, and use the corresponding affected area and duration of the gale disaster as labels to train the model to obtain a trained index prediction model;

[0054] Step 3: Collect the meteorological data of the area to be evaluated in the recent week, input it into the trained index prediction model, obtain the predicted affected area and predicted duration of the gale disaster weather, and use the rasterization method to collect the terrain height and slope data of the area to be evaluated;

[0055] Step 4: Use the collected terrain height and slope data of the area to be evaluated to correct the predicted affected area and predicted duration of the gale disaster weather, obtain the comprehensive weather index, compare the comprehensive weather index with the preset evaluation threshold, and divide the risk level according to the comparison result.

[0056] Compared with the prior art, the beneficial effects of the present invention are:

[0057] The present invention effectively solves the limitations of traditional prediction methods in dealing with complex non-linear relationships and time series data by introducing a long short-term memory network model based on deep learning. The LSTM model can learn the deep associations between factors such as wind speed, humidity, temperature, etc. and severe wind disaster events from historical meteorological data, improving the accuracy of prediction. At the same time, topographical factors have a significant impact on wind speed and direction, which may change the path and influence range of strong winds. Therefore, they are key variables for accurate prediction. The rasterization method is used to collect terrain height and slope data of the area to be evaluated. Through the correction of terrain data, the regulatory effect of terrain on severe wind disasters is fully considered, making the prediction results more practical and closer to the actual disaster situation. This prediction method that comprehensively considers meteorological factors and topographical features not only improves the accuracy and reliability of severe wind disaster prediction, but also provides more scientific and effective decision-making support for disaster prevention and emergency management, significantly enhancing the efficiency and effectiveness of disaster response. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a schematic diagram of the overall system module of the present invention;

[0059] Figure 2 It is a schematic diagram of the overall method flow of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with specific embodiments.

[0061] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those with ordinary skills in the field to which the present invention belongs. The "first", "second", and similar terms used in the present invention do not indicate any order, quantity, or importance, but are only used to distinguish different components. The terms such as "including" or "comprising" mean that the elements or objects appearing before this term cover the elements or objects listed after this term and their equivalents, without excluding other elements or objects. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0062] Embodiment:

[0063] Please refer to Figure 1 , the present invention provides a severe wind disaster weather index prediction system based on deep learning, including:

[0064] A data acquisition module is used to collect the affected area and duration of strong wind disaster events that have occurred in the area to be evaluated historically, and collect historical meteorological data for one week prior to the strong wind disaster events that have occurred in the area to be evaluated historically. The meteorological data includes wind speed, humidity, temperature, air pressure, precipitation, and cloud cover.

[0065] In this embodiment, the specific logic for collecting the wind speed, humidity, temperature, and air pressure in the week before the strong wind disaster event occurred in the area to be evaluated is as follows:

[0066] Obtain the wind speed, humidity, temperature, and air pressure in the week before the strong wind disaster event occurred in the area to be evaluated through the weather data recorded in the historical records of the meteorological bureau.

[0067] Select n measurement points around the area to be evaluated, covering the four directions of east, west, south, and north, and select the measurement points at equal intervals and equidistantly in each direction.

[0068] The specific logic for collecting the meteorological data in the most recent week in the area to be evaluated is as follows:

[0069] Install an anemometer, a hygrometer, a thermometer, and a barometer at the measurement points.

[0070] Record the wind speed data, humidity data, temperature data, and air pressure data at different measurement points to obtain the average wind speed, average humidity, average temperature, and average air pressure in the area to be evaluated.

[0071] The formula for obtaining the average wind speed is:

[0072]

[0073] Among them, is the average wind speed, v i represents the wind speed measured at the i-th measurement point;

[0074] The formula for obtaining the average humidity is:

[0075]

[0076] Among them, is the average humidity, RH i represents the humidity measured at the i-th measurement point;

[0077] The formula for obtaining the average temperature is:

[0078]

[0079] Among them, is the average temperature, Temp i represents the temperature measured at the i-th measurement point;

[0080] The formula for obtaining the average air pressure is as follows:

[0081]

[0082] where is the average air pressure, and PRES i represents the air pressure measured at the i-th measurement point;

[0083] The specific logic for collecting the terrain height and slope data of the area to be evaluated using the rasterization method is as follows:

[0084] The area to be evaluated is divided into regular small grids according to the area size, and then the average height and slope are calculated within each grid cell to generate a continuous terrain surface model. Each cell represents an average value. The heights and slopes of all grid cells are weighted and averaged again, and the final value obtained is used as the terrain height H and slope C of the area to be evaluated.

[0085] The specific logic for collecting the precipitation M and cloud cover N in the week before the occurrence of a strong wind disaster event in the area to be evaluated is as follows:

[0086] The daily average precipitation M and daily average cloud cover N in the week before the occurrence of a strong wind disaster event in the area to be evaluated are obtained through the historical weather data recorded by the meteorological bureau;

[0087] The collected cloud cover N is the proportion of the sky covered by clouds in the disaster-stricken area during the strong wind disaster, expressed as a percentage, ranging from 0% to 100%;

[0088] The specific logic for collecting the precipitation and cloud cover data in the most recent week in the area to be evaluated is as follows:

[0089] The precipitation and cloud cover in the most recent week in the area to be evaluated are collected through the historical weather data recorded by the meteorological bureau and the prediction results of weather forecasts;

[0090] The specific method for collecting the disaster area S is as follows:

[0091] Satellite images are used to identify and measure the disaster-stricken area, and image processing software is used to analyze the images to identify the disaster-stricken area and calculate its area;

[0092] The specific method for collecting the disaster duration T is as follows:

[0093] Remote sensing technology is used to monitor the development process of the disaster event. By analyzing consecutive remote sensing images, the disaster duration T can be determined;

[0094] Historical meteorological data is a key input for training the prediction model. By analyzing the patterns and trends in historical data, the model can learn the laws of weather changes, thereby improving the accuracy of predicting future weather.

[0095] A model construction module that establishes an exponential prediction model based on a long short-term memory network. The historical meteorological data is used as a training set and input into the exponential prediction model, and the affected area and duration of the corresponding gale disaster are used as labels to train the model, obtaining a trained exponential prediction model;

[0096] The process of establishing the exponential prediction model based on the long short-term memory network specifically includes:

[0097] The historical meteorological data is grouped into multiple time series data according to types and used as a training set. The affected area and duration of the gale disasters that occurred are used as labels to construct a deep learning model based on the long short-term memory network. The training set and labels are input into the deep learning model to train the model, obtaining a trained exponential prediction model. The historical meteorological data includes wind speed, humidity, temperature, air pressure, precipitation, and cloud cover in the area to be evaluated;

[0098] LSTM is a special type of recurrent neural network (RNN) that can effectively process and predict long-term dependencies in time series. In meteorological data, weather patterns often have complex time dependencies, and LSTM can capture these patterns, thereby improving the accuracy of prediction; The LSTM model can directly map from the original input data (wind speed, temperature, humidity, etc.) to the output labels (affected area and duration), achieving end-to-end learning. This direct mapping simplifies the model training process and may improve the accuracy of prediction.

[0099] A prediction module that collects the meteorological data of the area to be evaluated in the recent week, inputs it into the trained exponential prediction model, obtains the predicted affected area and predicted duration of the gale disaster weather, and uses the rasterization method to collect the terrain height and slope data of the area to be evaluated;

[0100] In this embodiment, the specific logic for obtaining the predicted affected area S and predicted duration T of the gale disaster weather is as follows:

[0101] Collect the meteorological data of the area to be evaluated in the recent week, including wind speed, humidity, temperature, air pressure, precipitation, and cloud cover in the area to be evaluated, input it into the trained exponential prediction model, and output to obtain the predicted affected area S and predicted duration T of the gale disaster weather;

[0102] Real-time collection of meteorological data can ensure that the input to the model is the latest information, which helps improve the timeliness and accuracy of predictions. Compared with technologies that rely on historical data, real-time data can better reflect the current weather conditions. Combining terrain height and slope data can more comprehensively evaluate the impact of strong wind disasters on different terrain areas. This multi-dimensional data fusion helps improve the prediction ability of the model because it takes into account the influence of terrain on wind speed and direction, enabling the model to better adapt to prediction requirements in different geographical environments and thus enhancing the generalization ability of the model.

[0103] A comprehensive evaluation module is used to correct the predicted affected area and predicted duration of strong wind disaster weather using the collected terrain height and slope data of the area to be evaluated, obtain a comprehensive weather index, compare the comprehensive weather index with a preset evaluation threshold, and divide the risk level according to the comparison result;

[0104] In this embodiment, the formula for obtaining the comprehensive weather index by correcting the predicted affected area and predicted duration of strong wind weather using the collected terrain height and slope data of the area to be evaluated is as follows:

[0105]

[0106] Among them, QS is the comprehensive weather index, H is the terrain height of the area to be evaluated, α is its preset proportional coefficient, C is the terrain slope of the area to be evaluated, β is its preset proportional coefficient, S is the predicted affected area, T is the predicted duration, ρ is the preset proportional coefficient of (S 2 +T 3 )), C1 is a constant correction index, and α, β, and ρ are all greater than zero; when the terrain height H of the area to be evaluated increases, the comprehensive weather index QS increases because it will cause a change in wind direction, resulting in uneven wind distribution in different terrain areas. This change in wind direction may cause some areas to be affected by stronger wind forces, thus increasing the risk of disasters; when the terrain slope C increases, the comprehensive weather index QS increases. Steep slopes may cause stronger local wind effects, such as slope winds or downslope winds. These local wind effects may generate extreme wind speeds in specific areas, increasing the risk of disasters; when the predicted affected area S increases, QS increases; when the predicted duration T increases, QS will also increase accordingly; that is, it shows that there is a positive correlation between H, C, S, T, and the comprehensive weather index QS;

[0107] The process of comparing the comprehensive weather index QS with the preset evaluation threshold and dividing the risk level according to the comparison result specifically includes: setting a series of thresholds corresponding to different strong wind disaster danger levels, comparing the obtained comprehensive weather index with the preset thresholds, and dividing the danger level according to the comparison result:

[0108] When QS < 30, it is predicted as low risk, and the public can carry out outdoor activities normally, but they need to pay attention to weather changes;

[0109] When 30 ≤ QS < 60, it is predicted as medium risk. It is recommended that the public reduce unnecessary outdoor activities and pay attention to fixing loose items to prevent them from being blown away by the wind;

[0110] When 60 ≤ QS < 80, it is predicted as high risk. It is recommended that the public avoid outdoor activities, and it is recommended that relevant departments strengthen inspections to ensure the stability of public safety facilities. Residents should ensure that doors and windows are closed tightly and outdoor items are fixed;

[0111] When QS ≥ 80, it is predicted as extremely high risk. It is recommended that the public stay indoors and avoid all outdoor activities. Relevant departments should initiate emergency response measures, such as evacuating residents in high-risk areas and closing public facilities that may be affected.

[0112] Among them, QS is the comprehensive weather index.

[0113] Please refer to Figure 2 , the present invention also provides a method for predicting the strong wind disaster weather index based on deep learning. The method for predicting the strong wind disaster weather index is executed by the above-mentioned strong wind disaster weather index prediction system. The specific steps include:

[0114] Step 1: Collect the affected area and duration of the strong wind disaster events that occurred in the area to be evaluated historically, and collect the historical meteorological data of the week before the strong wind disaster events that occurred in the area to be evaluated historically. The meteorological data includes wind speed, humidity, temperature, air pressure, precipitation, and cloud cover;

[0115] Step 2: Establish an index prediction model based on the long short-term memory network. Input the historical meteorological data as the training set into the index prediction model, and use the corresponding affected area and duration of the strong wind disaster as labels to train the model to obtain a trained index prediction model;

[0116] Step 3: Collect the meteorological data of the area to be evaluated in the most recent week, input it into the trained index prediction model to obtain the predicted affected area and predicted duration of the strong wind disaster weather, and use the rasterization method to collect the terrain height and slope data of the area to be evaluated;

[0117] Step 4: Use the collected terrain height and slope data of the area to be evaluated to correct the predicted affected area and predicted duration of the strong wind disaster weather to obtain the comprehensive weather index. Compare the comprehensive weather index with the preset evaluation threshold, and divide the risk level according to the comparison result.

[0118] The specific values of α, β, and ρ in the formula are generally determined by those skilled in the art according to the actual situation. Those skilled in the art collect multiple sets of sample data, set corresponding preset proportionality coefficients for each set of sample data, substitute the set preset proportionality coefficients and the collected sample data into the formula, through repeated experiments and parameter adjustments, observe the accuracy of the model output and the rationality of the results, gradually adjust these factor coefficients, compare the performance and effects of the model under different parameter settings, find the optimal coefficient combination, screen the calculated factor coefficients and take the average value to obtain the values of α, β, and ρ.

[0119] In addition, the magnitude of the preset factor coefficient is a specific value obtained by quantifying each parameter. It is for the convenience of subsequent comparison. Regarding the magnitude of the coefficient, it depends on the amount of sample data and the initial setting of the corresponding preset proportionality coefficient for each set of sample data by those skilled in the art, and it is not unique, as long as it does not affect the proportional relationship between the parameter and the quantified value.

[0120] The above formulas are all dimensionless and take their numerical calculations. The formula is a formula obtained by collecting a large amount of data and performing software simulation to approximate the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0121] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0122] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0123] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all should be covered by the protection scope of the present application.

Claims

1. A prediction system for severe wind disaster weather index based on a long short-term memory network model, characterized in that, Specifically, it includes: A data acquisition module, which is used to collect the affected area and duration of the severe wind disaster events that occurred in the area to be evaluated historically, and collect the historical meteorological data of the week before the severe wind disaster events that occurred in the area to be evaluated historically. The meteorological data includes wind speed, humidity, temperature, air pressure, precipitation, and cloud cover; A model construction module, which establishes an exponential prediction model based on a long short-term memory network, inputs the historical meteorological data as a training set into the exponential prediction model, uses the corresponding affected area and duration of the severe wind disaster as labels, trains the model, and obtains a trained exponential prediction model; A prediction module, which collects the meteorological data of the area to be evaluated in the most recent week, inputs it into the trained exponential prediction model, obtains the predicted affected area and predicted duration of the severe wind disaster weather, and uses the rasterization method to collect the terrain height and slope data of the area to be evaluated; A comprehensive evaluation module, which is used to correct the predicted affected area and predicted duration of the severe wind disaster weather by using the collected terrain height and slope data of the area to be evaluated, obtains a comprehensive weather index, compares the comprehensive weather index with a preset evaluation threshold, and divides the risk level according to the comparison result; The severe wind disaster weather index prediction system executes the following severe wind disaster weather index prediction method, including: Step 1: Collect the affected area and duration of the severe wind disaster events that occurred in the area to be evaluated historically, and collect the historical meteorological data of the week before the severe wind disaster events that occurred in the area to be evaluated historically. The meteorological data includes wind speed, humidity, temperature, air pressure, precipitation, and cloud cover; Step 2: Establish an exponential prediction model based on a long short-term memory network, input the historical meteorological data as a training set into the exponential prediction model, use the corresponding affected area and duration of the severe wind disaster as labels, train the model, and obtain a trained exponential prediction model; Step 3: Collect the meteorological data of the area to be evaluated in the most recent week, input it into the trained exponential prediction model, obtain the predicted affected area and predicted duration of the severe wind disaster weather, and use the rasterization method to collect the terrain height and slope data of the area to be evaluated; Step 4: Correct the predicted affected area and predicted duration of the severe wind disaster weather by using the collected terrain height and slope data of the area to be evaluated, obtain a comprehensive weather index, compare the comprehensive weather index with a preset evaluation threshold, and divide the risk level according to the comparison result; The specific method for collecting the affected area is: using satellite images to identify and measure the affected area, analyzing the images through image processing software, identifying the affected area and calculating its area; The specific method for collecting the disaster duration is: using remote sensing technology to monitor the development process of the disaster event, and determining the duration of the disaster event by analyzing consecutive remote sensing images.

2. The gale disaster weather index prediction system based on the long short-term memory network model according to claim 1, characterized in that: The specific logic for collecting the wind speed, humidity, temperature, and air pressure in the week before the severe wind disaster events that occurred in the area to be evaluated historically is: Obtain the wind speed, humidity, temperature, and air pressure in the week before the severe wind disaster events that occurred in the area to be evaluated historically through the weather data recorded in the historical records of the meteorological bureau; Select n measurement points around the area to be evaluated, covering the four directions of east, west, south, and north, and select the measurement points equidistantly and equally in each direction; The specific logic for collecting meteorological data in the area to be evaluated in the most recent week is as follows: Install an anemometer, a hygrometer, a thermometer, and a barometer at the measurement points; Record wind speed data, humidity data, temperature data, and barometric pressure data at different measurement points to obtain the average wind speed, average humidity, average temperature, and average barometric pressure in the area to be evaluated; The formula for obtaining the average wind speed is: Among them, is the average wind speed, and v i represents the wind speed measured at the i-th measurement point; The formula for obtaining the average humidity is: Among them, is the average humidity, RH i represents the humidity measured at the i-th measurement point; The formula for obtaining the average temperature is: Among them, is the average temperature, Temp i represents the temperature measured at the i-th measurement point; The formula for obtaining the average barometric pressure is: Among them, is the average air pressure, PRES i represents the air pressure measured at the i-th measurement point.

3. The gale disaster weather index prediction system based on the long short-term memory network model according to claim 1, characterized in that: The specific logic for collecting the terrain height H and slope data C of the area to be evaluated using the rasterization method is as follows: Divide the area to be evaluated into regular small grids according to the area size, then calculate the average height and slope in each grid unit to generate a continuous terrain surface model. Each unit represents an average value. Weightedly average the heights and slopes of all grid units again, and the finally obtained value is used as the terrain height H and slope C of the area to be evaluated.

4. The gale disaster weather index prediction system based on the long short-term memory network model according to claim 2, characterized in that: The specific logic for collecting the precipitation M and cloud cover N in the week before a severe wind disaster event occurred in the area to be evaluated is as follows: Obtain the daily average precipitation M and daily average cloud cover N in the week before a severe wind disaster event occurred in the area to be evaluated through the historical weather data recorded by the meteorological bureau; The collected cloud cover N is the proportion of the sky in the disaster-stricken area covered by clouds during a severe wind disaster, expressed as a percentage, ranging from 0% to 100%; The specific logic for collecting the precipitation and cloud cover data in the most recent week in the area to be evaluated is as follows: Collect the precipitation and cloud cover in the most recent week in the area to be evaluated through the historical weather data recorded by the meteorological bureau and the prediction results of weather forecasts.

5. The gale disaster weather index prediction system based on the long short-term memory network model according to claim 1, wherein: The process of establishing an exponential prediction model based on a long short-term memory network specifically includes: Construct multiple groups of time series data from historical meteorological data according to types and use them as the training set. Use the affected area and duration of the severe wind disasters that occurred as labels to build a deep learning model based on a long short-term memory network. Input the training set and labels into the deep learning model to train the model to obtain a trained exponential prediction model. The historical meteorological data includes the wind speed, humidity, temperature, barometric pressure, precipitation, and cloud cover in the area to be evaluated.

6. The prediction system for severe wind disaster weather index based on the long short-term memory network model according to claim 1, characterized in that: The specific logic for obtaining the predicted affected area S and predicted duration T of a severe wind disaster weather is as follows: Collect the meteorological data in the most recent week in the area to be evaluated, including the wind speed, humidity, temperature, barometric pressure, precipitation, and cloud cover in the area to be evaluated, and input it into the trained exponential prediction model to output the predicted affected area S and predicted duration T of a severe wind disaster weather.

7. The gale disaster weather index prediction system based on the long short-term memory network model according to claim 6, characterized in that: The formula for obtaining the comprehensive weather index by correcting the predicted affected area and predicted duration of a severe wind weather using the collected terrain height and slope data of the area to be evaluated is as follows: Among them, QS is the comprehensive weather index, H is the terrain height of the area to be evaluated, α is its preset proportional coefficient, C is the terrain slope of the area to be evaluated, β is its preset proportional coefficient, S is the predicted affected area, T is the predicted duration, ρ is the preset proportional coefficient of (S 2 +T 3 ), C1 is the constant correction index, and α, β, and ρ are all greater than zero.

8. The gale disaster weather index prediction system based on the long short-term memory network model according to claim 7, characterized in that: The process of comparing the comprehensive weather index QS with a preset evaluation threshold and dividing the risk level according to the comparison result specifically includes: setting a series of thresholds corresponding to different high wind disaster risk levels, comparing the obtained comprehensive weather index with the preset thresholds, and dividing the risk level according to the comparison result: When QS < 30, it is predicted as a low risk. The public can carry out outdoor activities normally, but need to pay attention to weather changes; When 30 ≤ QS < 60, it is predicted as a medium risk. It is recommended that the public reduce unnecessary outdoor activities and pay attention to fixing loose items to prevent them from being blown away; When 60 ≤ QS < 80, it is predicted as a high risk. It is recommended that the public avoid outdoor activities. It is recommended that relevant departments strengthen inspections to ensure the stability of public safety facilities. Residents should ensure that doors and windows are closed tightly and outdoor items are fixed; When QS ≥ 80, it is predicted as an extremely high risk. It is recommended that the public stay indoors and avoid all outdoor activities. Relevant departments should initiate emergency response measures, such as evacuating residents in high-risk areas and closing affected public facilities; Among them, QS is the comprehensive weather index.

9. Application of the high wind disaster weather index prediction system based on the long short-term memory network model according to any one of claims 1-8 in weather monitoring.

Images