Intelligent extreme weather monitoring method, system and equipment based on meteorological satellite and radar and medium
By integrating meteorological satellite and radar data and utilizing feature weights and machine learning models, the problem of insufficient data fusion in existing technologies has been solved, enabling accurate prediction and risk assessment of extreme weather and improving the protection capabilities of the power system.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-13
- Publication Date
- 2026-04-03
AI Technical Summary
Existing extreme weather monitoring methods rely on a single data source, making it difficult to integrate information such as satellite cloud images, temperature and humidity with radar reflectivity and radial velocity, resulting in inaccurate identification of extreme weather. Risk assessment still mainly relies on experience-based judgment, lacking data-driven quantitative analysis, and is unable to predict the specific impact range and risk level of extreme weather on the power distribution network system.
By integrating multi-source data from meteorological satellites and radar, and through spatiotemporal registration, preprocessing, feature weight calculation, and machine learning models, the probability of extreme weather is predicted. Combined with historical impact data of extreme weather, a comprehensive risk value is calculated to generate targeted early warning information.
It has achieved a quantitative mapping from meteorological data to risk levels, improved the accuracy of extreme weather identification and the scientific nature of distribution network risk assessment, provided precise protection support for the power system, and reduced distribution network failure losses caused by extreme weather.
Smart Images
Figure CN121784864A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power system safety operation and maintenance technology, specifically to an intelligent monitoring method, system, equipment, and medium for extreme weather based on meteorological satellites and radar. Background Technology
[0002] Extreme weather events such as severe convection, heavy rain, hail, and strong winds can easily lead to faults such as icing of power lines, tower collapse, and short circuits, causing widespread power outages, seriously threatening the safe operation of the power distribution network, and causing economic losses and social impact.
[0003] Current monitoring methods mostly rely on single data sources. Meteorological satellites can cover a wide area and provide information such as atmospheric temperature and humidity, but their updates are slow and cannot keep up with the rapid changes in short-term severe convection. Radar can capture details such as precipitation intensity and wind speed in local areas in real time, but its detection range is limited and cannot cover the entire region. Due to the lack of effective means to integrate satellite and radar data, existing monitoring systems struggle to combine satellite cloud images, temperature and humidity with radar reflectivity, radial velocity, and other information, resulting in inaccurate identification of extreme weather. At the same time, risk assessment still mainly relies on experience-based judgment and lacks data-driven quantitative analysis, making it impossible to predict the specific impact range and risk level of extreme weather on the power distribution network system, and hindering the power sector's ability to deploy preventative measures in advance.
[0004] Therefore, there is an urgent need for a system that integrates multi-source data from meteorological satellites and radar to make probabilistic predictions of extreme weather, conduct risk assessments and graded early warnings, and provide scientific decision support for the emergency protection of power distribution network systems. Summary of the Invention
[0005] In view of the above-mentioned problems, the present invention is proposed.
[0006] Therefore, the technical problem solved by this invention is that extreme weather monitoring methods rely on a single data source, making it difficult to integrate information such as satellite cloud images, temperature and humidity with radar reflectivity and radial velocity, resulting in inaccurate identification of extreme weather; risk assessment still mainly relies on experience-based judgment, lacking data-based quantitative analysis, and is unable to predict the specific impact range and risk level of extreme weather on the power distribution network system.
[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an intelligent monitoring method for extreme weather based on meteorological satellites and radar, comprising the following steps, Historical satellite meteorological data and historical radar meteorological data are collected, and the historical satellite meteorological data and historical radar meteorological data are spatiotemporally registered and preprocessed to obtain historical meteorological data. Acquire historical extreme weather data and combine it with historical meteorological data to calculate the feature weights of each feature in the historical meteorological data; Optimized meteorological data is obtained by filtering historical meteorological data based on feature weights. The probability of extreme weather events is predicted based on optimized meteorological data using machine learning models. The comprehensive risk value is calculated based on the probability of extreme weather occurrence and its impact. Early warning information is generated based on the comprehensive risk value and the corresponding extreme weather type.
[0008] As a preferred embodiment of the intelligent extreme weather monitoring method based on meteorological satellites and radar described in this invention, the step of obtaining historical meteorological data by performing spatiotemporal registration and preprocessing of historical satellite meteorological data and historical radar meteorological data includes: Collect historical satellite meteorological data and historical radar meteorological data and perform spatiotemporal registration; The historical satellite meteorological data and historical radar meteorological data after spatiotemporal registration are preprocessed to obtain preprocessed meteorological data. Historical meteorological data are obtained by normalizing the preprocessed meteorological data.
[0009] As a preferred embodiment of the intelligent extreme weather monitoring method based on meteorological satellites and radar described in this invention, the step of acquiring historical extreme weather data and calculating the feature weights of each feature in the historical meteorological data includes: Acquire historical extreme weather data corresponding to historical satellite meteorological data and historical radar meteorological data; The feature weights of each feature in the historical meteorological data corresponding to each type of extreme weather are calculated based on historical meteorological data and historical extreme weather data.
[0010] As a preferred embodiment of the intelligent extreme weather monitoring method based on meteorological satellites and radar described in this invention, the step of filtering historical meteorological data according to feature weights to obtain optimized meteorological data includes: Set a weight threshold, compare the feature weight with the weight threshold, and if the feature weight is greater than the feature weight threshold, it is determined to be a high feature weight, and the historical meteorological data corresponding to the feature weight is retained. If the feature weight is less than or equal to the feature weight threshold, it is determined to be a low feature weight, and the historical meteorological data corresponding to that feature weight is removed. Historical meteorological data, after removing low feature weights, are integrated into optimized meteorological data.
[0011] The beneficial effects of this preferred technical solution are as follows: by integrating historical meteorological data from multiple sources, including meteorological satellites and radar, it not only makes up for the limitations of monitoring from a single data source, but also eliminates redundant information and improves data utilization efficiency by calculating feature weights, thereby reducing the use of computing resources.
[0012] As a preferred embodiment of the intelligent extreme weather monitoring method based on meteorological satellites and radar described in this invention, the step of predicting the probability of extreme weather occurrence using a machine learning model based on optimized meteorological data includes: Using optimized meteorological data as input data and corresponding historical extreme weather data as output data, the input and output data are divided into training set, validation set and test set. The machine learning model is trained using the training set, hyperparameters are tuned using the validation set, and the accuracy of the machine learning model in predicting the probability of extreme weather events is evaluated using the test set. Collect real-time meteorological data, input it into a trained machine learning model, and predict the probability of extreme weather events.
[0013] As a preferred embodiment of the intelligent extreme weather monitoring method based on meteorological satellites and radar described in this invention, the formula for calculating the comprehensive risk value based on the probability of extreme weather occurrence and its impact is as follows: ; in, For the comprehensive risk value, Let be the probability of the i-th type of extreme weather. Let represent the severity of the consequences caused by the i-th type of extreme weather, and n represent the types of possible failure events.
[0014] The beneficial effects of this preferred technical solution are as follows: by using a machine learning model to predict the probability of extreme weather occurrence based on optimized meteorological data, and combining historical impact data of extreme weather to calculate the comprehensive risk value through a formula, a quantitative mapping from meteorological data to risk level is achieved, replacing the traditional manual qualitative assessment mode.
[0015] As a preferred embodiment of the intelligent extreme weather monitoring method based on meteorological satellites and radar described in this invention, the step of generating early warning information based on the comprehensive risk value and the corresponding extreme weather type includes: Set risk value thresholds and classify the comprehensive risk values into risk levels based on the risk value thresholds; The impact of extreme weather is determined based on historical impact data, and the output is the weather impact. The impacts of weather are categorized according to risk level, and early warning information is generated.
[0016] The beneficial effects of this preferred technical solution are: generating targeted early warning information based on comprehensive risk values and extreme weather types, significantly improving the accuracy of extreme weather identification and the scientific nature of distribution network risk assessment, providing accurate data support for the power system's protection against extreme weather, and reducing distribution network failure losses caused by extreme weather.
[0017] This invention provides an intelligent monitoring system for extreme weather based on meteorological satellites and radar.
[0018] To address the aforementioned technical problems, the present invention further provides the following technical solution: an intelligent monitoring system for extreme weather based on meteorological satellites and radar, comprising: Data acquisition module: Collects historical satellite meteorological data, historical radar meteorological data, historical extreme weather data, and real-time meteorological data; Data calculation module: performs spatiotemporal registration and preprocessing of historical satellite meteorological data and historical radar meteorological data to obtain historical meteorological data; acquires historical extreme weather data and calculates the feature weights of each feature in the historical meteorological data by combining the historical meteorological data; Data filtering module: Filters historical meteorological data based on feature weights to obtain optimized meteorological data; Extreme Weather Prediction Module: Predicts the probability of extreme weather events based on optimized meteorological data using machine learning models; Early warning generation module: Calculates a comprehensive risk value based on the probability of extreme weather occurrence and the impact of extreme weather, and generates early warning information based on the corresponding extreme weather type.
[0019] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the intelligent monitoring method for extreme weather based on meteorological satellites and radar.
[0020] The present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the intelligent extreme weather monitoring method based on meteorological satellites and radar.
[0021] The beneficial effects of this invention are as follows: By integrating historical meteorological data from multiple sources, including meteorological satellites and radar, this invention not only overcomes the limitations of monitoring from a single data source but also improves data utilization efficiency and reduces the use of computing resources by calculating feature weights to eliminate redundant information. Through machine learning models, it predicts the probability of extreme weather events based on optimized meteorological data and calculates a comprehensive risk value using formulas, combining historical impact data of extreme weather. This achieves a quantitative mapping from meteorological data to risk levels, replacing the traditional manual qualitative assessment model. Finally, it generates targeted early warning information based on the comprehensive risk value and the type of extreme weather, significantly improving the accuracy of extreme weather identification and the scientific nature of distribution network risk assessment. This provides precise data support for power system protection against extreme weather and reduces distribution network failure losses caused by extreme weather. Attached Figure Description
[0022] To more clearly illustrate the technical solutions of 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.
[0023] Figure 1 The present invention provides an overall flowchart of an intelligent extreme weather monitoring method based on meteorological satellites and radar, which is an embodiment of the present invention. Detailed Implementation
[0024] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.
[0025] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a method for intelligent monitoring of extreme weather based on meteorological satellites and radar, including: S100: Collects historical satellite meteorological data and historical radar meteorological data, performs spatiotemporal registration of historical satellite meteorological data and historical radar meteorological data and performs preprocessing to obtain historical meteorological data. S200: Acquire historical extreme weather data and combine it with historical meteorological data to calculate the feature weights of each feature in the historical meteorological data; S300: Optimized meteorological data is obtained by filtering historical meteorological data based on feature weights; S400: Predicts the probability of extreme weather events based on optimized meteorological data using machine learning models; S500: Calculates a comprehensive risk value based on the probability of extreme weather occurrence and the impact of extreme weather. S600: Generates early warning information based on the comprehensive risk value and the corresponding extreme weather type.
[0026] It should be noted that existing extreme weather monitoring methods rely on a single data source, making it difficult to integrate information such as satellite cloud images, temperature and humidity with radar reflectivity and radial velocity, resulting in inaccurate identification of extreme weather. Risk assessment still mainly relies on experience-based judgment, lacking data-driven quantitative analysis, and cannot predict the specific impact range and risk level of extreme weather on the power distribution network system. Therefore, it is very important to propose an intelligent extreme weather monitoring method based on meteorological satellites and radar.
[0027] Therefore, to address the aforementioned issues of extreme weather prediction and risk warning, a smart extreme weather monitoring method based on meteorological satellites and radar is constructed through steps S100-S600. By integrating multi-source historical meteorological data from meteorological satellites and radar, the limitations of monitoring from a single data source are overcome. Furthermore, redundant information is eliminated by calculating feature weights, improving data utilization efficiency and reducing computational resource consumption. A machine learning model predicts the probability of extreme weather occurrence based on optimized meteorological data. Combined with historical impact data of extreme weather, a comprehensive risk value is calculated using a formula, achieving a quantitative mapping from meteorological data to risk levels, replacing the traditional manual qualitative assessment model. Finally, targeted early warning information is generated based on the comprehensive risk value and extreme weather type, significantly improving the accuracy of extreme weather identification and the scientific nature of distribution network risk assessment. This provides precise data support for power system protection against extreme weather and reduces distribution network failure losses caused by extreme weather.
[0028] Example 2, refer to Figure 1 This is the second embodiment of the present invention, which provides an intelligent monitoring method for extreme weather based on meteorological satellites and radar.
[0029] In this embodiment of the application, step S100, which involves spatiotemporal registration of historical satellite meteorological data and historical radar meteorological data and preprocessing them to obtain historical meteorological data, includes the following steps A1 to A3: A1: Collect historical satellite meteorological data and historical radar meteorological data and perform spatiotemporal registration; In this embodiment, historical satellite meteorological data and historical radar meteorological data for a certain region and time period are collected. The historical satellite meteorological data is updated every 10 minutes with a spatial resolution of 1 km, and the historical radar meteorological data is updated every 6 minutes with a spatial resolution of 100 m. Through a time interpolation algorithm, the time series of the historical satellite meteorological data is resampled from a time interval of 10 minutes to a time interval of 6 minutes to keep it synchronized with the time interval of the historical radar meteorological data, thus achieving time interval alignment. Then, through spatial resampling technology, the 1 km spatial resolution of the historical satellite data is refined to a spatial resolution of 100 m, consistent with the historical radar data, thus achieving spatial structural alignment and spatiotemporal registration of the historical satellite meteorological data and the historical radar meteorological data.
[0030] A2: Preprocess the spatiotemporally registered historical satellite meteorological data and historical radar meteorological data to obtain preprocessed meteorological data; In this embodiment, the spatiotemporally registered historical satellite meteorological data is radiometrically calibrated and geometrically corrected to eliminate image position deviations caused by satellite attitude and Earth curvature; the spatiotemporally registered historical radar meteorological data is subjected to ground clutter suppression using a Kalman filter algorithm and the reflectivity factor is attenuated and corrected based on atmospheric humidity profile data; the preprocessed historical satellite meteorological data and historical radar meteorological data are output as preprocessed meteorological data. A3: Normalize the preprocessed meteorological data to obtain historical meteorological data.
[0031] In this embodiment of the application, the formula for normalizing the preprocessed meteorological data is as follows: ; in, For normalized preprocessed meteorological data, This is the minimum value in the preprocessed meteorological data. The maximum value in the preprocessed meteorological data. Preprocessed meteorological data before normalization; In this embodiment of the application, the minimum wind speed data included in the three sets of preprocessed meteorological data is set to 5 m / s and the maximum wind speed data is set to 30 m / s. The wind speed data in a certain set of preprocessed meteorological data is 20 m / s. Substituting it into the formula for normalizing the preprocessed meteorological data, the normalized wind speed data is calculated to be 0.6.
[0032] In this embodiment of the application, step S200, which involves acquiring historical extreme weather data and calculating the feature weights of each feature in the historical meteorological data, includes the following steps B1 to B2: In one alternative implementation, the normalization of preprocessed meteorological data can also be achieved through Z-score standardization. The Z-score standardization formula transforms the preprocessed meteorological data into a distribution with a mean of 0 and a standard deviation of 1, making it easier to fairly measure the contribution of each feature when calculating the feature weights in the subsequent process.
[0033] In another alternative implementation, the normalization of preprocessed meteorological data can also be achieved through maximum absolute value standardization. The normalized value is calculated by using the maximum absolute value standardization formula for the preprocessed meteorological data, which preserves the positive and negative characteristics of the data and compresses the data range to the interval [-1,1].
[0034] B1: Obtain historical extreme weather data corresponding to historical satellite meteorological data and historical radar meteorological data; In this application embodiment, extreme weather records of a certain region and time period are collected, including the occurrence of extreme weather events such as severe convection, rainstorms, and typhoons, to form historical extreme weather data that is spatiotemporally matched with historical satellite meteorological data and historical radar meteorological data.
[0035] B2: Calculate the feature weights of each feature in the historical meteorological data corresponding to each type of extreme weather based on historical meteorological data and historical extreme weather data.
[0036] In this embodiment of the application, the formula for calculating the feature weights of each feature in the historical meteorological data corresponding to each extreme weather event is as follows: ; in, Let the feature weight be the j-th feature in the historical meteorological data. Let be the measured value of the j-th feature in the k-th sample from historical meteorological data. is the extreme weather label value corresponding to the kth sample in the historical extreme weather data, with 1 for extreme weather occurring and 0 for no extreme weather occurring, and m is the total number of samples; It should be noted that the feature weight of the j-th feature in historical meteorological data is used to measure the contribution of that feature to extreme weather identification; the larger the absolute value, the higher the contribution. The sum of covariances indicates a stronger positive correlation between the feature and extreme weather, while a smaller sum indicates a weaker or negative correlation. The feature weight calculation formula can quantify the contribution of historical meteorological data and historical extreme weather data to each type of extreme weather, providing a basis for subsequent feature selection and model training.
[0037] In one alternative implementation, the feature weights of each feature in the historical meteorological data corresponding to each extreme weather event can also be calculated using random forest importance information. Taking severe convective weather identification as an example, multiple meteorological features are selected and input into the random forest model along with the severe convective weather label for training. The model constructs multiple decision trees, calculates the contribution of each feature when splitting at the decision tree node, and finally outputs the importance weights of each feature.
[0038] In another alternative implementation, the feature weights of each feature in the historical meteorological data corresponding to each extreme weather can also be calculated using a gradient boosting tree. Taking severe convective weather identification as an example, multiple meteorological features are selected, and a gradient boosting tree model is trained in combination with severe convective weather labels. A weak decision tree is generated iteratively, and the residual of the previous round is fitted each time. Finally, the feature weights are calculated based on the sum of the gains of node splits in all trees.
[0039] In this embodiment of the application, step S300, which filters historical meteorological data according to feature weights to obtain optimized meteorological data, includes the following steps C1 to C3: C1: Set a weight threshold, compare the feature weight with the weight threshold, if the feature weight is greater than the feature weight threshold, it is determined to be a high feature weight, and the historical meteorological data corresponding to the feature weight is retained; In this embodiment of the application, taking rainstorm weather as an example, the feature weights of each feature in the historical meteorological data for severe convective weather are calculated in step B2 as follows: air humidity 0.73, wind speed 0.66, cloud thickness 0.72, and cloud temperature 0.68. The weight threshold is set to 0.7. After comparison, the feature weights of air humidity and cloud thickness are greater than the weight threshold. Therefore, air humidity and cloud thickness are determined to be high feature weights, and the historical meteorological data corresponding to air humidity and cloud thickness are retained. C2: If the feature weight is less than or equal to the feature weight threshold, it is determined to be a low feature weight, and the historical meteorological data corresponding to that feature weight is removed. In this embodiment of the application, taking rainstorm weather as an example, the feature weights of each feature in the historical meteorological data for severe convective weather are calculated in step B2 as follows: air humidity 0.73, wind speed 0.66, cloud thickness 0.72, and cloud temperature 0.68. The weight threshold is set to 0.7. After comparison, the feature weights of wind speed and cloud temperature are less than the weight threshold. Therefore, wind speed and cloud temperature are determined to be high feature weights, and the historical meteorological data corresponding to wind speed and cloud temperature are removed. C3: Integrate historical meteorological data after removing low feature weights into optimized meteorological data.
[0040] In this embodiment of the application, the historical meteorological data of the two features of air humidity and cloud thickness are retained and re-integrated according to the sample dimension. Each sample contains the quantified values of the two features of air humidity and cloud thickness, and is output as optimized meteorological data.
[0041] In this embodiment of the application, step S400, which uses a machine learning model to predict the probability of extreme weather based on optimized meteorological data, includes the following steps D1 to D3: D1: Using optimized meteorological data as input data and corresponding historical extreme weather data as output data, the input and output data are divided into training set, validation set and test set. In this embodiment of the application, taking the probability prediction of rainstorm occurrence as an example, the optimized meteorological data obtained after screening in step S300 is used as input data, and the corresponding historical extreme weather data, i.e. whether rainstorm occurs, is used as output data. The input data and output data are divided into training set, validation set and test set in a ratio of 7:2:1.
[0042] D2: The machine learning model is trained using the training set, hyperparameters are tuned using the validation set, and the accuracy of the machine learning model in predicting the probability of extreme weather events is evaluated using the test set. In this embodiment, the machine learning model uses a recurrent neural network. The recurrent neural network model is trained using the training set defined in step D1, with 128 hidden layer neurons and a time step of 5. The learning rate and number of iterations are adjusted using the validation set defined in step D1. The prediction accuracy of the trained recurrent neural network model is evaluated using the test set defined in step D1. If the accuracy is greater than 90%, the training of the recurrent neural network model is considered successful and the model can be used. If the accuracy is less than or equal to 90%, the training of the recurrent neural network model is considered unsuccessful and needs to be retrained. In one alternative implementation, the machine learning model for predicting the probability of extreme weather can also employ a support vector machine. For predicting the probability of rainstorms, the support vector machine is trained with optimized meteorological data as input and rainstorm occurrence labels as output. A radial basis function is selected as the kernel function, and the regularization parameters and kernel function parameters are adjusted through a validation set. When evaluated with a test set, if the accuracy of rainstorm prediction is greater than 90%, the trained support vector machine is usable.
[0043] In another alternative implementation, the machine learning model for predicting the probability of extreme weather can also be implemented using a convolutional neural network. For the probability prediction of rainstorm weather, the convolutional neural network is trained with optimized meteorological data as input and rainstorm occurrence labels as output. The number of convolutional layers is set to 3, the kernel size is 3×3, and the pooling layer size is 2×2. Hyperparameters such as the number of kernels and the learning rate are adjusted through a validation set. If the accuracy of rainstorm prediction is greater than 90% when evaluated with a test set, the trained convolutional neural network can be used.
[0044] D3: Collect real-time meteorological data, input it into a trained machine learning model, and predict the probability of extreme weather events.
[0045] In this embodiment of the application, feature data corresponding to the high feature weights determined in step C1 of a certain area are collected by satellite and radar. The collected feature data is preprocessed and normalized in step S100 to obtain real-time meteorological data. The real-time meteorological data is input into a trained recurrent neural network model. The recurrent neural network model analyzes the real-time meteorological data based on the learned historical spatiotemporal patterns and outputs the probability of extreme weather occurring in the area in the future.
[0046] In this embodiment of the application, the formula for calculating the comprehensive risk value in step S500 based on the probability of extreme weather occurrence and the severity of the consequences caused by extreme weather is as follows: ; in, For the comprehensive risk value, Let be the probability of the i-th type of extreme weather. Let represent the severity of the consequences caused by the i-th type of extreme weather, and n represent the types of possible failure events.
[0047] In this embodiment of the application, taking two common extreme weather events in a certain region, severe convection and heavy rain, as examples, the probability of severe convection occurring is predicted to be 0.7 and the probability of heavy rain occurring is 0.6 by the machine learning model in step S400. The severity of the consequences of severe convection occurring in historical extreme weather data is 0.8 and the severity of the consequences of heavy rain occurring is 0.5. Substituting these values into the comprehensive risk value calculation formula, the comprehensive risk value is calculated to be 0.86.
[0048] In an alternative implementation, the comprehensive risk value can also be calculated through risk coupling, taking into account the interaction and risk superposition effect between different extreme weather events. Taking the scenario where an extreme weather event simultaneously causes conductor icing and tower collapse as an example, the probability and severity of the occurrence of conductor icing failure and the probability and severity of the occurrence of tower collapse failure are determined separately. Then, a coupling coefficient is introduced to quantify the degree to which the risks of these two extreme weather events mutually reinforce each other. The product of the probability and severity of each failure is multiplied and summed, and finally the coupling coefficient is calculated. The result is the comprehensive risk value of the extreme weather event.
[0049] In another optional implementation, the comprehensive risk value can also be calculated using risk entropy. Risk entropy quantifies the uncertainty of fault risk, and then the comprehensive risk value is derived by combining it with the severity of the impact. First, the three core faults that a certain extreme weather may induce are identified: conductor icing, tower collapse, and conductor breakage. Historical data is collected to determine the probability of each fault occurring, and the severity of the impact of each type of fault is assessed. Then, the risk entropy of the probability distribution of these three types of faults is calculated. The larger the entropy value, the stronger the uncertainty of the fault occurrence. The average severity of the impact of the three types of faults is calculated. Finally, the risk entropy is multiplied by the average severity of the impact, and the result is the comprehensive risk value of the extreme weather.
[0050] In this embodiment of the application, step S600, which generates early warning information based on the comprehensive risk value and the corresponding extreme weather type, includes the following steps E1 to E3: E1: Set a risk value threshold and classify the comprehensive risk value into risk levels based on the risk value threshold; In this application embodiment, the risk value threshold includes a first risk value threshold and a second risk value threshold. When the comprehensive risk value is less than the first risk value threshold, it is a low risk level. When the comprehensive risk value is between the first risk value threshold and the second risk value threshold, it is a medium risk level. When the comprehensive risk value is greater than the second risk value threshold, it is a high risk level. In this embodiment of the application, the first risk value threshold is 0.5, the second risk value threshold is 0.85, when the comprehensive risk value is less than 0.5, it is a low risk level, when 0.5 ≤ comprehensive risk value ≤ 0.85, it is a medium risk level, and when the comprehensive risk value is greater than 0.85, it is a high risk level. In step S500, the comprehensive risk value of the conductor icing is calculated to be 0.86, which is greater than the second risk value threshold of 0.85, so it is a high risk level.
[0051] E2: Determine the impact of extreme weather based on historical impact data, and output the weather impact; In this application embodiment, the impact of severe convective weather in historical data of a certain region is obtained, and the weather impact caused by severe convective weather includes damage to transformer windings, line tripping, and equipment damage such as tower collapse, resulting in large-scale power outages; the weather impact caused by rainstorms includes equipment damage such as decreased equipment insulation performance and loosening of tower foundations, resulting in local power outages.
[0052] E3: Classify weather impacts according to risk level and generate early warning information.
[0053] In this application implementation, taking severe convective weather as an example, if the risk level determined by step E1 is low risk, there may be slight overheating of a small number of transformer windings, short-term tripping of individual lines, and no tower collapse; the power supply impact is a short-term power outage affecting a small local area of about 500 users, lasting for 1 hour of rain, and the maintenance team needs to be notified to strengthen the daily inspection of the lines in the area. If the risk level determined by step E1 is medium risk, some transformer windings will be damaged, multiple lines will trip, and there will be potential tilting of individual towers; the power supply impact will be a medium-sized power outage affecting approximately 2,000 users, lasting 1 to 3 hours; maintenance personnel need to be arranged to carry out emergency repairs on the damaged equipment, reinforce the tilted towers, and send outage duration forecast text messages to users in the potential outage area. If the risk level determined by step E1 is high risk, a large number of transformer windings will be damaged, multiple lines will be permanently tripped, and multiple towers will collapse. The power supply impact will be a large-scale power outage affecting approximately 10,000 households, lasting for more than 3 hours. It is necessary to immediately dispatch emergency repair teams to the scene, activate backup transformers and temporary power supply facilities, and simultaneously issue emergency power outage notices and updates on repair progress to the affected areas.
[0054] Example 3, referring to Figure 1 This is the third embodiment of the present invention, which provides an intelligent monitoring system for extreme weather based on meteorological satellites and radar, comprising: Data acquisition module: Collects historical satellite meteorological data, historical radar meteorological data, historical extreme weather data, and real-time meteorological data; Data calculation module: performs spatiotemporal registration and preprocessing of historical satellite meteorological data and historical radar meteorological data to obtain historical meteorological data; acquires historical extreme weather data and calculates the feature weights of each feature in the historical meteorological data by combining the historical meteorological data; Data filtering module: Filters historical meteorological data based on feature weights to obtain optimized meteorological data; Extreme Weather Prediction Module: Predicts the probability of extreme weather events based on optimized meteorological data using machine learning models; Early warning generation module: Calculates a comprehensive risk value based on the probability of extreme weather occurrence and the impact of extreme weather, and generates early warning information based on the corresponding extreme weather type.
[0055] Example 4, the fourth embodiment of the present invention, differs from the previous three embodiments in that: if the function is implemented as a software functional unit and sold or used as an independent product, it 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.
[0056] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.
[0057] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.
[0058] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0059] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A method for intelligent monitoring of extreme weather based on meteorological satellites and radar, characterized in that, include, Historical satellite meteorological data and historical radar meteorological data are collected, and the historical satellite meteorological data and historical radar meteorological data are spatiotemporally registered and preprocessed to obtain historical meteorological data. Acquire historical extreme weather data and combine it with historical meteorological data to calculate the feature weights of each feature in the historical meteorological data; Optimized meteorological data is obtained by filtering historical meteorological data based on feature weights. The probability of extreme weather events is predicted using machine learning models based on optimized meteorological data. The comprehensive risk value is calculated based on the probability of extreme weather occurrence and its impact. Early warning information is generated based on the comprehensive risk value and the corresponding extreme weather type.
2. The intelligent monitoring method for extreme weather based on meteorological satellites and radar as described in claim 1, characterized in that, The steps involved in obtaining historical meteorological data by performing spatiotemporal registration and preprocessing on historical satellite meteorological data and historical radar meteorological data include: Collect historical satellite meteorological data and historical radar meteorological data and perform spatiotemporal registration; The historical satellite meteorological data and historical radar meteorological data after spatiotemporal registration are preprocessed to obtain preprocessed meteorological data. Historical meteorological data are obtained by normalizing the preprocessed meteorological data.
3. The intelligent monitoring method for extreme weather based on meteorological satellites and radar as described in claim 2, characterized in that, The steps for obtaining historical extreme weather data and calculating the feature weights of each feature in the historical meteorological data include: Acquire historical extreme weather data corresponding to historical satellite meteorological data and historical radar meteorological data; The feature weights of each feature in the historical meteorological data corresponding to each type of extreme weather are calculated based on historical meteorological data and historical extreme weather data.
4. The intelligent monitoring method for extreme weather based on meteorological satellites and radar as described in claim 3, characterized in that, The steps for filtering historical meteorological data based on feature weights to obtain optimized meteorological data include: Set a weight threshold, compare the feature weight with the weight threshold, and if the feature weight is greater than the feature weight threshold, it is determined to be a high feature weight, and the historical meteorological data corresponding to the feature weight is retained. If the feature weight is less than or equal to the feature weight threshold, it is determined to be a low feature weight, and the historical meteorological data corresponding to that feature weight is removed. Historical meteorological data, after removing low feature weights, are integrated into optimized meteorological data.
5. The intelligent monitoring method for extreme weather based on meteorological satellites and radar as described in claim 4, characterized in that, The steps involved in predicting the probability of extreme weather events using machine learning models based on optimized meteorological data include: Using optimized meteorological data as input data and corresponding historical extreme weather data as output data, the input and output data are divided into training set, validation set and test set. The machine learning model is trained using the training set, hyperparameters are tuned using the validation set, and the accuracy of the machine learning model in predicting the probability of extreme weather events is evaluated using the test set. Collect real-time meteorological data, input it into a trained machine learning model, and predict the probability of extreme weather events.
6. The intelligent monitoring method for extreme weather based on meteorological satellites and radar as described in claim 5, characterized in that, The formula for calculating the comprehensive risk value based on the probability of extreme weather occurrence and its impact is as follows: ； in, For the comprehensive risk value, Let be the probability of the i-th type of extreme weather. Let represent the severity of the consequences caused by the i-th type of extreme weather, and n represent the types of possible failure events.
7. The intelligent monitoring method for extreme weather based on meteorological satellites and radar as described in claim 6, characterized in that, The steps for generating early warning information based on the comprehensive risk value and the corresponding extreme weather type include: Set risk value thresholds and classify the comprehensive risk values into risk levels based on the risk value thresholds; The impact of extreme weather is determined based on historical impact data, and the output is the weather impact. The impact of weather is categorized according to risk level, and early warning information is generated.
8. An intelligent monitoring system for extreme weather based on meteorological satellites and radar, employing the intelligent monitoring method for extreme weather based on meteorological satellites and radar as described in any one of claims 1 to 7, characterized in that, include: Data acquisition module: Collects historical satellite meteorological data, historical radar meteorological data, historical extreme weather data, and real-time meteorological data; Data calculation module: performs spatiotemporal registration and preprocessing of historical satellite meteorological data and historical radar meteorological data to obtain historical meteorological data; acquires historical extreme weather data and calculates the feature weights of each feature in the historical meteorological data by combining the historical meteorological data; Data filtering module: Filters historical meteorological data based on feature weights to obtain optimized meteorological data; Extreme Weather Prediction Module: Predicts the probability of extreme weather events based on optimized meteorological data using machine learning models; Early warning generation module: Calculates a comprehensive risk value based on the probability of extreme weather occurrence and the impact of extreme weather, and generates early warning information based on the corresponding extreme weather type.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent extreme weather monitoring method based on meteorological satellites and radar as described in any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent extreme weather monitoring method based on meteorological satellites and radar as described in any one of claims 1 to 7.