AI-based high and cold region wind direction and wind speed historical data intelligent restoration method and system

Abstract

The embodiment of the invention discloses an AI-based high and cold region wind direction and wind speed historical data intelligent restoration method and system, and belongs to the technical field of data processing. The embodiment of the invention aims to effectively improve the timeliness and accuracy of the meteorological data by updating the meteorological monitoring data set, applying the advanced data recovery algorithm and performing multiple rounds of wind load state vector mining based on the feature focusing strategy. By applying the embodiment of the invention, not only can the problem of meteorological data updating lag be solved, but also the overall quality of the meteorological monitoring data can be remarkably improved through deep data mining and intelligent restoration, so that a more reliable data basis is provided for meteorological prediction, disaster early warning and the like, and the method has remarkable technical progress and application value.

Description

Technical Field

[0001] The present application belongs to the field of data processing technology, and specifically relates to an AI-based intelligent repair method and system for historical data of wind direction and wind speed in high-cold areas. Background Art

[0002] In the field of meteorological monitoring and forecasting, the timeliness and accuracy of data are crucial. However, existing technologies face several key problems: First, meteorological monitoring data sets are often difficult to update in real time, resulting in insufficient data timeliness; second, there are many outliers in past meteorological data due to instrument errors, environmental interference and other factors. If these data are not effectively corrected, they will seriously affect the accuracy of meteorological forecasts; third, even if data corrections are made, there is often a lack of efficient data mining strategies to further improve data quality, especially in the monitoring and correction of key meteorological elements such as wind loads. Summary of the invention

[0003] The present application provides an AI-based intelligent repair method and system for historical data of wind direction and speed in high-cold areas, which can solve or partially solve the technical problems involved in the above-mentioned background technology.

[0004] The embodiment of the present application provides an AI-based intelligent repair method for historical data of wind direction and speed in high-cold areas, which is applied to a data intelligent repair system, and the method includes: obtaining a target meteorological monitoring data set; the target meteorological monitoring data set is obtained by updating the first spatiotemporal data segment in the past meteorological monitoring data set to a second spatiotemporal data segment; the first spatiotemporal data segment includes at least one first wind load monitoring information; the second spatiotemporal data segment includes at least one second wind load monitoring information; obtaining a past monitoring correction data set generated by a data repair algorithm based on the past meteorological monitoring data set, and a first vector reinforcement coefficient obtained by performing multiple rounds of first wind load state vector mining based on a feature focusing strategy when generating the past monitoring correction data set; implementing data intelligent repair processing through the data repair algorithm to obtain a first target monitoring correction data set corresponding to the target meteorological monitoring data set; wherein the data intelligent repair processing includes: performing multiple rounds of second wind load state mining based on a feature focusing strategy based on the target meteorological monitoring data set and the past monitoring correction data set. Vector mining, obtaining the second vector reinforcement coefficient corresponding to each round of second wind load state vector mining, and obtaining the wind load state spatiotemporal vector of each round of second wind load state vector mining based on the first wind load state vector corresponding to the target meteorological monitoring data set and the target vector reinforcement coefficient corresponding to each round of second wind load state vector mining, and generating the first target monitoring correction data set corresponding to the target meteorological monitoring data set based on the wind load state spatiotemporal vector of the last round of second wind load state vector mining; wherein, the target vector reinforcement coefficient corresponding to any second wind load state vector mining is the second vector reinforcement coefficient or the third vector reinforcement coefficient; the target confidence vector corresponding to at least one round of second wind load state vector mining is the third vector reinforcement coefficient obtained based on the following steps: based on the vector reinforcement coefficient corresponding to the second spatiotemporal data segment in the second vector reinforcement coefficient corresponding to the second wind load state vector mining of the target wheel, the vector reinforcement coefficient corresponding to the first spatiotemporal data segment in the first vector reinforcement coefficient of the target wheel is adjusted to obtain the third vector reinforcement coefficient.

[0005] An embodiment of the present application provides a data intelligent repair system, comprising at least one processor and a memory; the memory stores computer-executable instructions; the at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the above method.

[0006] An embodiment of the present application provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the above method are implemented.

[0007] The embodiment of the present application aims to effectively improve the timeliness and accuracy of meteorological data by updating meteorological monitoring data sets, applying advanced data repair algorithms, and multiple rounds of wind load state vector mining based on feature focusing strategies. The application of the embodiment of the present application can not only solve the problem of delayed meteorological data updates, but also significantly improve the overall quality of meteorological monitoring data through in-depth data mining and intelligent repair, thereby providing a more reliable data foundation for meteorological forecasts, disaster warnings, etc., with significant technical progress and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A flowchart of an AI-based intelligent repair method for historical data of wind direction and speed in high-cold areas provided in an embodiment of the present application.

[0009] Figure 2 A schematic diagram of the structure of a data intelligent repair system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0010] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0011] The terms "first", "second", etc. in this application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are usually a class, and the number of objects is not limited. For example, the first object can be one or more. In addition, in this application, "and / or" represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0012] Figure 1 An AI-based intelligent repair method for historical data of wind direction and speed in high-cold areas is shown, which is applied to a data intelligent repair system. The method includes the following steps 110 to 130.

[0013] Step 110, obtain a target meteorological monitoring data set; the target meteorological monitoring data set is obtained by updating a first spatiotemporal data segment in a past meteorological monitoring data set to a second spatiotemporal data segment; the first spatiotemporal data segment includes at least one first wind load monitoring information; the second spatiotemporal data segment includes at least one second wind load monitoring information.

[0014] Step 120, obtaining a past monitoring correction data set generated by a data repair algorithm based on the past meteorological monitoring data set, and a first vector enhancement coefficient obtained by performing multiple rounds of first wind load state vector mining based on a feature focusing strategy when generating the past monitoring correction data set.

[0015] Step 130: Implement intelligent data repair processing through the data repair algorithm to obtain a first target monitoring correction data set corresponding to the target meteorological monitoring data set.

[0016] Among them, the data intelligent repair processing includes: according to the target meteorological monitoring data set and the past monitoring correction data set, performing multiple rounds of second wind-load state vector mining based on the feature focusing strategy, obtaining the second vector reinforcement coefficient corresponding to each round of the second wind-load state vector mining, and according to the first wind-load state vector corresponding to the target meteorological monitoring data set and the target vector reinforcement coefficient corresponding to each round of the second wind-load state vector mining, obtaining the wind-load state spatiotemporal vector of each round of the second wind-load state vector mining, and generating the first target monitoring correction data set corresponding to the target meteorological monitoring data set based on the wind-load state spatiotemporal vector of the last round of the second wind-load state vector mining; wherein, the target vector reinforcement coefficient corresponding to any second wind-load state vector mining is the second vector reinforcement coefficient or the third vector reinforcement coefficient; the target confidence vector corresponding to at least one round of the second wind-load state vector mining is the third vector reinforcement coefficient obtained based on the following steps: based on the vector reinforcement coefficient corresponding to the second spatiotemporal data segment in the second vector reinforcement coefficient corresponding to the second wind-load state vector mining of the target wheel, the vector reinforcement coefficient corresponding to the first spatiotemporal data segment in the first vector reinforcement coefficient of the target wheel is adjusted to obtain the third vector reinforcement coefficient.

[0017] In high-altitude cold regions, historical data on wind direction and speed are extremely valuable for meteorological research, climate model construction, and prediction of extreme weather events. However, due to the extreme environmental conditions, meteorological monitoring equipment often faces problems such as freezing and jamming, resulting in errors or missing data collected. In order to solve this problem, the data intelligent repair system of the present application repairs these problematic data in an automated and intelligent manner.

[0018] The following describes in detail, in conjunction with a specific application scenario example, how the data intelligent repair system implements steps 110 to 130 to repair historical data of wind direction and speed in high-altitude cold areas.

[0019] First, the data intelligent repair system obtains the target meteorological monitoring data set. The target meteorological monitoring data set is obtained by updating the past meteorological monitoring data set, in which the first spatiotemporal data segment is updated to the second spatiotemporal data segment. The first spatiotemporal data segment contains at least one first wind load monitoring information, which is the actual monitoring data of meteorological parameters such as wind speed and wind direction in a specific high-cold area in a certain period of time in the past. The second spatiotemporal data segment contains at least one second wind load monitoring information, which is the wind load monitoring data of a new time period and location, which may be generated due to equipment updates, location changes or the passage of time.

[0020] Next, the data intelligent repair system obtains the past monitoring correction data set generated by the data repair algorithm based on the past meteorological monitoring data set. The past monitoring correction data set is the result of repairing the erroneous or missing data in the past data set. At the same time, the data intelligent repair system also obtains the first vector reinforcement coefficient obtained by multiple rounds of first wind load state vector mining based on the feature focusing strategy when generating the past monitoring correction data set. The feature focusing strategy is an advanced data mining technology that improves the efficiency and accuracy of data processing by analyzing key features in the data, such as the changing patterns of wind speed and wind direction. In this scenario, it is used to mine and strengthen data features related to wind load status in order to better repair problems in the data set.

[0021] Then, the data intelligent repair system implements the data intelligent repair processing through the data repair algorithm, and obtains the first target monitoring corrected data set corresponding to the target meteorological monitoring data set.

[0022] It is worth mentioning that the process of intelligent data repair is complex and sophisticated, and it includes multiple steps of iteration and optimization.

[0023] First, the data intelligent repair system conducted multiple rounds of second wind load state vector mining based on the feature focusing strategy based on the target meteorological monitoring data set and the past monitoring correction data set. In this process, the data intelligent repair system deeply analyzed the wind load state characteristics in the target data set and tried to find similar patterns and rules in the past correction data set. Through multiple rounds of mining and iteration, the data intelligent repair system obtained the second vector reinforcement coefficient corresponding to each round of second wind load state vector mining. The second vector reinforcement coefficient reflects the importance and repair potential of each wind load state vector in different rounds of data mining.

[0024] Subsequently, the data intelligent repair system mines the target vector reinforcement coefficients corresponding to the first wind load state vector corresponding to the target meteorological monitoring data set and the second wind load state vector of each round, and obtains the wind load state spatiotemporal vectors of each round of second wind load state vector mining. The wind load state spatiotemporal vectors are the representations of the wind load state data obtained through reinforcement and repair in different rounds of data mining. They not only contain the specific values ​​of wind speed and wind direction, but also metadata related to time and space, making the repaired data more accurate and complete.

[0025] In this process, the target vector reinforcement coefficient corresponding to any second wind load state vector mining can be the second vector reinforcement coefficient or the third vector reinforcement coefficient. The second vector reinforcement coefficient is directly calculated in the current round of data mining, which reflects the repair potential of each wind load state vector in the current round. The third vector reinforcement coefficient is obtained by adjusting the vector reinforcement coefficient corresponding to the first spatiotemporal data segment in the first vector reinforcement coefficient of the target wheel based on the vector reinforcement coefficient corresponding to the second spatiotemporal data segment in the second vector reinforcement coefficient corresponding to the second wind load state vector mining of the target wheel. This adjustment process takes into account the correlation and difference between the new and old data segments, so that the repair algorithm can better adapt to the update and change of the data set.

[0026] Finally, based on the wind load state spatiotemporal vectors mined from the last round of the second wind load state vector, the data intelligent repair system generated the first target monitoring correction data set corresponding to the target meteorological monitoring data set. The first target monitoring correction data set is the result of repairing the problems in the original target meteorological monitoring data set, and it provides more accurate and complete wind load monitoring information. Through this process, the data intelligent repair system successfully solved the problems existing in the historical data of wind direction and wind speed in high-altitude cold areas, and provided more powerful data support for meteorological research, climate model construction, and prediction of extreme weather events.

[0027] In this way, the data intelligent repair system plays an important role in the application scenario of intelligent repair of historical data of wind direction and wind speed in high-altitude cold regions. Through automated data processing and intelligent algorithm optimization, it successfully repairs problematic meteorological monitoring data and improves the accuracy and completeness of the data. This not only provides a more reliable data foundation for scientific research and business applications in the field of meteorology, but also provides strong support for climate change research and disaster warning in high-altitude cold regions.

[0028] In the embodiment of the present application, step 120 is one of the core invention points of the present application, and its technical solution is detailed as follows: First, the data intelligent repair system will perform in-depth analysis and processing based on the past meteorological monitoring data set. The past meteorological monitoring data set contains a large amount of historical meteorological data, recording key meteorological parameters such as wind speed and wind direction in the alpine region in the past. The data intelligent repair system will conduct a comprehensive review and evaluation of these data sets to identify errors, missing or abnormal data. Secondly, in order to repair these problematic data, the data intelligent repair system will perform multiple rounds of first wind load state vector mining based on feature focusing strategy. Feature focusing strategy is an advanced data mining technology that allows the algorithm to focus on key features in the data, that is, those features that are most important for describing and predicting wind load states. In the process of mining the first wind load state vector, the algorithm analyzes the change patterns of features such as wind speed and wind direction in the past meteorological monitoring data set, and tries to find the intrinsic correlation and law between them. Then, through multiple rounds of mining and iteration, the data intelligent repair system will obtain the first vector reinforcement coefficient. These coefficients reflect the importance and repair potential of each wind load state vector in different rounds of data mining. Specifically, the first vector reinforcement coefficient is a quantitative representation of the contribution and influence of each wind load state vector in the repair process. The higher the coefficient, the more important and effective the vector is in repairing problems in the data set. Finally, based on these first vector reinforcement coefficients, the data intelligent repair system will generate a past monitoring correction data set. The corrected data set is the result of repairing the problems in the original past meteorological monitoring data set. In the corrected data set, erroneous, missing or abnormal data will be accurately repaired or filled, so that the overall quality and accuracy of the data set are significantly improved. At the same time, due to the introduction of the first vector reinforcement coefficient, the corrected data set retains the characteristics of the original data while enhancing the expression and prediction capabilities of key features related to the wind load state. This provides more reliable and powerful data support for subsequent meteorological research, climate model construction, and prediction of extreme weather events.

[0029] In addition, step 130, as another core invention point of the present application, elaborates on the technical scheme of data intelligent repair processing: First, the data intelligent repair system will perform in-depth analysis and comparison based on the target meteorological monitoring data set and the past monitoring correction data set. The target meteorological monitoring data set is a data set containing new spatiotemporal data segments, while the past monitoring correction data set is the result obtained after repairing the past meteorological monitoring data. The data intelligent repair system will compare the two data sets to identify possible problems in the target meteorological monitoring data set, such as errors, missing or abnormal data. Secondly, in order to repair these problems, the data intelligent repair system will perform multiple rounds of second wind load state vector mining based on feature focus strategy. In this process, the algorithm will focus on the wind load state characteristics in the target meteorological monitoring data set, that is, the change pattern of key meteorological parameters such as wind speed and wind direction. Through multiple rounds of mining and iteration, the algorithm will gradually reveal the inherent correlation and law between these data features and the past monitoring correction data set. Then, the data intelligent repair system will obtain the second vector reinforcement coefficients corresponding to each round of second wind load state vector mining. These coefficients reflect the importance and repair potential of each wind load state vector in different rounds of data mining. Similar to the first vector reinforcement coefficient, the second vector reinforcement coefficient is also a quantitative representation of the contribution and influence of each wind load state vector in the repair process. At the same time, in order to further improve the accuracy and reliability of the repair, the data repair algorithm will also calculate the third vector reinforcement coefficient. This coefficient is obtained by adjusting the vector reinforcement coefficient corresponding to the first spatiotemporal data segment in the first vector reinforcement coefficient of the target wheel based on the vector reinforcement coefficient corresponding to the second spatiotemporal data segment in the second vector reinforcement coefficient corresponding to the second wind load state vector mining of the target wheel. The adjustment process takes into account the correlation and difference between the new and old data segments, so that the repair algorithm can better adapt to the update and change of the data set. Next, the data intelligent repair system will mine the target vector reinforcement coefficients (i.e., the second vector reinforcement coefficient or the third vector reinforcement coefficient) corresponding to the first wind load state vector corresponding to the target meteorological monitoring data set and the second wind load state vector of each round, and obtain the wind load state spatiotemporal vectors mined by the second wind load state vector of each round. These vectors are representations of the wind load state data obtained by reinforcement and repair in different rounds of data mining. They not only contain the specific values ​​of wind speed and wind direction, but also metadata related to time and space, making the repaired data more accurate and complete. Finally, based on the wind load state spatiotemporal vectors mined from the last round of the second wind load state vector, the data intelligent repair system generates a first target monitoring correction data set corresponding to the target meteorological monitoring data set. The correction data set is the result of repairing the problems in the original target meteorological monitoring data set. Through this process, the data intelligent repair system successfully solves the problems in the target meteorological monitoring data set and improves the accuracy and completeness of the data.This provides more reliable and powerful data support for meteorological research, climate model construction and prediction of extreme weather events.

[0030] It should be noted that the wind load state space-time vector, as the core concept in the data intelligent repair system, integrates the specific values ​​of key meteorological parameters such as wind speed and wind direction as well as metadata closely related to time and space, providing a more accurate and complete representation for the repaired data.

[0031] First, in the target meteorological monitoring data set, each wind load monitoring information can be represented as a wind load state vector. This vector contains basic numerical features such as wind speed and wind direction. For example, a wind load state vector may be represented as [wind speed: 5m / s, wind direction: 90°], where 5m / s represents the magnitude of the wind speed and 90° represents the angle of the wind direction. This representation method enables the wind load state vector to clearly describe the wind load conditions at a specific time and place.

[0032] Secondly, during the data intelligent repair process, the data intelligent repair system will conduct in-depth analysis and comparison of the target meteorological monitoring data set and the past monitoring and correction data set. Through multiple rounds of second wind load state vector mining based on feature focus strategy, the data intelligent repair system can gradually reveal the inherent correlation and regularity between the wind load state characteristics in the target data set and the past correction data set. In this process, the data intelligent repair system will calculate the second vector reinforcement coefficients corresponding to each round of second wind load state vector mining. These coefficients reflect the importance and repair potential of each wind load state vector in different rounds of data mining.

[0033] Then, in order to further improve the accuracy and reliability of the repair, the data intelligent repair system will calculate the third vector reinforcement coefficient. This coefficient is obtained by adjusting the vector reinforcement coefficient corresponding to the first spatiotemporal data segment in the first vector reinforcement coefficient of the target wheel based on the vector reinforcement coefficient corresponding to the second spatiotemporal data segment in the second vector reinforcement coefficient corresponding to the second wind load state vector mining of the target wheel. In the first round of data mining, the first vector reinforcement coefficient is [0.8, 0.6], where 0.8 represents the reinforcement coefficient of the wind speed vector, and 0.6 represents the reinforcement coefficient of the wind direction vector. In the second round of data mining, for the second spatiotemporal data segment, the second vector reinforcement coefficient calculated by the data intelligent repair system is [0.9, 0.7]. Then, based on these two coefficients, the data intelligent repair system can adjust the third vector reinforcement coefficient to [0.85, 0.65]. The adjusted coefficient not only takes into account the correlation between the new and old data segments, but also retains the importance of the original data features.

[0034] Finally, based on these vector reinforcement coefficients and the first wind load state vector corresponding to the target meteorological monitoring data set, the data intelligent repair system can obtain the wind load state spatiotemporal vectors mined from each round of the second wind load state vector. These vectors not only contain the specific values ​​of wind speed and wind direction, but also incorporate metadata of time and space. For example, a wind load state spatiotemporal vector may be expressed as [timestamp: 2023-04-01 12:00:00, location: alpine region A, wind speed: 5.1m / s, wind direction: 92°, reinforcement coefficient: 0.85, 0.65]. This representation method makes the repaired data more accurate and complete, providing strong data support for subsequent meteorological research, climate model construction, and prediction of extreme weather events.

[0035] It can be seen that the wind load state space-time vector, as an important concept in the data intelligent repair system, provides a more accurate and complete representation of the repaired data by integrating the specific values ​​of key meteorological parameters such as wind speed and wind direction and metadata related to time and space. This not only improves the accuracy and completeness of the data, but also provides a more reliable data foundation for scientific research and business applications in the field of meteorology.

[0036] For example, when constructing a target meteorological monitoring dataset, the first thing to consider is the comprehensiveness and accuracy of the dataset. It should cover a variety of meteorological elements, such as temperature, humidity, air pressure, wind speed, wind direction, and precipitation, and these data need to come from monitoring stations in different geographical locations to ensure the broad representativeness of the data. In addition, the dataset should also contain time series information to analyze the changing trends of meteorological conditions.

[0037] A specific example of a target meteorological monitoring dataset is shown below: The target meteorological monitoring dataset is the "Global Meteorological Monitoring Dataset 2023", which contains real-time data from 1,000 meteorological monitoring stations around the world. These monitoring stations are distributed on seven continents, covering a variety of climate types from tropical rainforests to polar ice sheets. The dataset is updated every hour, recording the temperature (degrees Celsius), relative humidity (percentage), air pressure (hPa), wind speed (meters per second), wind direction (degrees), and precipitation (mm) in the past hour at each monitoring station. In addition, the dataset also includes the geographic location information (longitude and latitude) and altitude (meters) of each monitoring station for spatial analysis.

[0038] In order to further improve the accuracy and reliability of meteorological forecasts, the original monitoring data needs to be corrected to eliminate outliers caused by instrument errors, environmental factors or data transmission problems. This leads to the concept of the first target monitoring correction data set.

[0039] The first target monitoring revised dataset corresponding to the "Global Meteorological Monitoring Dataset 2023" has the following specific examples: The first target monitoring revised dataset is named "Global Meteorological Monitoring Revised Dataset 2023", which is based on the original monitoring dataset and undergoes comprehensive quality control and outlier detection. First, by comparing the data of adjacent monitoring stations, those abnormal records that deviate significantly from the normal range are identified and eliminated. For example, if the temperature reading of a monitoring station suddenly rises by 20 degrees Celsius compared with the surrounding stations, the reading will be considered abnormal and eliminated. Secondly, using time series analysis techniques, long-term trend errors caused by instrument drift or calibration problems are detected and corrected. For example, if it is found that the temperature readings of a monitoring station have been rising steadily over the past few months, and this increase cannot be explained by climate change, then all temperature readings of the station will be adjusted downward accordingly. Finally, for missing data, the revised dataset uses interpolation methods to fill in to ensure the integrity and continuity of the data. The interpolation method takes into account the data of adjacent stations and adjacent time points, as well as seasonal and diurnal variation patterns, to generate estimates that are as accurate as possible.

[0040] Through the above correction process, the "Global Weather Monitoring Revised Dataset 2023" provides a more accurate and reliable data basis for meteorological researchers and prediction models. This will not only help improve the accuracy of weather forecasts, but also provide stronger support for climate change research, extreme weather event analysis, and applications in agriculture, aviation and other fields.

[0041] This application achieves the improvement of the timeliness and accuracy of meteorological data by innovatively acquiring and updating the target meteorological monitoring data set. Specifically, the first spatiotemporal data segment in the past meteorological monitoring data set is updated to the second spatiotemporal data segment containing new monitoring information, ensuring the timeliness of the data set. At the same time, the data repair algorithm is used not only to generate the past monitoring correction data set, but also to obtain the reinforcement coefficient through multiple rounds of wind load state vector mining based on the feature focusing strategy in the process, thereby further improving the data quality. After obtaining the target meteorological monitoring data set, the data repair algorithm is continued to be used for intelligent repair processing. Through multiple rounds of mining and the application of reinforcement coefficients, the first target monitoring correction data set is finally generated, which significantly improves the accuracy and reliability of meteorological data, and provides more solid data support for meteorological forecasting, disaster warning, etc.

[0042] In some optional embodiments, multiple rounds of second wind-load state vector mining based on a feature focusing strategy are performed based on the target meteorological monitoring data set and the past monitoring correction data set to obtain the second vector enhancement coefficient corresponding to each round of second wind-load state vector mining, including: extracting the first wind-load state vector corresponding to the target meteorological monitoring data set, and extracting the second wind-load state vector of the past monitoring correction data set; determining the regional environmental attribute vector corresponding to each round of second wind-load state vector mining based on the first wind-load state vector; determining the second vector enhancement coefficient corresponding to each round of second wind-load state vector mining based on the index representation vector and the regional environmental attribute vector corresponding to each round of second wind-load state vector mining; wherein, the index representation vector corresponding to the first round of second wind-load state vector mining is determined based on the second wind-load state vector; the index representation vector corresponding to the second wind-load state vector mining other than the first round is determined based on the wind-load state spatiotemporal vector of the previous round of second wind-load state vector mining.

[0043] In the above embodiment, the first wind load state vector refers to a feature vector extracted from the target meteorological monitoring data set, representing the wind load condition within a specific time and space range. This vector contains information on key meteorological elements such as wind speed, wind direction, and wind force, and is an important basic data in meteorological monitoring and forecasting.

[0044] The second wind load state vector is extracted from the past monitoring and correction data set, and also represents the characteristic vector of the wind load condition. Compared with the first wind load state vector, the second wind load state vector contains the wind load information of the past period, and its accuracy is improved after data correction.

[0045] The regional environmental attribute vector is a feature vector related to a specific region, which contains various environmental factors that affect the wind load conditions, such as terrain, vegetation cover, building distribution, etc. This vector plays an important role in wind load state vector mining because it can help the model better understand the wind load characteristics of different regions.

[0046] The index representation vector is a feature vector used to represent and index the wind load state vector. In each round of second wind load state vector mining, the index representation vector is updated according to the current wind load state spatiotemporal vector to ensure the continuity and accuracy of the mining process.

[0047] Based on the above content, firstly, the first wind load state vector is extracted from the target meteorological monitoring data set. This vector contains the latest wind load monitoring information and is the basis for subsequent mining work. At the same time, the second wind load state vector is extracted from the past monitoring correction data set. This vector contains the corrected past wind load information, providing rich historical data support for mining work.

[0048] Next, the regional environmental attribute vectors corresponding to the second wind load state vectors in each round of mining are determined based on the first wind load state vector. This step is the key to mining because it combines the latest wind load conditions with regional environmental factors, providing a more accurate and specific direction for subsequent mining work. Specifically, the regional environmental attributes that should be paid attention to in each round of mining can be determined by analyzing the wind speed, wind direction and other elements in the first wind load state vector, combined with environmental factors such as the terrain and vegetation of the region.

[0049] Then, based on the index representation vector and the regional environmental attribute vector corresponding to each round of the second wind load state vector mining, the second vector reinforcement coefficient corresponding to each round of the second wind load state vector mining is determined. This step is the core of the mining work. It evaluates the accuracy and importance of the wind load state vector obtained in each round of mining by calculating the similarity or correlation between the index representation vector and the regional environmental attribute vector. Specifically, the similarity or correlation between the two vectors can be calculated through a machine learning algorithm or a statistical model to obtain a reinforcement coefficient. This coefficient reflects the importance and accuracy of the wind load state vector obtained by the current round of mining in the overall data set, and is an important basis for subsequent data repair and prediction work.

[0050] It is worth noting that in the first round of mining the second wind load state vector, the index representation vector is determined based on the second wind load state vector. This is because in this round of mining, the data intelligent repair system has not obtained any new wind load state spatiotemporal vectors, so it can only determine the index representation vector based on past wind load information. Starting from the second round, the index representation vector corresponding to each round of mining is determined based on the wind load state spatiotemporal vector obtained in the previous round of mining. This is because as the mining work deepens, the data intelligent repair system will obtain more and more new wind load information, which has a more direct and significant impact on determining the index representation vector and evaluating the accuracy and importance of the wind load state vector.

[0051] Through such multiple rounds of mining and calculation of reinforcement coefficients, the data intelligent repair system can obtain a series of optimized and enhanced second wind load state vectors. These vectors not only contain wind load information from past periods, but also have been adjusted for regional environmental attributes and weighted by reinforcement coefficients, so they have higher accuracy and reliability. In subsequent data repair and prediction work, the data intelligent repair system can use these optimized wind load state vectors to generate more accurate and reliable meteorological monitoring data sets and prediction models.

[0052] In this way, through multiple rounds of second wind load state vector mining and reinforcement coefficient calculation based on feature focusing strategy, deep mining and optimization processing of meteorological monitoring data are achieved. It not only improves the accuracy and reliability of data, but also provides a more solid data foundation and technical support for subsequent meteorological forecasts and disaster warnings.

[0053] In some examples, if the target confidence vector corresponding to a round of second wind-load state vector mining is the third vector enhancement coefficient, then the target confidence vectors corresponding to subsequent rounds of second wind-load state vector mining are all the third vector enhancement coefficient.

[0054] Based on this example, an innovative strategy is adopted to determine the target confidence vector in the second wind load state vector mining process, that is, if the target confidence vector corresponding to a round of second wind load state vector mining is the third vector reinforcement coefficient, then the third vector reinforcement coefficient will be used as the target confidence vector in each subsequent round of second wind load state vector mining. The implementation of this strategy aims to improve the accuracy and efficiency of the second wind load state vector mining by dynamically adjusting the target confidence vector.

[0055] First, in the process of mining the second wind load state vector, the target confidence vector is a key parameter, which is used to measure and evaluate the accuracy and reliability of the wind load state vector obtained by the current round of mining. The third vector reinforcement coefficient is a new reinforcement coefficient obtained by adjusting and optimizing the second vector reinforcement coefficient corresponding to the second wind load state vector mining of the target wheel under specific conditions. This new reinforcement coefficient not only contains the information of the original second vector reinforcement coefficient, but also incorporates the new wind load state information related to the target spatiotemporal data segment, so it has higher accuracy and applicability.

[0056] Furthermore, based on the following considerations, the third vector enhancement coefficient is selected as the target confidence vector for mining the second wind load state vector in subsequent rounds:

[0057] 1. The third vector reinforcement coefficient contains richer and more accurate wind load state information. By adjusting and optimizing the second vector reinforcement coefficient corresponding to the second wind load state vector mining of the target wheel, the data intelligent repair system can integrate the new wind load state information related to the target spatiotemporal data segment into the reinforcement coefficient, thereby obtaining a more accurate and comprehensive representation of the wind load state. Such a reinforcement coefficient can better reflect the current wind load conditions and provide more accurate direction and guidance for subsequent mining work;

[0058] 2. Using the third vector reinforcement coefficient as the target confidence vector can improve the efficiency and accuracy of mining. In the process of mining the second wind load state vector, the data intelligent repair system needs to continuously evaluate and adjust the mining results to ensure the accuracy and effectiveness of mining. Using the third vector reinforcement coefficient as the target confidence vector can provide a more accurate and reliable evaluation standard for the data intelligent repair system, so that the data intelligent repair system can find the accurate mining direction more quickly and improve the efficiency of mining;

[0059] 3. Using the third vector reinforcement coefficient as the target confidence vector can also enhance the stability and continuity of the mining work. During the mining of the second wind load state vector, the mining results may fluctuate and be unstable due to various factors. Using the third vector reinforcement coefficient as the target confidence vector can provide a stable reference benchmark for the data intelligent repair system, so that the mining work can maintain continuity and stability between different rounds, thereby improving the overall mining effect.

[0060] In the specific implementation process, the data intelligent repair system determines that the target confidence vector corresponding to a round of second wind load state vector mining is the third vector reinforcement coefficient mainly through the following steps: First, the data intelligent repair system needs to analyze and evaluate the second vector reinforcement coefficient corresponding to the second wind load state vector mining of the target round. This step is the basis and prerequisite for determining the third vector reinforcement coefficient. The data intelligent repair system needs to conduct a comprehensive analysis and evaluation of the second vector reinforcement coefficient to understand the wind load state information it contains and the potential problems and deficiencies. Secondly, based on the results of the analysis and evaluation, the data intelligent repair system needs to adjust and optimize the second vector reinforcement coefficient. This step is the key and core of determining the third vector reinforcement coefficient. The data intelligent repair system needs to make targeted adjustments and optimizations to the second vector reinforcement coefficient according to the wind load state information of the target spatiotemporal data segment and the actual needs of the mining work, so as to obtain a more accurate and comprehensive representation of the wind load state. Finally, the data intelligent repair system needs to use the adjusted and optimized third vector reinforcement coefficient as the target confidence vector for the subsequent rounds of second wind load state vector mining. This step is the last step of the implementation strategy and the most critical step. The data intelligent repair system needs to apply the third vector reinforcement coefficient to subsequent mining work as an important basis for evaluating and guiding mining work, thereby improving the accuracy and efficiency of mining work. Through such strategies and implementation steps, the data intelligent repair system can effectively determine and use the third vector reinforcement coefficient as the target confidence vector for subsequent rounds of second wind load state vector mining. The implementation of this strategy can not only improve the accuracy and efficiency of mining work, but also enhance the stability and continuity of mining work, providing more accurate and reliable data support for meteorological monitoring and forecasting work.

[0061] In a specific application scenario, the above embodiment can obtain more accurate and comprehensive wind load state information through deep mining and optimization of the second wind load state vector. This information has important reference value for meteorological forecasting, disaster warning and related decision-making. At the same time, by using the third vector reinforcement coefficient as the target confidence vector, the data intelligent repair system can further improve the accuracy and reliability of the meteorological monitoring data set, providing a more solid data foundation and technical support for meteorological scientific research and application.

[0062] In addition, through deep mining and optimization of the second wind load state vector, the data intelligent repair system can continuously explore and discover new meteorological characteristics and laws, and provide new ideas and methods for the improvement and optimization of meteorological models. At the same time, by using the third vector reinforcement coefficient as the target confidence vector, the data intelligent repair system can further improve the prediction accuracy and stability of the meteorological model, and provide more reliable technical support for meteorological forecasting and disaster warning.

[0063] In another optional embodiment, the first vector reinforcement coefficient includes a disturbance assessment feature corresponding to each first wind load monitoring information in the past meteorological monitoring data set; the second vector reinforcement coefficient includes a disturbance assessment feature corresponding to each second wind load monitoring information in the target meteorological monitoring data set. If the number of second wind load monitoring information included in the second spatiotemporal data segment is equal to the number of first wind load monitoring information included in the first spatiotemporal data segment, the vector reinforcement coefficient corresponding to the second spatiotemporal data segment in the second vector reinforcement coefficient corresponding to the second wind load state vector of the target wheel is mined, and the vector reinforcement coefficient corresponding to the first spatiotemporal data segment in the first vector reinforcement coefficient of the target wheel is adjusted to obtain the third vector reinforcement coefficient, including: determining the first disturbance assessment feature in the first vector reinforcement coefficient corresponding to at least one first wind load monitoring information of the first spatiotemporal data segment; determining the second disturbance assessment feature in the second vector reinforcement coefficient corresponding to at least one second wind load monitoring information of the second spatiotemporal data segment; and updating at least one first disturbance assessment feature in the first vector reinforcement coefficient by using the at least one second disturbance assessment feature to obtain the third vector reinforcement coefficient.

[0064] It can be understood that this embodiment involves how to adjust and optimize the vector enhancement coefficient based on specific conditions to improve the analysis and processing effects of the meteorological monitoring data set, and particularly focuses on the adjustment process of the first vector enhancement coefficient and the second vector enhancement coefficient, and how to obtain the third vector enhancement coefficient through this adjustment.

[0065] First, both the first vector reinforcement coefficient and the second vector reinforcement coefficient contain the disturbance assessment features corresponding to the meteorological monitoring information. These features are a quantitative representation of the meteorological monitoring data, which can reflect the changes in meteorological conditions and disturbances. Specifically, the first vector reinforcement coefficient covers the disturbance assessment features corresponding to each first wind load monitoring information in the past meteorological monitoring data set, while the second vector reinforcement coefficient contains the disturbance assessment features corresponding to each second wind load monitoring information in the target meteorological monitoring data set.

[0066] On this basis, if the number of second wind load monitoring information contained in the second spatiotemporal data segment is equal to the number of first wind load monitoring information contained in the first spatiotemporal data segment, the data intelligent repair system has the premise to adjust the vector reinforcement coefficient. This means that the two spatiotemporal data segments match in the number of wind load monitoring information, so the data intelligent repair system can compare and adjust them one by one.

[0067] Next, the data intelligent repair system specifically explains how to mine the corresponding second vector reinforcement coefficient based on the second wind load state vector of the target wheel, and adjust the first vector reinforcement coefficient of the target wheel to obtain the third vector reinforcement coefficient.

[0068] First, the data intelligent repair system needs to determine the first disturbance assessment features in the first vector enhancement coefficient that correspond to at least one first wind load monitoring information of the first spatiotemporal data segment. These features reflect the changes and disturbances in the wind load state in the past meteorological monitoring data set and are the basis for the data intelligent repair system to make adjustments.

[0069] Secondly, the data intelligent repair system needs to determine the second disturbance assessment features in the second vector enhancement coefficient that correspond to at least one second wind load monitoring information of the second spatiotemporal data segment. These features represent the wind load state changes and disturbance conditions in the target meteorological monitoring data set, and are the basis for the data intelligent repair system to update the first vector enhancement coefficient.

[0070] Finally, the data intelligent repair system updates at least one first disturbance assessment feature in the first vector reinforcement coefficient by using at least one second disturbance assessment feature, thereby obtaining a third vector reinforcement coefficient. This updating process is based on the correspondence between the wind load monitoring information of the second spatiotemporal data segment and the wind load monitoring information of the first spatiotemporal data segment. The data intelligent repair system adjusts the first disturbance assessment feature to make it closer to or consistent with the second disturbance assessment feature, thereby obtaining a more accurate and reliable third vector reinforcement coefficient.

[0071] Specifically, this updating process can involve fine-tuning the value of the first disturbance assessment feature, or redistributing the first disturbance assessment feature according to the distribution of the second disturbance assessment feature. Regardless of the method used, the goal of the data intelligent repair system is to make the third vector reinforcement coefficient more accurately reflect the actual wind load state and disturbance of the meteorological monitoring data set.

[0072] To illustrate this process more specifically, in the first spatiotemporal data segment, the data intelligent repair system has a disturbance assessment feature corresponding to the first wind load monitoring information of [0.5, 0.3, 0.2], and in the second spatiotemporal data segment, the corresponding disturbance assessment feature of the second wind load monitoring information is [0.6, 0.4, 0.1]. In this case, the data intelligent repair system can adjust the first disturbance assessment feature to make it closer to the second disturbance assessment feature, thereby obtaining the disturbance assessment feature of the wind load monitoring information corresponding to the third vector enhancement coefficient. For example, the data intelligent repair system can adjust the first disturbance assessment feature to [0.55, 0.35, 0.15], so that it is closer to the second disturbance assessment feature.

[0073] Through such an adjustment process, the data intelligent repair system can obtain a more accurate and reliable third vector reinforcement coefficient, which can better reflect the actual wind load state and disturbance of the meteorological monitoring data set. The application of this coefficient will help improve the accuracy and efficiency of meteorological monitoring and prediction, and provide a more solid data foundation and technical support for meteorological scientific research and application.

[0074] In some other possible embodiments, the first vector enhancement coefficient includes the disturbance assessment feature corresponding to each first wind load monitoring information in the past meteorological monitoring data set; the second vector enhancement coefficient includes the disturbance assessment feature corresponding to each second wind load monitoring information in the target meteorological monitoring data set. If the number of second wind load monitoring information included in the second space-time data segment is not equal to the number of first wind load monitoring information included in the first space-time data segment, the vector reinforcement coefficient corresponding to the second space-time data segment in the second vector reinforcement coefficient corresponding to the second wind load state vector of the target wheel is mined, and the vector reinforcement coefficient corresponding to the first space-time data segment in the first vector reinforcement coefficient of the target wheel is adjusted to obtain the third vector reinforcement coefficient, including: determining the first disturbance assessment feature in the first vector reinforcement coefficient that corresponds to at least one first wind load monitoring information of the first space-time data segment; determining the second disturbance assessment feature in the second vector reinforcement coefficient that corresponds to at least one second wind load monitoring information of the second space-time data segment; determining the first disturbance influence degree of each second disturbance assessment feature in the second space-time data segment according to the first number of the first wind load monitoring information in the first space-time data segment and the second number of the second wind load monitoring information in the second space-time data segment; using the first disturbance influence degree of each second disturbance assessment feature, and using the at least one second disturbance assessment feature to update at least one first disturbance assessment feature in the first vector reinforcement coefficient to obtain the third vector reinforcement coefficient.

[0075] It is worth mentioning that this embodiment involves a more complex but equally important scenario, that is, when the number of second wind load monitoring information included in the second spatiotemporal data segment is not equal to the number of first wind load monitoring information included in the first spatiotemporal data segment, how to adjust the first vector enhancement coefficient to obtain the third vector enhancement coefficient. The handling of this problem is of vital importance to improving the accuracy of meteorological monitoring and forecasting, because it allows the data intelligent repair system to effectively compare and adjust between different numbers of wind load monitoring information.

[0076] When faced with the situation that the amount of wind load monitoring information in the second spatiotemporal data segment is not equal to that in the first spatiotemporal data segment, the data intelligent repair system cannot directly perform a one-to-one comparison and adjustment. Therefore, the data intelligent repair system needs to adopt a more complex method to deal with this problem.

[0077] First, the data intelligent repair system needs to determine the first disturbance assessment features in the first vector enhancement coefficient that correspond to at least one first wind load monitoring information of the first spatiotemporal data segment. These features reflect the changes and disturbances in the wind load state in the past meteorological monitoring data set and are the basis for the data intelligent repair system to make adjustments.

[0078] Secondly, the data intelligent repair system needs to determine the second disturbance assessment features in the second vector enhancement coefficient that correspond to at least one second wind load monitoring information of the second spatiotemporal data segment. These features represent the wind load state changes and disturbance conditions in the target meteorological monitoring data set, and are the basis for the data intelligent repair system to update the first vector enhancement coefficient.

[0079] However, since the number of wind load monitoring information in the two spatiotemporal data segments is not equal, the data intelligent repair system cannot directly correspond and update the features. Therefore, the data intelligent repair system needs to determine the first disturbance impact of each second disturbance assessment feature in the second spatiotemporal data segment based on the first number of the first wind load monitoring information in the first spatiotemporal data segment and the second number of the second wind load monitoring information in the second spatiotemporal data segment. This step is critical because it allows the data intelligent repair system to evaluate the relative importance of each second disturbance assessment feature in the overall wind load state.

[0080] Specifically, the data intelligent repair system can use a weighted average method to calculate the first disturbance influence of each second disturbance assessment feature. The data intelligent repair system can multiply the value of each second disturbance assessment feature by the number of second wind load monitoring information corresponding to it, then add all the products, and finally divide by the total number of all second wind load monitoring information in the second spatiotemporal data segment. In this way, the data intelligent repair system can obtain a weighted average second disturbance assessment feature value, which represents the relative contribution of each feature in the overall wind load state.

[0081] After determining the first disturbance impact of each second disturbance assessment feature, the data intelligent repair system can use these features to update at least one first disturbance assessment feature in the first vector reinforcement coefficient. This update process may involve fine-tuning the value of the first disturbance assessment feature, or reallocating the first disturbance assessment feature according to the distribution of the second disturbance assessment feature. Regardless of the method used, the goal of the data intelligent repair system is to enable the third vector reinforcement coefficient to more accurately reflect the actual wind load state and disturbance of the meteorological monitoring data set.

[0082] To illustrate this process more specifically, in the first spatiotemporal data segment, the data intelligent repair system has three first wind load monitoring information corresponding to the disturbance assessment features, which are [0.5, 0.3, 0.2], and in the second spatiotemporal data segment, the data intelligent repair system has four second wind load monitoring information corresponding to the disturbance assessment features, which are [0.6, 0.4, 0.1, 0.3]. In this case, the data intelligent repair system needs to first calculate the first disturbance impact of each second disturbance assessment feature. The data intelligent repair system uses a weighted average method, so the first disturbance impact of each second disturbance assessment feature may be 0.24, 0.16, 0.04 and 0.12 respectively.

[0083] Then, the data intelligent repair system can use these first disturbance influences to update the first disturbance assessment feature in the first vector reinforcement coefficient. For example, the data intelligent repair system can adjust the first disturbance assessment feature [0.5, 0.3, 0.2] to [0.5+0.24 adjustment coefficient, 0.3+0.16 adjustment coefficient, 0.2+(0.04+0.12) / 2*adjustment coefficient], where the adjustment coefficient is a constant determined according to actual needs and is used to control the update amplitude. Through such adjustments, the data intelligent repair system can obtain a more accurate and reliable third vector reinforcement coefficient.

[0084] It can be seen that when the number of wind load monitoring information in the second spatiotemporal data segment is different from that in the first spatiotemporal data segment, the data intelligent repair system determines the first disturbance assessment feature and the second disturbance assessment feature, calculates the first disturbance influence degree of each second disturbance assessment feature, and uses these features to update the first disturbance assessment feature in the first vector reinforcement coefficient, thereby obtaining the third vector reinforcement coefficient. This process allows the data intelligent repair system to effectively compare and adjust between different numbers of wind load monitoring information, thereby improving the accuracy and efficiency of meteorological monitoring and prediction.

[0085] In the next step, if the number of second wind load monitoring information included in the second space-time data segment is greater than the number of first wind load monitoring information included in the first space-time data segment, the target vector reinforcement coefficient is any one of the second vector reinforcement coefficient, the third vector reinforcement coefficient or the fourth vector reinforcement coefficient; the target confidence vector corresponding to at least one round of second wind load state vector mining is the fourth vector reinforcement coefficient; the fourth vector reinforcement coefficient is obtained through the following steps: adding at least one third disturbance assessment feature to the third vector reinforcement coefficient to obtain the fourth vector reinforcement coefficient; the at least one third disturbance assessment feature includes the at least one second disturbance assessment feature.

[0086] It is understandable that when the number of second wind load monitoring information included in the second spatiotemporal data segment is greater than the number of first wind load monitoring information included in the first spatiotemporal data segment, how to determine the target vector reinforcement coefficient and then obtain the fourth vector reinforcement coefficient. The handling of this problem is particularly important for improving the accuracy of meteorological monitoring and prediction, because it enables the data intelligent repair system to establish effective connections and comparisons between different numbers of wind load monitoring information.

[0087] First, the first vector reinforcement coefficient, the second vector reinforcement coefficient, the third vector reinforcement coefficient and the fourth vector reinforcement coefficient all contain disturbance assessment features corresponding to meteorological monitoring information. These features are a quantitative representation of meteorological monitoring data, which can reflect changes in meteorological conditions and disturbances. Specifically, the first vector reinforcement coefficient covers the disturbance assessment features corresponding to each first wind load monitoring information in the past meteorological monitoring data set, while the second vector reinforcement coefficient contains the disturbance assessment features corresponding to each second wind load monitoring information in the target meteorological monitoring data set. The third vector reinforcement coefficient is obtained after adjusting the first vector reinforcement coefficient and the second vector reinforcement coefficient, and it integrates the information of the past and target meteorological monitoring data sets. The fourth vector reinforcement coefficient is obtained by further adding at least one third disturbance assessment feature on the basis of the third vector reinforcement coefficient, and it contains more meteorological state changes and disturbance information.

[0088] When the amount of wind load monitoring information in the second spatiotemporal data segment is different from that in the first spatiotemporal data segment, and the amount of information in the second spatiotemporal data segment is greater than that in the first spatiotemporal data segment, the data intelligent repair system cannot directly perform a one-to-one comparison and adjustment. Therefore, a more complex method is needed to deal with this problem.

[0089] At this time, the target vector reinforcement coefficient can be any of the second vector reinforcement coefficient, the third vector reinforcement coefficient, or the fourth vector reinforcement coefficient. This means that the data intelligent repair system can select the appropriate vector reinforcement coefficient as the target according to different needs and scenarios. However, at least one round of the second wind load state vector mining corresponds to the fourth vector reinforcement coefficient. This is because the fourth vector reinforcement coefficient adds at least one third disturbance assessment feature on the basis of the third vector reinforcement coefficient, which contains more meteorological state changes and disturbance information, and can therefore provide more accurate and reliable meteorological monitoring and forecasting results.

[0090] Based on this, a series of steps to obtain the fourth vector reinforcement coefficient are as follows:

[0091] First, the data intelligent repair system needs to obtain the third vector reinforcement coefficient. This process involves adjusting the first vector reinforcement coefficient and the second vector reinforcement coefficient. Specifically, the data intelligent repair system needs to determine the first disturbance assessment feature in the first vector reinforcement coefficient that corresponds to at least one first wind load monitoring information of the first space-time data segment, and the second disturbance assessment feature in the second vector reinforcement coefficient that corresponds to at least one second wind load monitoring information of the second space-time data segment. Then, the data intelligent repair system uses at least one second disturbance assessment feature to update at least one first disturbance assessment feature in the first vector reinforcement coefficient to obtain the third vector reinforcement coefficient. This process allows the data intelligent repair system to effectively compare and adjust between different numbers of wind load monitoring information, so that the third vector reinforcement coefficient can more accurately reflect the actual wind load status and disturbance conditions of the meteorological monitoring data set;

[0092] Next, the data intelligent repair system adds at least one third disturbance assessment feature on the basis of the third vector reinforcement coefficient to obtain a fourth vector reinforcement coefficient. This step is critical because it enables the data intelligent repair system to incorporate more meteorological state changes and disturbance information into the fourth vector reinforcement coefficient. Specifically, the data intelligent repair system can select at least one second disturbance assessment feature as the third disturbance assessment feature and add it to the third vector reinforcement coefficient. In this way, the fourth vector reinforcement coefficient contains all the information of the third vector reinforcement coefficient, as well as the information of at least one additional second disturbance assessment feature.

[0093] By adding the third disturbance assessment feature, the fourth vector enhancement coefficient can more comprehensively reflect the wind load state and disturbance of the meteorological monitoring data set. It integrates the information of the past and target meteorological monitoring data sets and contains more meteorological state changes and disturbance characteristics. This makes the fourth vector enhancement coefficient more accurate and reliable in meteorological monitoring and forecasting.

[0094] For example, in the first spatiotemporal data segment, the data intelligent repair system has three first wind load monitoring information corresponding to the disturbance assessment features [0.5, 0.3, 0.2], and in the second spatiotemporal data segment, the data intelligent repair system has five second wind load monitoring information corresponding to the disturbance assessment features [0.6, 0.4, 0.1, 0.3, 0.2]. In this case, the data intelligent repair system needs to first obtain the third vector reinforcement coefficient. Through the adjustment process, the data intelligent repair system obtains the third vector reinforcement coefficient of [0.55, 0.35, 0.25].

[0095] Then, the data intelligent repair system adds at least one third disturbance assessment feature on the basis of the third vector reinforcement coefficient to obtain a fourth vector reinforcement coefficient. The data intelligent repair system selects 0.3 in the second disturbance assessment feature as the added third disturbance assessment feature, so the fourth vector reinforcement coefficient is [0.55, 0.35, 0.25, 0.3]. In this way, the fourth vector reinforcement coefficient contains all the information of the third vector reinforcement coefficient and the information of an additional second disturbance assessment feature.

[0096] Therefore, when the number of second wind load monitoring information contained in the second spatiotemporal data segment is greater than the number of first wind load monitoring information contained in the first spatiotemporal data segment, the data intelligent repair system determines that the target vector reinforcement coefficient is any one of the second vector reinforcement coefficient, the third vector reinforcement coefficient or the fourth vector reinforcement coefficient, and at least one round of second wind load state vector mining corresponds to the target confidence vector as the fourth vector reinforcement coefficient. The data intelligent repair system obtains the fourth vector reinforcement coefficient through a series of steps, including first obtaining the third vector reinforcement coefficient, and then adding at least one third disturbance assessment feature on the basis of the third vector reinforcement coefficient. This process allows the data intelligent repair system to establish effective connections and comparisons between different numbers of wind load monitoring information, thereby improving the accuracy and efficiency of meteorological monitoring and forecasting.

[0097] Under a preferred design concept, at least one third disturbance assessment feature is added to the third vector enhancement coefficient to obtain the fourth vector enhancement coefficient, including: using the first disturbance influence of each second disturbance assessment feature to determine the second disturbance influence of each third disturbance assessment feature; using the second disturbance influence of each third disturbance assessment feature to add at least one third disturbance assessment feature to the third vector enhancement coefficient to obtain the fourth vector enhancement coefficient.

[0098] Based on this design idea, when it is necessary to add at least one third disturbance assessment feature to the third vector enhancement coefficient to obtain the fourth vector enhancement coefficient, the whole process needs to be carried out in a delicate and orderly manner. This is not only to ensure that the newly added features can effectively improve the expression ability of the vector, but also to maintain the accuracy and stability of the vector enhancement coefficient in meteorological monitoring and forecasting applications.

[0099] First, it is necessary to clarify the first disturbance impact of the second disturbance assessment feature. This step is the basis for all subsequent operations, because the data intelligent repair system will determine the second disturbance impact of the third disturbance assessment feature based on these impacts. Specifically, the first disturbance impact reflects the size and importance of the role played by each second disturbance assessment feature in the change and disturbance of the meteorological state. By deeply analyzing the wind load monitoring information in the second spatiotemporal data segment, the data intelligent repair system can quantify the impact of each second disturbance assessment feature on the meteorological state, thereby obtaining its first disturbance impact.

[0100] With the first disturbance impact of the second disturbance assessment feature, the data intelligent repair system can further determine the second disturbance impact of each third disturbance assessment feature. This step is critical because it directly determines which third disturbance assessment features will be added to the third vector reinforcement coefficient, as well as the weights of these features in the fourth vector reinforcement coefficient. In order to determine the second disturbance impact, the data intelligent repair system needs to consider multiple factors, including the correlation between the second disturbance assessment feature and changes in meteorological conditions, the interaction between features, and the distribution of features in time and space. Through the comprehensive use of data mining, machine learning, statistical analysis and other methods, the data intelligent repair system can accurately evaluate the second disturbance impact of each third disturbance assessment feature, providing strong support for subsequent additional operations.

[0101] After determining the second disturbance impact of the third disturbance assessment feature, the data intelligent repair system can add at least one third disturbance assessment feature to the third vector reinforcement coefficient to obtain the fourth vector reinforcement coefficient. This step is the final stage of the entire process and is also a key link in achieving improved meteorological monitoring and forecasting capabilities. During the addition process, the data intelligent repair system needs to select the most appropriate third disturbance assessment feature for addition based on the size and importance of the second disturbance impact. At the same time, the data intelligent repair system also needs to consider the stability and generalization ability of the fourth vector reinforcement coefficient after the addition to ensure that it can maintain accurate and reliable prediction results under different meteorological conditions and time and space backgrounds.

[0102] To achieve this goal, the data intelligent repair system can adopt a variety of strategies and methods. For example, the data intelligent repair system can use feature selection algorithms to screen out the most representative and predictive third disturbance assessment features; the data intelligent repair system can also use regularization technology to prevent overfitting and improve the generalization ability of the model; in addition, the data intelligent repair system can also use methods such as cross-validation to evaluate the effects of different augmentation schemes and select the optimal augmentation strategy.

[0103] Therefore, the process of adding at least one third disturbance assessment feature to the third vector reinforcement coefficient to obtain the fourth vector reinforcement coefficient is a complex and orderly task. It requires the data intelligent repair system to deeply analyze the first disturbance impact of the second disturbance assessment feature, accurately evaluate the second disturbance impact of the third disturbance assessment feature, and select the most appropriate feature for addition on this basis. Through this series of operations and strategy selection, the data intelligent repair system can obtain a more expressive and predictive fourth vector reinforcement coefficient, providing more accurate and reliable technical support for meteorological monitoring and forecasting.

[0104] For example, accurate meteorological monitoring and forecasting capabilities are crucial in the prediction and response to extreme weather events. By applying this technical solution, the data intelligent repair system can more accurately capture and quantify the changes and disturbance characteristics of meteorological conditions, and improve the accuracy and timeliness of extreme weather event predictions. This will provide decision-makers with more accurate and timely meteorological information support, help formulate more scientific and effective response strategies, and reduce the impact of extreme weather events on society and the economy.

[0105] In addition, this technical solution will also play an important role in the field of climate change research and meteorological science research. By deeply analyzing and mining the disturbance assessment characteristics in the meteorological monitoring data set, the data intelligent repair system can have a deeper understanding of the laws and mechanisms of climate change, and provide a more solid data foundation and technical support for climate change research and meteorological science research. At the same time, this technical solution can also be combined with other meteorological monitoring and prediction technologies to form a more complete and accurate meteorological monitoring and prediction system, providing a broader space and possibility for the development and application of meteorological science.

[0106] To sum up, by adding at least one third disturbance assessment feature to the third vector enhancement coefficient to obtain the fourth vector enhancement coefficient, the data intelligent repair system will be able to effectively improve the ability and accuracy of meteorological monitoring and forecasting.

[0107] At the same time, in practical applications, the data intelligent repair system needs to flexibly select and adjust the third disturbance assessment feature according to specific meteorological conditions and monitoring requirements to ensure the optimal performance and prediction effect of the fourth vector reinforcement coefficient. In addition, the data intelligent repair system also needs to continuously update and optimize the meteorological monitoring data set and disturbance assessment feature library to adapt to the ever-changing climate environment and meteorological conditions.

[0108] In summary, in the field of meteorological monitoring and forecasting, by adding at least one third disturbance assessment feature to the third vector enhancement coefficient to obtain the fourth vector enhancement coefficient, the data intelligent repair system will provide more solid and powerful support for achieving this goal.

[0109] In other examples, if the target confidence vector of a round of second wind-load state vector mining is the third vector reinforcement coefficient, then the target confidence vectors corresponding to subsequent rounds of second wind-load state vector mining are the third vector reinforcement coefficient or the fourth vector reinforcement coefficient; if the target confidence vector corresponding to a round of second wind-load state vector mining is the fourth vector reinforcement coefficient, then the target confidence vectors corresponding to subsequent rounds of second wind-load state vector mining are all the fourth vector reinforcement coefficient.

[0110] Based on this example, regarding the selection strategy of the target confidence vector for mining the second wind load state vector, the data intelligent repair system adopts a dynamic adjustment method to ensure that the accuracy and reliability of meteorological monitoring and forecasting can be continuously and stably improved during continuous rounds of mining.

[0111] First, if the target confidence vector of a round of second wind load state vector mining is the third vector reinforcement coefficient, then this means that in the current meteorological monitoring and forecasting task, the third vector reinforcement coefficient is considered to be a relatively reliable and effective target. Therefore, in subsequent rounds of second wind load state vector mining, the data intelligent repair system will continue to select the third vector reinforcement coefficient as the target confidence vector, or select the fourth vector reinforcement coefficient that is further equipped with at least one third disturbance assessment feature on its basis. Such a selection strategy is intended to maintain the stability and continuity of meteorological monitoring and forecasting, while allowing the accuracy of the forecast to be further improved by introducing more disturbance assessment features when necessary.

[0112] Specifically, when the data intelligent repair system selects the third vector reinforcement coefficient as the target confidence vector for subsequent rounds, the data intelligent repair system is actually conducting further mining and optimization on the basis of keeping the core features of the current meteorological monitoring and forecasting model unchanged. This helps to ensure the stability of the model and the consistency of the forecast. When the data intelligent repair system selects the fourth vector reinforcement coefficient, the data intelligent repair system is actually introducing more meteorological state changes and disturbance information in the hope of capturing more meteorological features, thereby improving the accuracy and comprehensiveness of the forecast.

[0113] Secondly, if the target confidence vector corresponding to a round of second wind load state vector mining is the fourth vector reinforcement coefficient, then this means that in the current meteorological monitoring and forecasting task, the fourth vector reinforcement coefficient is considered to be a more accurate and effective target. Therefore, in subsequent rounds of second wind load state vector mining, the data intelligent repair system will insist on selecting the fourth vector reinforcement coefficient as the target confidence vector. This selection strategy is intended to ensure that the accuracy of meteorological monitoring and forecasting can continue to improve and remain at a high level.

[0114] Specifically, when the data intelligent repair system selects the fourth vector reinforcement coefficient as the target confidence vector for subsequent rounds, the data intelligent repair system is actually confirming that the current meteorological monitoring and forecasting model has been significantly improved by introducing more disturbance assessment features. Therefore, the data intelligent repair system will continue to adhere to this choice to ensure that the model can continue to capture more meteorological features and continuously improve the accuracy and comprehensiveness of the forecast. At the same time, such a choice also helps to maintain the stability and consistency of the model and avoid forecast fluctuations and uncertainties caused by frequent changes in the target confidence vector.

[0115] It can be seen that this dynamically adjusted target confidence vector selection strategy allows the data intelligent repair system to flexibly select the most suitable target confidence vector in the process of continuous multiple rounds of second wind load state vector mining according to the current meteorological monitoring and prediction task requirements and model performance. Such a strategy not only helps to maintain the stability and continuity of meteorological monitoring and prediction, but also can continuously improve the accuracy and comprehensiveness of the prediction.

[0116] In addition, this dynamic adjustment strategy will also play an important role. By deeply analyzing and mining the disturbance assessment features in the meteorological monitoring data set and dynamically adjusting the target confidence vector, the data intelligent repair system can have a deeper understanding of the laws and mechanisms of climate change, and provide a more solid data foundation and technical support for climate change research and meteorological science research. At the same time, this strategy can also be combined with other meteorological monitoring and forecasting technologies to form a more complete and accurate meteorological monitoring and forecasting system, providing a broader space and possibility for the development and application of meteorological science.

[0117] Of course, this dynamic adjustment strategy still needs to be continuously explored and improved in its implementation and application. For example, the data intelligent repair system needs to further study how to more accurately evaluate the second disturbance impact of the third disturbance assessment feature to determine its weight and importance in the fourth vector reinforcement coefficient. At the same time, it is also necessary to explore how to more effectively add the third disturbance assessment feature to the third vector reinforcement coefficient to obtain a more expressive and predictive fourth vector reinforcement coefficient. In addition, the data intelligent repair system also needs to pay attention to the quality and integrity of the meteorological monitoring data set to ensure the accuracy and reliability of the disturbance assessment features.

[0118] In order to solve these problems and challenges, a series of measures and methods can be taken. For example, the data intelligent repair system can use machine learning and data mining techniques to more accurately evaluate the impact of the second disturbance on the third disturbance assessment feature. By building a prediction model and performing cross-validation, the data intelligent repair system can evaluate the predictive ability of different features for changes in meteorological conditions and determine their weights and importance in the fourth vector reinforcement coefficient. At the same time, the data intelligent repair system can also use feature selection algorithms to screen the most representative and predictive third disturbance assessment features, and add them to the third vector reinforcement coefficient to obtain a more expressive and predictive fourth vector reinforcement coefficient.

[0119] In addition, the data intelligent repair system can also take a series of measures to improve the quality and integrity of meteorological monitoring data sets. For example, the data intelligent repair system can strengthen the construction and maintenance of meteorological monitoring sites to ensure the accuracy and reliability of monitoring data. At the same time, the data intelligent repair system can also use data interpolation and fusion technology to fill in missing values ​​in monitoring data and improve the temporal and spatial resolution of data. Through the application of these measures and methods, the data intelligent repair system can further improve the accuracy and reliability of meteorological monitoring and forecasting.

[0120] In this way, by dynamically adjusting the target confidence vector selection strategy of the second wind load state vector mining, and combining the application of machine learning and data mining technology, feature selection algorithm, data interpolation and fusion technology, the data intelligent repair system can effectively improve the accuracy and reliability of meteorological monitoring and forecasting. This will provide more accurate and timely meteorological information support for government decision-making departments, and help formulate more scientific and effective response strategies to reduce the impact of extreme weather events on society and the economy. At the same time, this will also provide a more solid data foundation and technical support for climate change research and meteorological science research, and promote the development and application of meteorological science.

[0121] In other preferred embodiments, the method of performing multiple rounds of second wind load state vector mining based on a feature focusing strategy based on the target meteorological monitoring data set and the past monitoring correction data set includes: performing multiple rounds of interpolation and smoothing processing based on the target meteorological monitoring data set and the past monitoring correction data set, each interpolation and smoothing processing includes multiple rounds of second wind load state vector mining based on a feature focusing strategy; wherein, the time series period corresponding to the multiple rounds of interpolation and smoothing processing includes a first time series period; the first time series period is a first set period; the first time series period represents the proportion of the cumulative value of the first interpolation and smoothing processing in the global cumulative value of each round of interpolation and smoothing processing; the first interpolation and smoothing processing represents that the target confidence vectors corresponding to each round of second wind load state vector mining included are all interpolation and smoothing processing corresponding to the third vector enhancement coefficient.

[0122] In this embodiment, based on the target meteorological monitoring data set and the past monitoring correction data set, multiple rounds of second wind load state vector mining based on the feature focusing strategy not only involves in-depth mining of the data, but also includes multiple rounds of interpolation and smoothing processing to ensure the accuracy and stability of the mining results.

[0123] First, multiple rounds of interpolation and smoothing are the core of the entire mining process. Each interpolation and smoothing process includes multiple rounds of second wind load state vector mining based on feature focus strategy, which means that in each round of interpolation and smoothing, the data intelligent repair system will conduct in-depth mining and analysis of the target meteorological monitoring data set and the past monitoring and correction data set to extract the wind load state vector that is of great value to meteorological forecasting.

[0124] The time series period corresponding to the multiple rounds of interpolation and smoothing is a key indicator used by the data intelligent repair system to measure and evaluate the importance of each round of interpolation and smoothing in the global context. Among them, the first time series period is a particularly important concept, which represents the proportion of the cumulative value of the first interpolation and smoothing process in the global cumulative value of each round of interpolation and smoothing. In other words, the first time series period reflects the relative importance of the first interpolation and smoothing process in the entire multi-round interpolation and smoothing process.

[0125] Furthermore, the first interpolation smoothing process indicates that the target confidence vectors corresponding to each round of the second wind load state vector mining are all interpolation smoothing processes corresponding to the third vector reinforcement coefficient. Here, the third vector reinforcement coefficient is a key target confidence vector, which represents an optimization goal that the data intelligent repair system pursues when mining the second wind load state vector. Therefore, the first interpolation smoothing process is actually a series of second wind load state vector mining with the third vector reinforcement coefficient as the target confidence vector during the entire multi-round interpolation smoothing process.

[0126] Next, the data intelligent repair system will elaborate on the specific implementation method of the entire multi-round second wind load state vector mining based on the feature focusing strategy.

[0127] First, the data intelligent repair system will perform the first round of interpolation and smoothing based on the target meteorological monitoring data set and the past monitoring and correction data set. In this round of processing, the data intelligent repair system will adopt a feature focus strategy to focus on those features that have an important impact on meteorological forecasts and conduct in-depth mining and analysis. At the same time, the data intelligent repair system will use the third vector reinforcement coefficient as the target confidence vector and conduct multiple rounds of second wind load state vector mining to extract wind load state vectors that are valuable for meteorological forecasts.

[0128] After completing the first round of interpolation and smoothing, the data intelligent repair system will obtain a series of second wind load state vectors, which reflect the changes and characteristics of the meteorological state in the first time series period. At this time, the data intelligent repair system will update and correct the target meteorological monitoring data set and the past monitoring correction data set based on the mining results of these vectors to improve the accuracy and efficiency of subsequent mining.

[0129] Then, the data intelligent repair system will enter the second round of interpolation and smoothing processing. In this round of processing, the data intelligent repair system will continue to adopt the feature focus strategy, but at this time the data intelligent repair system will refocus and select the features based on the processing results of the first round and the updated data set. At the same time, the data intelligent repair system will still use the third vector reinforcement coefficient as the target confidence vector and conduct multiple rounds of second wind load state vector mining. Unlike the first round, the mining of the data intelligent repair system at this time will be more in-depth and detailed, because the data intelligent repair system already has the processing results of the first round as a reference and basis.

[0130] Through the continuous iteration and update of multiple rounds of interpolation and smoothing processing, the data intelligent repair system can gradually improve the accuracy and reliability of meteorological monitoring and prediction. Each round of interpolation and smoothing processing will conduct more in-depth and detailed mining and analysis based on the processing results of the previous round. At the same time, by continuously adjusting and optimizing the target confidence vector and feature focus strategy, the data intelligent repair system can better capture the changes and characteristics of meteorological conditions, providing more accurate and powerful support for meteorological forecasting.

[0131] In addition, during the process of multiple rounds of interpolation and smoothing, the data intelligent repair system will also conduct a comprehensive evaluation and analysis of the processing results of each round. By comparing the processing results of different rounds and the actual changes in the weather, the data intelligent repair system can evaluate the accuracy and effectiveness of each round of interpolation and smoothing. At the same time, the data intelligent repair system can also make timely adjustments and optimizations to the subsequent processing process based on the evaluation results to ensure the stability and reliability of the entire mining process.

[0132] For example, the following wind load state vectors are mined in the first round of interpolation and smoothing: {wind speed 1: 5m / s, wind speed 2: 6m / s, wind speed 3: 7m / s, ...}, which reflect the gradual increase of wind speed in the first time series period. Then, in the second round of interpolation and smoothing, the data intelligent repair system may mine more detailed and comprehensive wind load state vectors, such as: {wind speed 1: 5m / s, air pressure 1: 1013hPa, wind direction 1: northeast; wind speed 2: 6m / s, air pressure 2: 1012hPa, wind direction 2: east; ...}, which not only include the changes in wind speed, but also the changes in air pressure and wind direction that are closely related to wind speed.

[0133] It can be seen that this multi-round second wind load state vector mining method based on feature focusing strategy not only fully considers the changes and characteristics of meteorological states, but also gradually improves the accuracy and reliability of meteorological monitoring and prediction through continuous iteration and updating of multi-round interpolation smoothing processing. This method not only provides more accurate and powerful support for meteorological forecasting, but also opens up new ideas and directions for the research and application of meteorological science.

[0134] In some possible embodiments, if the number of second wind load monitoring information included in the second spatiotemporal data segment is greater than the number of first wind load monitoring information included in the first spatiotemporal data segment, the target vector reinforcement coefficient is any one of the second vector reinforcement coefficient, the third vector reinforcement coefficient or the fourth vector reinforcement coefficient; the target confidence vector corresponding to at least one round of second wind load state vector mining is the fourth vector reinforcement coefficient; the time series period also includes a second time series period; the second time series period is a second set period; the second time series period represents the proportion of the cumulative value of the second interpolation smoothing processing in the cumulative value of the first interpolation smoothing processing; the second interpolation smoothing processing indicates that the target confidence vectors corresponding to each round of second wind load state vector mining are all interpolation smoothing processing corresponding to the fourth vector reinforcement coefficient.

[0135] It can be understood that this embodiment faces a specific situation: the amount of second wind load monitoring information contained in the second spatiotemporal data segment exceeds the amount of first wind load monitoring information in the first spatiotemporal data segment. In this case, the selection of the target vector reinforcement coefficient by the data intelligent repair system becomes particularly critical, because it will directly affect the accuracy and efficiency of the subsequent wind load state vector mining work of the data intelligent repair system.

[0136] First, when the amount of second wind load monitoring information increases, it means that the data intelligent repair system has more data to reveal the changes and characteristics of the wind load state. Therefore, the data intelligent repair system needs a more sophisticated and powerful vector reinforcement coefficient to guide the mining work of the data intelligent repair system. At this time, the target vector reinforcement coefficient can be any of the second vector reinforcement coefficient, the third vector reinforcement coefficient, or the fourth vector reinforcement coefficient. These three vector reinforcement coefficients have the ability to process more and more complex wind load monitoring information, and they can help the data intelligent repair system to more deeply dig out the inherent laws and characteristics of the wind load state vector.

[0137] However, among these three vector reinforcement coefficients, the data intelligent repair system pays special attention to the fourth vector reinforcement coefficient. This is because there is at least one round of mining of the second wind load state vector, and its corresponding target confidence vector is the fourth vector reinforcement coefficient. This means that when the data intelligent repair system uses the fourth vector reinforcement coefficient for mining, the data intelligent repair system can obtain a wind load state vector with a high confidence level, which can accurately reflect the changes and characteristics of the wind load state.

[0138] Next, the data intelligent repair system will introduce the concept of timing cycle. The timing cycle is an important indicator to measure the progress and effect of the interpolation and smoothing processing of the data intelligent repair system. Here, the data intelligent repair system pays special attention to the second timing cycle. The second timing cycle is a set cycle, which represents the proportion of the cumulative value of the second interpolation and smoothing processing in the cumulative value of the first interpolation and smoothing processing. In other words, the second timing cycle reflects how much work the data intelligent repair system has completed when performing the second interpolation and smoothing processing, and the relative importance of these tasks in the entire interpolation and smoothing process.

[0139] Furthermore, the target confidence vectors corresponding to each round of the second wind load state vector mining included in the second interpolation smoothing process are all interpolation smoothing processes corresponding to the fourth vector reinforcement coefficient. Here, the data intelligent repair system uses the fourth vector reinforcement coefficient as the target confidence vector, which means that when the data intelligent repair system mines the second wind load state vector, it will pay special attention to those wind load state characteristics that are closely related to the fourth vector reinforcement coefficient. Through this mining method, the data intelligent repair system can obtain a series of more accurate and more detailed wind load state vectors, which can better reflect the changes and characteristics of the wind load state.

[0140] In the actual application process, the data intelligent repair system will select the appropriate vector enhancement coefficient and time series period according to the difference in the quantity of wind load monitoring information in the second space-time data segment and the first space-time data segment, as well as the data intelligent repair system's requirements for the accuracy and precision of wind load state vector mining. Specifically, when the data intelligent repair system needs to process more wind load monitoring information, or needs to obtain a more accurate and more refined wind load state vector, the data intelligent repair system will select the fourth vector enhancement coefficient as the target confidence vector and perform a second interpolation smoothing process. Through this processing method, the data intelligent repair system can make full use of the rich information in the second space-time data segment to reveal more characteristics and changing laws of the wind load state.

[0141] At the same time, although the fourth vector reinforcement coefficient and the second interpolation smoothing process can bring more accurate and refined wind load state vectors to the data intelligent repair system, they also require more computing resources and time. Therefore, in practical applications, the data intelligent repair system needs to make trade-offs and choices based on specific needs and resource conditions. If the data intelligent repair system does not have particularly high requirements for the accuracy and precision of the wind load state vector, or the computing resources of the data intelligent repair system are limited, then the data intelligent repair system can choose to use the second vector reinforcement coefficient or the third vector reinforcement coefficient for mining, and appropriately adjust the timing period to adapt to the needs of the data intelligent repair system.

[0142] In this way, the data intelligent repair system selects the appropriate vector reinforcement coefficient and time series period according to the difference in the quantity of wind load monitoring information in the second spatiotemporal data segment and the first spatiotemporal data segment, as well as the data intelligent repair system's requirements for the accuracy and precision of wind load state vector mining. The data intelligent repair system pays special attention to the fourth vector reinforcement coefficient and the second interpolation smoothing process, because they can bring more accurate and detailed wind load state vectors to the data intelligent repair system. At the same time, the data intelligent repair system must also make trade-offs and choices based on specific needs and resource conditions to ensure that the mining work of the data intelligent repair system is both accurate and efficient. Through such a technical solution, the data intelligent repair system can better utilize the wind load monitoring information in the spatiotemporal data segment, reveal more characteristics and changing laws of the wind load state, and provide more accurate and powerful support for meteorological forecasting, wind energy utilization and other fields.

[0143] In some other independent embodiments, the method also includes: determining the correlation between the first device status data in the past monitoring and correction data set and the second device status data in the first target monitoring and correction data set by jointly analyzing the first target monitoring and correction data set with the past monitoring and correction data set; the first device status data includes data in the past monitoring and correction data set except the first environmental wind load data corresponding to the first spatiotemporal data segment; the second device status data includes data in the first target monitoring and correction data set except the second environmental wind load data corresponding to the second spatiotemporal data segment; if the correlation meets the set requirements, the first target monitoring and correction data set is used as the second target monitoring and correction data set corresponding to the target meteorological monitoring data set; if the correlation does not meet the set requirements, the timing period is changed, and the first target monitoring and correction data set corresponding to the changed timing period is obtained, and the correlation corresponding to the changed first target monitoring and correction data set is determined until the determined correlation meets the set requirements.

[0144] It can be understood that the method involved in the embodiment of the present application is not only limited to the mining and processing of wind load information in the spatiotemporal data segment, but also further extends to the in-depth analysis of the equipment status data. Specifically, an attempt is made to reveal the correlation between the hidden equipment status data by jointly analyzing the first target monitoring correction data set and the past monitoring correction data set. The introduction of this step provides a new perspective and possibility for meteorological monitoring and forecasting.

[0145] First, the equipment status data does not include the environmental wind load data corresponding to the spatiotemporal data segment. In the first target monitoring and correction data set, the data intelligent repair system focuses on data other than the second environmental wind load data corresponding to the second spatiotemporal data segment, which is called the second equipment status data. Similarly, in the past monitoring and correction data set, the data intelligent repair system focuses on data other than the first environmental wind load data corresponding to the first spatiotemporal data segment, which is called the first equipment status data. This setting enables the data intelligent repair system to focus on mining the correlation between equipment status data that are relatively independent of environmental wind load data.

[0146] Next, the data intelligent repair system performs a joint analysis process. The core of this step is to determine whether there is a certain correlation between the first device status data and the second device status data by comparing and matching them. This correlation may be manifested as consistency in data change trends, a certain functional relationship between data values, or similarity in data features. Regardless of the form, as long as this correlation can meet the requirements set by the data intelligent repair system, it can be considered that the first target monitoring correction data set is the second target monitoring correction data set corresponding to the target meteorological monitoring data set.

[0147] If the correlation does not meet the set requirements, the data intelligent repair system needs to make a series of adjustments and optimizations. First, the data intelligent repair system will change the timing cycle. The timing cycle is an important indicator to measure the progress and effect of the interpolation and smoothing processing of the data intelligent repair system. By adjusting the timing cycle, the data intelligent repair system can change the strategy and focus of the interpolation and smoothing processing, thereby affecting the correlation between the device status data finally mined. After changing the timing cycle, the data intelligent repair system will re-acquire the corresponding first target monitoring correction data set and determine the correlation again. This process may be repeated many times until the correlation determined by the data intelligent repair system meets the set requirements.

[0148] Through such a technical solution, the data intelligent repair system can more comprehensively mine and utilize the information in the meteorological monitoring data set. Not only limited to the mining and processing of environmental wind load data, the data intelligent repair system can also go deep into the level of equipment status data and reveal the correlation between different equipment status data. The mining and utilization of this correlation is of great significance to improving the accuracy and reliability of meteorological monitoring and forecasting. For example, in some cases, changes in equipment status data may reflect the changing trend of environmental wind load data in advance, thereby providing earlier and more accurate warning information for the data intelligent repair system.

[0149] At the same time, the implementation of this technical solution requires certain computing resources and time costs. Because the correlation mining between equipment status data is a relatively complex process, it requires a lot of data comparison and matching work. Therefore, in practical applications, the data intelligent repair system needs to make trade-offs and choices based on specific needs and resource conditions. If the data intelligent repair system does not have particularly high requirements for the correlation between equipment status data, or the computing resources are limited, then the data intelligent repair system can appropriately simplify the correlation mining process, or choose to use other more efficient data mining methods.

[0150] In this way, the data intelligent repair system reveals the correlation between the hidden equipment status data by jointly analyzing the first target monitoring and correction data set and the past monitoring and correction data set. If the correlation meets the set requirements, the data intelligent repair system can use the first target monitoring and correction data set as the second target monitoring and correction data set corresponding to the target meteorological monitoring data set; if the correlation does not meet the set requirements, the data intelligent repair system will perform a series of adjustments and optimizations until the correlation meets the requirements. Through such a technical solution, the data intelligent repair system can more comprehensively mine and utilize the information in the meteorological monitoring data set, and improve the accuracy and reliability of meteorological monitoring and forecasting. At the same time, the data intelligent repair system also needs to make reasonable choices and trade-offs based on actual conditions and needs to ensure that the method is both accurate and efficient.

[0151] In some other independent embodiments, the correlation between the first device status data in the past monitoring correction data set and the second device status data in the first target monitoring correction data set is determined by jointly analyzing the first target monitoring correction data set and the past monitoring correction data set, including: identifying the first wind speed and wind direction information corresponding to the past monitoring correction data set and the first spatiotemporal data segment, setting the feature description variable of the first wind speed and wind direction information in the past monitoring correction data set as the set description variable, and extracting the first device status data in the past monitoring correction data set; identifying the second wind speed and wind direction information corresponding to the second spatiotemporal data segment in the first target monitoring correction data set. wind direction information, extract the target wind speed and direction trend vector corresponding to the second wind speed and direction information, and extract the target meteorological monitoring trend vector of the target meteorological monitoring data set; perform correlation analysis on the target wind speed and direction trend vector and the target meteorological monitoring trend vector to obtain a first correlation; set the feature description variable of the second wind speed and direction information in the first target monitoring correction data set to the set description variable, and extract the second device status data in the first target monitoring correction data set; perform correlation analysis on the data features of the first device status data and the data features of the second device status data to obtain a second correlation; determine the correlation based on the first correlation and the second correlation.

[0152] It can be understood that this embodiment deeply explores the intrinsic connection between the first target monitoring correction data set and the past monitoring correction data set, especially the correlation between the device status data respectively contained therein.

[0153] First, the data intelligent repair system identifies the first wind speed and direction information corresponding to the first spatiotemporal data segment from the past monitoring and correction data set. This step is crucial because it provides the data intelligent repair system with a basis for analysis - the wind speed and direction characteristics under a specific spatiotemporal background. In order to understand these characteristics more deeply, the data intelligent repair system sets the characteristic description variables of the first wind speed and direction information in the past monitoring and correction data set as set description variables. This setting enables the data intelligent repair system to more accurately extract the first equipment status data in the past monitoring and correction data set, laying a solid foundation for subsequent analysis.

[0154] Next, the data intelligent repair system turns to the first target monitoring correction data set and identifies the second wind speed and direction information corresponding to the second spatiotemporal data segment. Similar to past data, the data intelligent repair system also extracts the target wind speed and direction trend vector corresponding to the second wind speed and direction information. But this step does not stop there. The data intelligent repair system further extracts the target meteorological monitoring trend vector of the target meteorological monitoring data set. The extraction of this vector provides another important analysis dimension for the data intelligent repair system - the overall trend of the meteorological monitoring data.

[0155] With these two trend vectors, the data intelligent repair system can carry out the next step of correlation analysis. The data intelligent repair system correlates the target wind speed and direction trend vector with the target meteorological monitoring trend vector to obtain the first correlation. This correlation reflects the inherent connection between wind speed and direction information and meteorological monitoring data in the second spatiotemporal data segment. It is an important part of the data intelligent repair system to understand the correlation of equipment status data.

[0156] However, the data intelligent repair system's analysis does not stop there. In order to more fully understand the relevance of the equipment status data, the data intelligent repair system also needs to consider the characteristic description variables of the second wind speed and direction information in the first target monitoring and correction data set. The data intelligent repair system sets these characteristic description variables to the same set description variables as those in the previous monitoring and correction data set. This setting enables the data intelligent repair system to more accurately extract the second equipment status data in the first target monitoring and correction data set.

[0157] With the first device status data and the second device status data, the data intelligent repair system can perform the next step of correlation analysis. The data intelligent repair system performs correlation analysis on the data features of the first device status data and the data features of the second device status data, thereby obtaining the second correlation. This correlation reflects the inherent connection between the device status data in different time and space data segments. It is an indispensable part of the data intelligent repair system to understand the correlation of device status data.

[0158] Finally, the data intelligent repair system determines the overall correlation based on the first correlation and the second correlation. This step is a summary and refinement of all the previous analysis work. By comprehensively considering the correlation between wind speed and direction information and meteorological monitoring data, as well as the correlation between equipment status data in different time and space data segments, the data intelligent repair system can more accurately understand the intrinsic connection between the first target monitoring correction data set and the past monitoring correction data set.

[0159] In this way, the data intelligent repair system uses a series of carefully designed steps to deeply explore the correlation between the first target monitoring and correction data set and the past monitoring and correction data set. The data intelligent repair system starts with identifying wind speed and direction information, and gradually reveals the inherent connection between equipment status data through multiple steps such as extracting feature description variables, trend vectors, and performing association analysis. This exploration process not only provides a deeper understanding of the data intelligent repair system, but also provides new perspectives and methods for meteorological monitoring and forecasting. Through this method, the data intelligent repair system can more accurately grasp the laws of meteorological changes and provide more powerful support for meteorological monitoring and forecasting.

[0160] Further, Figure 2 FIG. 2 is a schematic diagram of a data intelligent repair system 200 provided in an embodiment of the present application. Figure 2 The data intelligent repair system 200 shown includes a processor 210, which can call and run a computer program from a memory to implement the method in the embodiment of the present application.

[0161] Alternatively, if Figure 2 As shown, the data intelligent repair system 200 may further include a memory 230. The processor 210 may call and run a computer program from the memory 230 to implement the method in the embodiment of the present application.

[0162] The memory 230 may be a separate device independent of the processor 210 , or may be integrated into the processor 210 .

[0163] Alternatively, if Figure 2 As shown, the data intelligent repair system 200 may further include a transceiver 220, and the processor 210 may control the transceiver 220 to interact with other devices, specifically, may send information or data to other devices, or receive information or data sent by other devices.

[0164] Optionally, the data intelligent repair system 200 can implement the corresponding processes corresponding to the storage engine or components in the storage engine (such as a processing module) or a device deployed with a storage engine in each method of the embodiments of the present application. For the sake of brevity, they are not repeated here.

[0165] It should be understood that the processor of the embodiment of the present application may be an integrated circuit chip with signal processing capabilities.

[0166] It is understood that the memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory. It should be noted that the memory of the systems and methods described herein is intended to include but is not limited to suitable types of memory.

[0167] Based on the above, a readable storage medium is provided, on which a program or instruction is stored, and the program or instruction implements the steps of the above method when executed by a processor.

[0168] It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0169] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, a disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0170] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the present application, all of which are within the protection of the present application.

Claims

1. An AI-based intelligent repair method for historical data of wind direction and wind speed in high-cold areas, characterized in that: The method is applied to a data intelligent repair system, and the method comprises: Acquire a target meteorological monitoring data set; the target meteorological monitoring data set is obtained by updating a first spatiotemporal data segment in a past meteorological monitoring data set to a second spatiotemporal data segment; the first spatiotemporal data segment includes at least one first wind load monitoring information; the second spatiotemporal data segment includes at least one second wind load monitoring information; Acquire a past monitoring correction data set generated by a data repair algorithm based on the past meteorological monitoring data set, and a first vector reinforcement coefficient obtained by performing multiple rounds of first wind load state vector mining based on a feature focusing strategy when generating the past monitoring correction data set; Implementing intelligent data repair processing through the data repair algorithm to obtain a first target monitoring correction data set corresponding to the target meteorological monitoring data set; Wherein, the data intelligent repair processing includes: according to the target meteorological monitoring data set and the past monitoring correction data set, multiple rounds of second wind load state vector mining based on the feature focusing strategy are performed to obtain the second vector reinforcement coefficients corresponding to each round of second wind load state vector mining, and according to the first wind load state vector corresponding to the target meteorological monitoring data set and the target vector reinforcement coefficients corresponding to each round of second wind load state vector mining, the wind load state spatiotemporal vectors of each round of second wind load state vector mining are obtained, and based on the wind load state spatiotemporal vectors of the last round of second wind load state vector mining, the first target monitoring correction data set corresponding to the target meteorological monitoring data set is generated; Among them, the target vector reinforcement coefficient corresponding to any second wind-load state vector mining is the second vector reinforcement coefficient or the third vector reinforcement coefficient; the target confidence vector corresponding to at least one round of second wind-load state vector mining is the third vector reinforcement coefficient obtained based on the following steps: based on the vector reinforcement coefficient corresponding to the second space-time data segment in the second vector reinforcement coefficient corresponding to the second wind-load state vector mining of the target wheel, the vector reinforcement coefficient corresponding to the first space-time data segment in the first vector reinforcement coefficient of the target wheel is adjusted to obtain the third vector reinforcement coefficient.

2. The method according to claim 1, characterized in that The method of performing multiple rounds of second wind load state vector mining based on a feature focusing strategy according to the target meteorological monitoring data set and the past monitoring correction data set to obtain the second vector enhancement coefficients corresponding to each round of second wind load state vector mining includes: Extracting a first wind load state vector corresponding to the target meteorological monitoring data set, and extracting a second wind load state vector of the past monitoring correction data set; Determine the regional environment attribute vectors corresponding to the second wind load state vectors of each round according to the first wind load state vector; Based on the index representation vectors and regional environmental attribute vectors corresponding to each round of second wind load state vector mining, the second vector reinforcement coefficients corresponding to each round of second wind load state vector mining are determined; wherein, the index representation vector corresponding to the first round of second wind load state vector mining is determined based on the second wind load state vector; the index representation vector corresponding to the second wind load state vector mining other than the first round is determined based on the wind load state spatiotemporal vector of the previous round of second wind load state vector mining.

3. The method according to claim 1, characterized in that If the target confidence vector corresponding to a round of second wind-load state vector mining is the third vector enhancement coefficient, then the target confidence vectors corresponding to subsequent rounds of second wind-load state vector mining are all the third vector enhancement coefficient.

4. The method according to claim 1, characterized in that: The first vector enhancement coefficient includes a disturbance assessment feature corresponding to each first wind load monitoring information in the past meteorological monitoring data set; The second vector enhancement coefficient includes a disturbance assessment feature corresponding to each second wind load monitoring information in the target meteorological monitoring data set; If the number of second wind load monitoring information included in the second spatiotemporal data segment is equal to the number of first wind load monitoring information included in the first spatiotemporal data segment, the vector reinforcement coefficient corresponding to the second spatiotemporal data segment in the second vector reinforcement coefficient corresponding to the second wind load state vector of the target wheel is mined, and the vector reinforcement coefficient corresponding to the first spatiotemporal data segment in the first vector reinforcement coefficient of the target wheel is adjusted to obtain the third vector reinforcement coefficient, including: Determine first disturbance assessment features in the first vector enhancement coefficient that respectively correspond to at least one first wind load monitoring information of the first spatiotemporal data segment; Determine a second disturbance assessment feature in the second vector enhancement coefficient that corresponds to at least one second wind load monitoring information of the second spatiotemporal data segment; The third vector enhancement coefficient is obtained by updating at least one first disturbance assessment feature in the first vector enhancement coefficient using the at least one second disturbance assessment feature.

5. The method according to claim 1, characterized in that The first vector enhancement coefficient includes a disturbance assessment feature corresponding to each first wind load monitoring information in the past meteorological monitoring data set; The second vector enhancement coefficient includes a disturbance assessment feature corresponding to each second wind load monitoring information in the target meteorological monitoring data set; If the number of second wind load monitoring information included in the second spatiotemporal data segment is not equal to the number of first wind load monitoring information included in the first spatiotemporal data segment, the vector reinforcement coefficient corresponding to the second spatiotemporal data segment in the second vector reinforcement coefficient corresponding to the second wind load state vector of the target wheel is mined, and the vector reinforcement coefficient corresponding to the first spatiotemporal data segment in the first vector reinforcement coefficient of the target wheel is adjusted to obtain the third vector reinforcement coefficient, including: Determine first disturbance assessment features in the first vector enhancement coefficient that respectively correspond to at least one first wind load monitoring information of the first spatiotemporal data segment; Determine a second disturbance assessment feature in the second vector enhancement coefficient that corresponds to at least one second wind load monitoring information of the second spatiotemporal data segment; Determining a first disturbance influence degree of each second disturbance assessment feature in the second space-time data segment according to a first number of the first wind load monitoring information in the first space-time data segment and a second number of the second wind load monitoring information in the second space-time data segment; Using the first disturbance influence degree of each second disturbance assessment feature, and using the at least one second disturbance assessment feature, at least one first disturbance assessment feature in the first vector enhancement coefficient is updated to obtain the third vector enhancement coefficient.

6. The method according to claim 5, characterized in that If the number of second wind load monitoring information included in the second spatiotemporal data segment is greater than the number of first wind load monitoring information included in the first spatiotemporal data segment, the target vector enhancement coefficient is any one of the second vector enhancement coefficient, the third vector enhancement coefficient or the fourth vector enhancement coefficient; the target confidence vector corresponding to at least one round of second wind load state vector mining is the fourth vector enhancement coefficient; The fourth vector enhancement coefficient is obtained by the following steps: adding at least one third disturbance assessment feature to the third vector enhancement coefficient to obtain the fourth vector enhancement coefficient; the at least one third disturbance assessment feature includes the at least one second disturbance assessment feature.

7. The method according to claim 6, characterized in that The step of adding at least one third disturbance assessment feature to the third vector enhancement coefficient to obtain the fourth vector enhancement coefficient comprises: Determining a second disturbance influence degree of each third disturbance assessment feature by using the first disturbance influence degree of each second disturbance assessment feature; By using the second disturbance influence degree of each third disturbance assessment feature, at least one third disturbance assessment feature is added to the third vector enhancement coefficient to obtain the fourth vector enhancement coefficient.

8. The method according to claim 6, characterized in that If the target confidence vector of a round of second wind-load state vector mining is the third vector reinforcement coefficient, then the target confidence vectors corresponding to the subsequent rounds of second wind-load state vector mining are the third vector reinforcement coefficient or the fourth vector reinforcement coefficient; if the target confidence vector corresponding to a round of second wind-load state vector mining is the fourth vector reinforcement coefficient, then the target confidence vectors corresponding to the subsequent rounds of second wind-load state vector mining are all the fourth vector reinforcement coefficient.

9. The method according to claim 1, characterized in that: The method of performing multiple rounds of second wind load state vector mining based on a feature focusing strategy according to the target meteorological monitoring data set and the past monitoring correction data set includes: Multiple rounds of interpolation and smoothing processing are performed based on the target meteorological monitoring data set and the past monitoring and correction data set, and each interpolation and smoothing processing includes multiple rounds of second wind load state vector mining based on a feature focusing strategy; wherein the time series period corresponding to the multiple rounds of interpolation and smoothing processing includes a first time series period; the first time series period is a first set period; the first time series period represents the proportion of the cumulative value of the first interpolation and smoothing processing in the global cumulative value of each round of interpolation and smoothing processing; the first interpolation and smoothing processing represents that the target confidence vectors corresponding to each round of second wind load state vector mining included are all interpolation and smoothing processing corresponding to the third vector enhancement coefficient; If the number of second wind load monitoring information included in the second spatiotemporal data segment is greater than the number of first wind load monitoring information included in the first spatiotemporal data segment, the target vector enhancement coefficient is any one of the second vector enhancement coefficient, the third vector enhancement coefficient or the fourth vector enhancement coefficient; the target confidence vector corresponding to at least one round of second wind load state vector mining is the fourth vector enhancement coefficient; the time series period also includes a second time series period; the second time series period is a second set period; the second time series period represents the proportion of the cumulative value of the second interpolation smoothing processing in the cumulative value of the first interpolation smoothing processing; the second interpolation smoothing processing represents that the target confidence vectors corresponding to each round of second wind load state vector mining are all interpolation smoothing processing corresponding to the fourth vector enhancement coefficient.

10. A data intelligent repair system, characterized in that: The method comprises at least one processor and a memory; the memory stores computer-executable instructions; the at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the method according to any one of claims 1 to 9.

Images