Full-time-sequence artificial cloud elimination and rain reduction operation early warning analysis method and system

Abstract

The invention discloses a full-time-sequence artificial cloud elimination and rain reduction operation early warning analysis method and system, and belongs to the technical field of artificial intelligence and intelligent meteorology. According to the method, the to-be-analyzed multi-dimensional meteorological element data is acquired and is subjected to comprehensive early warning, research and judgment processing, so that comprehensive and deep analysis of meteorological conditions is innovatively realized. In detail, the early warning operation event label distribution data corresponding to the meteorological monitoring representation vectors are obtained, and the early warning operation event label discrimination weight, the first cloud cluster dynamic analysis viewpoint and the second cloud cluster dynamic analysis viewpoint are introduced, so that the target early warning operation event can be identified more accurately. According to the rich information, the target early warning operation event label corresponding to each strategy mode type is accurately determined, scientific and effective artificial cloud elimination and rain reduction indication information is finally generated, the efficiency and effect of cloud elimination and rain reduction operation are remarkably improved, and a new solution is provided for reducing the influence of meteorological disasters on the human society.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence and intelligent meteorological technology, and specifically relates to a full-time artificial cloud dissipation and rain reduction operation early warning analysis method and system. Background Art

[0002] In the field of meteorology, the analysis and application of multidimensional meteorological element data has always been a challenging topic. Traditional meteorological analysis methods often focus on the study of a single meteorological element, which makes it difficult to fully and accurately reflect the complexity and variability of meteorological conditions. At the same time, in artificial cloud dissipation and rain reduction operations, due to the lack of scientific operation guidance and comprehensive meteorological information, the efficiency and effectiveness of operations are often restricted. Therefore, how to conduct a comprehensive analysis of multidimensional meteorological element data and generate accurate artificial cloud dissipation and rain reduction instruction information based on it has become a technical problem that needs to be solved urgently. Summary of the invention

[0003] The present application provides a full-time artificial cloud dissipation and rain reduction operation early warning analysis method and system, which can solve or partially solve the technical problems involved in the above-mentioned background technology.

[0004] The embodiment of the present application provides a full-time artificial cloud dissipation and rain reduction operation early warning analysis method, which is applied to an early warning analysis system. The method comprises: obtaining multi-dimensional meteorological element data to be analyzed;

[0005] Performing comprehensive early warning analysis and judgment processing on the multidimensional meteorological element data to be analyzed, obtaining early warning operation event label distribution data, early warning operation event label discrimination weights, a first cloud group dynamic analysis viewpoint, and a second cloud group dynamic analysis viewpoint corresponding to each of the multiple meteorological monitoring characterization vectors of the multidimensional meteorological element data to be analyzed; wherein, the target early warning operation event includes at least one full-time manual operation project that activates a non-linked operation task or a linked operation task, the early warning operation event label discrimination weight represents the probability value that the early warning operation event label corresponding to the meteorological monitoring characterization vector is the target early warning operation event label, the first cloud group dynamic analysis viewpoint represents the probability value that the target early warning operation event belongs to the target cloud group trend type or the probability value that the strategy mode of the target early warning operation event belongs to the strategy mode type of the target linked operation task, and the second cloud group dynamic analysis viewpoint represents the probability value that the strategy mode of the target early warning operation event belongs to the strategy mode type of the target non-linked operation task;

[0006] Determine a target warning operation event label corresponding to each strategy mode type from the warning operation event labels corresponding to the multiple meteorological monitoring characterization vectors according to the warning operation event label distribution data, the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint;

[0007] The artificial cloud dissipation and rain reduction indication information of the multidimensional meteorological element data to be analyzed is generated based on the warning operation event label distribution data of the target warning operation event label, the warning operation event label discrimination weight, the target cloud group trend type and the strategy mode type.

[0008] In one implementation, the comprehensive early warning analysis and judgment processing is performed on the multidimensional meteorological element data to be analyzed to obtain early warning operation event label distribution data, early warning operation event label discrimination weights, first cloud group dynamic analysis viewpoints, and second cloud group dynamic analysis viewpoints corresponding to multiple meteorological monitoring characterization vectors of the multidimensional meteorological element data to be analyzed, including:

[0009] Extracting cross meteorological element vectors from the multidimensional meteorological element data to be analyzed to obtain a cross meteorological element vector set;

[0010] Performing early warning identification based on the AI ​​model on the cross meteorological element vector set, obtaining early warning operation event label distribution data, early warning operation event label discrimination weights, a first cloud group dynamic analysis viewpoint and a second cloud group dynamic analysis viewpoint corresponding to each of the multiple meteorological monitoring characterization vectors, wherein the first cloud group dynamic analysis viewpoint includes discrimination weights that the target early warning operation event belongs to multiple operation project keywords respectively and discrimination weights that the strategy mode of the target early warning operation event belongs to multiple strategy mode types of linkage operation tasks respectively, and the second cloud group dynamic analysis viewpoint includes discrimination weights that the strategy mode of the target early warning operation event belongs to multiple strategy mode types of non-linkage operation tasks respectively;

[0011] Determine the target cloud cluster trend type or the target linkage operation task from the multiple operation project keywords and the multiple linkage operation tasks according to the first cloud cluster dynamic analysis viewpoint, and determine the discrimination weight of the target cloud cluster trend type or the target linkage operation task as the first cloud cluster dynamic analysis viewpoint;

[0012] The target non-linked operation task is determined from the multiple non-linked operation tasks according to the second cloud group dynamic analysis viewpoint, and the discrimination weight corresponding to the target non-linked operation task is determined as the second cloud group dynamic analysis viewpoint.

[0013] In one implementation, the determining of the target warning operation event label corresponding to each strategy mode type from the warning operation event labels corresponding to the plurality of meteorological monitoring characterization vectors based on the warning operation event label distribution data, the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint includes:

[0014] Determine, from the warning operation event labels corresponding to the multiple meteorological monitoring characterization vectors, multiple preliminary screening operation event labels and the types of strategy patterns corresponding to the multiple preliminary screening operation event labels according to the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint;

[0015] According to the warning operation event label distribution data and the original cloud cluster dynamic analysis viewpoint, the preliminary screening operation event label corresponding to each strategy pattern type is optimized to obtain the target warning operation event label corresponding to each strategy pattern type.

[0016] In one implementation, the determining of a plurality of preliminary screening operation event labels and the types of strategy patterns corresponding to the plurality of preliminary screening operation event labels from the warning operation event labels corresponding to the plurality of meteorological monitoring characterization vectors based on the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint includes:

[0017] The warning operation event label discrimination weights and the first cloud cluster dynamic analysis viewpoints of each warning operation event label are respectively divided into groups to obtain the original cloud cluster dynamic analysis viewpoints corresponding to each warning operation event label;

[0018] Determine the early warning operation event label corresponding to the original cloud dynamic analysis viewpoint that meets the viewpoint discrimination requirements as the preliminary screening operation event label;

[0019] If the first cloud group dynamic analysis viewpoint reflects that the strategy mode of the target warning operation event is the target linkage operation task, the strategy mode type of the target linkage operation task is determined as the strategy mode type corresponding to the initial screening operation event label.

[0020] In one implementation, the method further includes:

[0021] If the first cloud cluster dynamic analysis viewpoint reflects that the target warning operation event is a target cloud cluster trend type, based on the second cloud cluster dynamic analysis viewpoint, the strategy mode type of the target non-linked operation task is determined as the strategy mode type corresponding to the initial screening operation event label.

[0022] In one implementation, the preliminary screening operation event labels corresponding to each strategy mode type are optimized based on the warning operation event label distribution data and the original cloud group dynamic analysis viewpoint, and the target warning operation event labels corresponding to each strategy mode type are obtained, including:

[0023] According to the warning operation event label distribution data and the original cloud cluster dynamic analysis viewpoint, redundant noise cleaning is performed on each preliminary screening operation event label corresponding to each strategy pattern type to filter out redundant warning operation event labels in each preliminary screening operation event label, and obtain the target warning operation event label corresponding to each strategy pattern type.

[0024] In one implementation, the full-time artificial cloud dissipation and rain reduction operation early warning analysis method is implemented based on a comprehensive early warning analysis algorithm model, and the method further includes:

[0025] Acquire an algorithm model learning sample set, wherein the algorithm model learning sample set includes a multidimensional meteorological element data sample and a priori warning operation event annotation, a priori cloud trend type annotation, and a priori strategy mode type annotation corresponding to the multidimensional meteorological element data sample;

[0026] Input the multi-dimensional meteorological element data samples into the joint neural network to be debugged respectively, and obtain early warning operation event label distribution data samples, early warning operation event label discrimination weight samples, first cloud group dynamic analysis viewpoint samples, and second cloud group dynamic analysis viewpoint samples corresponding to multiple meteorological monitoring characterization vectors of the multi-dimensional meteorological element data samples;

[0027] Determine a benchmark warning operation event label corresponding to each strategy mode type from the warning operation event label samples corresponding to the multiple meteorological monitoring characterization vectors according to the warning operation event label distribution data sample, the warning operation event label discrimination weight sample, the first cloud group dynamic analysis viewpoint sample, and the second cloud group dynamic analysis viewpoint sample;

[0028] Determine a first training learning error according to the warning operation event label discrimination weight samples corresponding to each of the plurality of meteorological monitoring characterization vectors and the priori warning operation event annotations;

[0029] Determine a second training learning error according to the warning operation event label distribution data sample corresponding to the benchmark warning operation event label, the first cloud group dynamic analysis viewpoint sample and the second cloud group dynamic analysis viewpoint sample, and the prior warning operation event annotation, the prior cloud group trend type annotation and the prior strategy mode type annotation;

[0030] The joint neural network to be debugged is debugged based on the first training learning error and the second training learning error to obtain the comprehensive early warning analysis algorithm model.

[0031] In one implementation, the second training learning error includes a first local training learning error, a second local training learning error, and a third local training learning error. According to the warning operation event label distribution data sample corresponding to the benchmark warning operation event label, the first cloud group dynamic analysis viewpoint sample and the second cloud group dynamic analysis viewpoint sample, and the prior warning operation event annotation, the prior cloud group trend type annotation, and the prior strategy mode type annotation, determining the second training learning error includes:

[0032] Determining the first local training learning error according to the warning operation event label distribution data corresponding to the prior warning operation event annotation and the benchmark warning operation event label, wherein the first local training learning error represents the event label recognition offset between the benchmark warning operation event label and the warning operation event authentication label;

[0033] Determine the second local training learning error according to the prior cloud trend type annotation, the prior strategy pattern type annotation and the first cloud dynamic analysis viewpoint example corresponding to the benchmark warning operation event label, wherein the second local training learning error represents the discrimination offset between the first cloud dynamic analysis viewpoint example corresponding to the benchmark warning operation event label and the strategy pattern authentication type;

[0034] Based on the prior strategy pattern type annotation, and the first cloud cluster dynamic analysis viewpoint sample and the second cloud cluster dynamic analysis viewpoint sample corresponding to the benchmark warning operation event label, the third local training learning error is determined, and the third local training learning error characterizes the discrimination offset between the second cloud cluster dynamic analysis viewpoint sample corresponding to the benchmark warning operation event label and the strategy pattern authentication type.

[0035] An embodiment of the present application provides an early warning analysis 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 executes the above method.

[0036] 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.

[0037] In the embodiment of the present application, by acquiring the multi-dimensional meteorological element data to be analyzed and conducting comprehensive early warning research and judgment processing on it, a comprehensive and in-depth analysis of meteorological conditions is innovatively achieved. In detail, not only the early warning operation event label distribution data corresponding to each meteorological monitoring characterization vector is obtained, but also the early warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint and the second cloud group dynamic analysis viewpoint are further introduced, so as to more accurately identify the target early warning operation event. Then, based on this rich information, the target early warning operation event label corresponding to each strategy mode type is accurately determined, and finally a scientific and effective artificial cloud dissipation and rain reduction instruction information is generated, which significantly improves the efficiency and effect of cloud dissipation and rain reduction operations, and provides a new solution for reducing the impact of meteorological disasters on human society. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A flowchart of a full-time artificial cloud dissipation and rain reduction operation early warning analysis method provided in an embodiment of the present application.

[0039] Figure 2 A schematic diagram of the structure of an early warning analysis system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0040] 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.

[0041] 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.

[0042] Figure 1 A full-time artificial cloud dissipation and rain reduction operation early warning analysis method is shown, which is applied to an early warning analysis system. The method includes the following steps 110 to 140.

[0043] In the field of meteorological science and technology, artificial cloud dissipation and rain reduction operations, as an important means of meteorological intervention, are of great significance for alleviating the impact of extreme weather, ensuring agricultural production, optimizing water resource allocation, etc. The technical solution provided in the embodiment of the present application provides accurate indication information for artificial cloud dissipation and rain reduction operations through comprehensive analysis of multi-dimensional meteorological data.

[0044] Step 110: Obtain multidimensional meteorological element data to be analyzed.

[0045] In the embodiment of the present application, the primary task of the early warning analysis system is to obtain multidimensional meteorological element data to be analyzed. The multidimensional meteorological element data to be analyzed includes, but is not limited to: satellite cloud images, radar echoes, ground observation station data, high-altitude detection data, and numerical weather forecast products. The multidimensional meteorological element data to be analyzed covers all-round meteorological information from the ground to the sky, from the microscopic to the macroscopic, and is the basis for the early warning analysis system to conduct comprehensive research and judgment.

[0046] For example, the early warning analysis system can obtain the latest meteorological observation and forecast data from various meteorological data sources in real time or quasi-real time through a network interface or data exchange platform. After acquisition, these data will undergo a series of quality control and data preprocessing steps to ensure the accuracy and consistency of the data, providing reliable data support for subsequent comprehensive early warning analysis.

[0047] Step 120, perform comprehensive early warning analysis and processing on the multidimensional meteorological element data to be analyzed, and obtain early warning operation event label distribution data, early warning operation event label discrimination weights, first cloud group dynamic analysis viewpoints, and second cloud group dynamic analysis viewpoints corresponding to multiple meteorological monitoring characterization vectors of the multidimensional meteorological element data to be analyzed.

[0048] Among them, the target early warning operation event includes at least one full-time manual operation project that activates a non-linked operation task or a linked operation task, the early warning operation event label discrimination weight represents the probability value of the early warning operation event label corresponding to the meteorological monitoring characterization vector being the target early warning operation event label, the first cloud group dynamic analysis viewpoint represents the probability value of the target early warning operation event belonging to the target cloud group trend type or the probability value of the strategy pattern of the target early warning operation event belonging to the strategy pattern type of the target linked operation task, and the second cloud group dynamic analysis viewpoint represents the probability value of the strategy pattern of the target early warning operation event belonging to the strategy pattern type of the target non-linked operation task.

[0049] In an embodiment of the present application, after obtaining the multidimensional meteorological element data to be analyzed, the early warning analysis system can perform comprehensive early warning analysis and processing on these data. This processing process involves complex data mining, machine learning algorithms, and meteorological expertise, and aims to extract information that is instructive for artificial cloud dissipation and rain reduction operations from massive data.

[0050] The core task of comprehensive early warning analysis and processing is to generate meteorological monitoring characterization vectors and assign corresponding early warning operation event label distribution data, early warning operation event label discrimination weights, first cloud group dynamic analysis viewpoints and second cloud group dynamic analysis viewpoints to these vectors.

[0051] In detail, the warning operation event label distribution data characterizes the degree of association between the meteorological monitoring representation vector and different warning operation events. For example, a certain meteorological monitoring representation vector may have different degrees of association with warning operation events such as "activating non-linked operation tasks" or "linked operation tasks". The warning operation event label discrimination weight is a probability value, which indicates the possibility that the warning operation event label corresponding to the meteorological monitoring representation vector is the target warning operation event label. The higher the weight value, the more likely the meteorological monitoring representation vector is to trigger the target warning operation event. The first cloud cluster dynamic analysis viewpoint characterizes the probability value of the target warning operation event belonging to a certain cloud cluster trend type, or the probability value of the target warning operation event strategy mode belonging to a certain linkage operation task strategy mode type. For example, it may indicate that a certain warning operation event is highly correlated with cloud cluster trends such as "cloud clusters are developing rapidly" or "cloud clusters are about to dissipate". The second cloud cluster dynamic analysis viewpoint focuses on the probability value of the target warning operation event strategy mode belonging to a certain non-linked operation task strategy mode type, which helps the warning analysis system consider the possibility of non-linked operation tasks when formulating operation strategies.

[0052] Step 130: Determine the target warning operation event label corresponding to each strategy mode type from the warning operation event labels corresponding to the multiple meteorological monitoring characterization vectors based on the warning operation event label distribution data, the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint.

[0053] In an embodiment of the present application, after completing the comprehensive warning analysis and processing, the warning analysis system can determine the target warning operation event label corresponding to each strategy mode type from the warning operation event labels corresponding to multiple meteorological monitoring characterization vectors based on the warning operation event label distribution data, the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint. This process involves complex decision logic and algorithm optimization to ensure that the final target warning operation event label is both accurate and reliable. For example, the warning analysis system may select the meteorological monitoring characterization vector that is most likely to trigger the target warning operation event based on the comprehensive evaluation results of the first cloud group dynamic analysis viewpoint and the second cloud group dynamic analysis viewpoint, and use its corresponding warning operation event label as the target warning operation event label.

[0054] Step 140: Generate artificial cloud dissipation and rain reduction indication information for the multidimensional meteorological element data to be analyzed based on the warning operation event label distribution data of the target warning operation event label, the warning operation event label discrimination weight, the target cloud trend type and the strategy mode type.

[0055] Finally, the early warning analysis system can generate artificial cloud dissipation and rain reduction indication information for the multi-dimensional meteorological element data to be analyzed based on the warning operation event label distribution data of the target warning operation event label, the warning operation event label discrimination weight, the target cloud trend type and the strategy mode type.

[0056] The artificial cloud dissipation and rain reduction instruction information contains all the key information required for artificial cloud dissipation and rain reduction operations, such as operation time, operation location, operation method (linked or non-linked), expected results, etc. It can provide clear guidance for meteorological departments or related operation units, helping them to formulate scientific and reasonable artificial cloud dissipation and rain reduction operation plans.

[0057] For example, if the early warning analysis system determines that a certain area is about to experience meteorological conditions that are conducive to artificial cloud dispersing and rain reduction operations (such as moderate cloud thickness and sufficient water vapor in the clouds), and the first cloud cluster dynamic analysis viewpoint shows that the clouds in the area have a trend of rapid development, then the system will generate an instruction message, recommending artificial cloud dispersing and rain reduction operations for linked operation tasks in the area, and give specific operation time and operation method recommendations.

[0058] It can be seen that the application of the early warning analysis system in artificial cloud dissipation and rain reduction operations is a complex and delicate process. It provides scientific and accurate operation instruction information for meteorological departments and related operation units through comprehensive analysis of multi-dimensional meteorological data, complex algorithm processing and application of professional meteorological knowledge. In this way, it not only improves the efficiency and effect of artificial cloud dissipation and rain reduction operations, but also provides strong technical support for ensuring agricultural production and optimizing water resource allocation.

[0059] In another possible application scenario, combined with the above steps, the embodiment of the present application proposes a new full-time artificial cloud dissipation and rain reduction operation early warning and analysis method, which deeply integrates multi-dimensional meteorological element data, advanced data processing technology, comprehensive early warning indicator analysis model and full-time management strategy, aiming to provide scientific and accurate early warning and guidance for artificial cloud dissipation and rain reduction operations.

[0060] The early warning analysis system first uses diversified meteorological observation methods such as ground-based weather radars, meteorological satellites, ground meteorological stations, and atmospheric composition monitoring equipment to comprehensively collect data on various meteorological elements, including wind speed, wind direction, temperature, humidity, air pressure, cloud distribution, and the internal physical structure of clouds. These data undergo strict quality control and preprocessing processes, such as filtering and denoising, format unification, and data fusion, to ensure the accuracy and reliability of the data, providing a solid foundation for subsequent comprehensive early warning analysis.

[0061] After obtaining accurate and reliable multi-dimensional meteorological element data, the early warning analysis system can conduct in-depth comprehensive early warning analysis and processing on these data. This processing process integrates multiple methods such as physical models, statistical models and machine learning models to dynamically analyze the generation, development, maturity and dissipation stages of clouds, and output accurate cloud dissipation and rain reduction operation timing and area. At the same time, the system will also generate multiple meteorological monitoring characterization vectors, and assign corresponding early warning operation event label distribution data, early warning operation event label discrimination weights, first cloud group dynamic analysis viewpoints and second cloud group dynamic analysis viewpoints to these vectors, so as to fully reflect the relationship between meteorological conditions and early warning operation events.

[0062] After completing the comprehensive warning analysis and processing, the warning analysis system can accurately determine the target warning operation event label corresponding to each strategy mode type from the warning operation event labels corresponding to multiple meteorological monitoring characterization vectors based on the warning operation event label distribution data, warning operation event label discrimination weight, the first cloud group dynamic analysis point of view and the second cloud group dynamic analysis point of view. This process involves complex decision logic and algorithm optimization to ensure that the final target warning operation event label is both accurate and reliable.

[0063] Finally, the early warning analysis system can generate detailed artificial cloud dispersal and rain reduction instruction information based on the warning operation event label distribution data of the target warning operation event label, the warning operation event label discrimination weight, the target cloud trend type and the strategy mode type. This information includes all the key information required for artificial cloud dispersal and rain reduction operations, such as operation time, operation location, operation method (linked or non-linked), expected effect, etc., which can provide clear guidance for meteorological departments or relevant operation units and help them formulate scientific and reasonable artificial cloud dispersal and rain reduction operation plans.

[0064] It can be seen that the full-time artificial cloud dispersing and rain reduction operation early warning and analysis technical solution provided in the embodiment of the present application realizes accurate early warning and scientific guidance of artificial cloud dispersing and rain reduction operations through deep integration of multi-dimensional meteorological element data, advanced data processing technology, comprehensive early warning indicator analysis model and full-time management strategy. In this way, it not only improves the efficiency and effectiveness of artificial cloud dispersing and rain reduction operations, but also provides strong technical support for ensuring agricultural production and optimizing water resources allocation.

[0065] It is worth mentioning that the core invention of the embodiment of the present application lies in its unique comprehensive early warning analysis and processing method, which can deeply analyze and interpret multi-dimensional meteorological element data and provide accurate early warning and guidance for artificial cloud dissipation and rain reduction operations. The following is a detailed introduction to the core invention:

[0066] 1. Core Process of Comprehensive Early Warning, Analysis and Processing

[0067] Data integration and preprocessing: The early warning analysis system first collects data on various meteorological elements including wind speed, wind direction, temperature, humidity, air pressure, cloud distribution, and the physical structure inside the clouds. These data undergo quality control and preprocessing, such as filtering and denoising, format unification, and data fusion, to ensure the accuracy and reliability of the data.

[0068] Construction of a comprehensive early warning indicator model: Combining physical models, statistical models and machine learning models, a comprehensive early warning indicator model is constructed. This model dynamically analyzes the formation, development, maturity and dissipation stages of clouds.

[0069] Warning operation event label and weight generation: Assign the corresponding warning operation event label distribution data to each meteorological monitoring characterization vector. Calculate the warning operation event label discrimination weight to represent the probability value of the warning operation event label corresponding to the meteorological monitoring characterization vector being the target warning operation event label.

[0070] Cloud cluster dynamic analysis viewpoint generation: Generate the first cloud cluster dynamic analysis viewpoint, which represents the probability value of the target early warning operation event belonging to the target cloud cluster trend type, or the probability value of the target early warning operation event's strategy pattern belonging to the target linkage operation task's strategy pattern type. Generate the second cloud cluster dynamic analysis viewpoint, which represents the probability value of the target early warning operation event's strategy pattern belonging to the target non-linkage operation task's strategy pattern type.

[0071] 2. Technological innovation and advantages of core inventions

[0072] Deep fusion of multi-dimensional meteorological element data: The embodiment of the present application realizes comprehensive and accurate perception of meteorological conditions by collecting and processing multi-dimensional meteorological element data. The deep fusion of data provides a solid foundation for subsequent early warning research and judgment.

[0073] Construction of a comprehensive early warning indicator model: Combining physical models, statistical models and machine learning models, the embodiment of the present application constructs a comprehensive early warning indicator model with powerful analytical capabilities. The model can dynamically analyze each stage of the cloud cluster and output accurate cloud dissipation and rain reduction operation timing and area.

[0074] Innovative application of warning operation event labels and weights: The embodiment of the present application introduces the concept of warning operation event labels and weights, and provides a clear warning operation event label distribution and probability value for each meteorological monitoring characterization vector, which enables the warning system to more accurately judge the relationship between meteorological conditions and warning operation events.

[0075] Unique generation of cloud cluster dynamic analysis viewpoints: The embodiment of the present application innovatively generates a first cloud cluster dynamic analysis viewpoint and a second cloud cluster dynamic analysis viewpoint. These two viewpoints respectively characterize the probability values ​​of the target warning operation event belonging to different cloud cluster trend types or strategy pattern types, which provides strong support for formulating more scientific and reasonable cloud dissipation and rain reduction operation strategies.

[0076] In this way, through the deep integration of multi-dimensional meteorological element data, the construction of a comprehensive early warning indicator analysis model, the innovative application of early warning operation event labels and weights, and the unique generation of cloud group dynamic analysis viewpoints and other technical means, accurate early warning and scientific guidance of artificial cloud dissipation and rain reduction operations have been achieved. This not only improves the efficiency and effectiveness of the operations, but also provides strong technical support for the development of related fields.

[0077] Next, the core invention points of the embodiments of the present application will be further introduced, especially the part about how to determine the target warning operation event label and how to generate artificial cloud dissipation and rain reduction indication information.

[0078] 1. Determine the target warning operation event label

[0079] Data integration and analysis: The early warning analysis system first integrates the early warning operation event label distribution data, the early warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint and the second cloud group dynamic analysis viewpoint obtained in step 120. These data together constitute a comprehensive interpretation of the multi-dimensional meteorological element data.

[0080] Identification of strategy pattern types: The system identifies different strategy pattern types, which represent different cloud and rain reduction operation strategies. Each strategy pattern type corresponds to a specific set of operation methods and processes.

[0081] Determination of target warning operation event labels: Based on the warning operation event label distribution data and the warning operation event label discrimination weights, the system determines the most likely warning operation event labels for each strategy mode type. These labels represent the meteorological conditions or operation events that require the most attention and response under this strategy mode type.

[0082] 2. Generate artificial cloud dissipation and rain reduction instruction information

[0083] Composition of instruction information: Artificial cloud dissipation and rain reduction instruction information consists of multiple key parts, including warning operation event labels, warning operation event label discrimination weights, target cloud trend types and strategy mode types. These parts together constitute comprehensive guidance for cloud dissipation and rain reduction operations.

[0084] Generation of instruction information: The system determines the operation area and operation timing that need the most attention based on the warning operation event label distribution data and warning operation event label discrimination weight of the target warning operation event label. Combined with the target cloud trend type and strategy mode type, the system generates specific artificial cloud dissipation and rain reduction operation instructions, including operation methods, operation intensity, operation time, etc.

[0085] Output and application of instruction information: The generated artificial cloud dispersal and rain reduction instruction information is output to the operators or relevant decision-making departments in an intuitive and easy-to-understand manner. This information provides scientific operation guidance for the operators, helping them to carry out effective cloud dispersal and rain reduction operations at the appropriate time and area.

[0086] 3. Technological innovation and advantages of core inventions

[0087] Accurate determination of target warning operation event labels: By comprehensively considering factors such as multi-dimensional meteorological element data, warning operation event label distribution data and discrimination weights, the embodiment of the present application can accurately determine the target warning operation event label corresponding to each strategy mode type, which provides a clear guiding direction for subsequent cloud dissipation and rain reduction operations.

[0088] Comprehensive generation of artificial cloud dissipation and rain reduction instruction information: The artificial cloud dissipation and rain reduction instruction information generated by the embodiment of the present application covers multiple aspects such as the operation area, operation timing, operation method and operation intensity. The comprehensiveness of this information ensures that the operators can perform scientific and effective operations according to the instructions.

[0089] Improve operation efficiency and effectiveness: Through accurate target warning operation event label determination and comprehensive artificial cloud dispersal and rain reduction instruction information generation, the embodiment of the present application significantly improves the efficiency and effectiveness of cloud dispersal and rain reduction operations, which not only helps to reduce the impact of meteorological disasters on human society, but also provides strong support for scientific research and technological development in related fields.

[0090] It can be understood that the core invention of the embodiment of the present application lies in its unique method for determining target warning operation event labels and comprehensive artificial cloud dispersal and rain reduction indication information generation technology. These technological innovations not only improve the efficiency and effectiveness of cloud dispersal and rain reduction operations.

[0091] In the embodiment of the present application, the meteorological monitoring characterization vector, as an important technical feature of the embodiment of the present application, carries the essence of multi-dimensional meteorological element data and provides a comprehensive and accurate description of meteorological conditions for the early warning analysis system.

[0092] 1. Composition of meteorological monitoring representation vector

[0093] The meteorological monitoring characterization vector is composed of multiple numerical features, each of which corresponds to a meteorological element, such as wind speed, wind direction, temperature, humidity, air pressure, etc. These numerical features together constitute a comprehensive description of the meteorological conditions.

[0094] Taking wind speed as an example, its numerical characteristics may include multiple dimensions such as instantaneous wind speed, average wind speed, maximum wind speed, etc. Each dimension uses a specific value to represent different aspects of wind speed. Similarly, meteorological elements such as temperature, humidity, and air pressure also have their own corresponding numerical characteristics.

[0095] In addition to basic meteorological elements, meteorological monitoring characterization vectors may also include some more complex features, such as the physical structure of clouds, the dynamic trends of cloud clusters, etc. These features are obtained through advanced observation technology and data processing methods, providing richer and more in-depth meteorological information for the early warning analysis system.

[0096] 2. Numerical eigenvector example [

[0098] Wind speed: [Instantaneous wind speed: 10m / s, average wind speed: 8m / s, maximum wind speed: 15m / s],

[0099] Wind direction: [Dominant wind direction: northeast wind, wind direction change range: 30°],

[0100] Temperature: [Current temperature: 25℃, Maximum temperature: 30℃, Minimum temperature: 20℃],

[0101] Humidity: [Current humidity: 60%, Maximum humidity: 80%, Minimum humidity: 40%],

[0102] Air pressure: [Current air pressure: 1013hPa, Air pressure change trend: stable],

[0103] Cloud physical structure: [cloud thickness: 500m, temperature gradient inside the cloud: 0.5℃ / 100m, water content of the cloud: medium], cloud dynamic trend: [cloud movement direction: westward, cloud development speed: slow, cloud dissipation trend: no obvious signs of dissipation] ]

[0105] In the above example, the meteorological monitoring characterization vector includes the numerical characteristics of multiple basic meteorological elements such as wind speed, wind direction, temperature, humidity, and air pressure. At the same time, it also includes more complex meteorological characteristics such as the physical structure of clouds and the dynamic trend of clouds.

[0106] The above numerical features together constitute a comprehensive and accurate description of the current meteorological conditions, providing rich data support for the early warning analysis system. The early warning analysis system can conduct accurate comprehensive early warning analysis and processing based on these numerical features, combined with information such as the distribution data of early warning operation event labels and the discrimination weights of early warning operation event labels.

[0107] 3. Application of meteorological monitoring characterization vector

[0108] Meteorological monitoring characterization vectors play a vital role in the early warning analysis system. They are not only the basic data for the system to conduct comprehensive early warning analysis and processing, but also an important basis for generating artificial cloud dissipation and rain reduction instruction information.

[0109] By deeply analyzing the numerical features in the meteorological monitoring representation vector, the early warning analysis system can accurately judge the current meteorological conditions and identify possible meteorological disaster risks. At the same time, the system can also predict future meteorological conditions based on the changing trend of numerical features, providing strong support for the formulation of scientific cloud dissipation and rain reduction operation strategies.

[0110] Therefore, the meteorological monitoring characterization vector, as one of the core technical features of the embodiment of the present application, provides strong data support for the early warning analysis system with its comprehensive and accurate data description capabilities, which not only improves the efficiency and effectiveness of cloud dispelling and rain reduction operations, but also brings new opportunities and challenges to research and development in related fields.

[0111] The embodiment of the present application innovatively realizes a comprehensive and in-depth analysis of meteorological conditions by acquiring multi-dimensional meteorological element data to be analyzed and conducting comprehensive early warning research and judgment processing on it. In detail, not only the early warning operation event label distribution data corresponding to each meteorological monitoring characterization vector is obtained, but also the early warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint and the second cloud group dynamic analysis viewpoint are further introduced, so as to more accurately identify the target early warning operation event. Then, based on this rich information, the target early warning operation event label corresponding to each strategy mode type is accurately determined, and finally scientific and effective artificial cloud dissipation and rain reduction instruction information is generated, which significantly improves the efficiency and effect of cloud dissipation and rain reduction operations, and provides a new solution for reducing the impact of meteorological disasters on human society.

[0112] In some optional embodiments, the comprehensive early warning analysis and processing of the multidimensional meteorological element data to be analyzed to obtain early warning operation event label distribution data, early warning operation event label discrimination weights, first cloud group dynamic analysis viewpoints and second cloud group dynamic analysis viewpoints corresponding to each of the multiple meteorological monitoring characterization vectors of the multidimensional meteorological element data to be analyzed include: performing cross-meteorological element vector extraction on the multidimensional meteorological element data to be analyzed to obtain a cross-meteorological element vector set; performing AI model-based early warning identification on the cross-meteorological element vector set to obtain early warning operation event label distribution data, early warning operation event label discrimination weights, first cloud group dynamic analysis viewpoints and second cloud group dynamic analysis viewpoints corresponding to each of the multiple meteorological monitoring characterization vectors, the first cloud group dynamic analysis viewpoint including that the target early warning operation event belongs to multiple operation project related The discrimination weights of keywords and the strategy pattern of the target early warning operation event respectively belong to the discrimination weights of the strategy pattern types of multiple linkage operation tasks. The second cloud cluster dynamic analysis viewpoint includes the discrimination weights of the strategy pattern of the target early warning operation event respectively belonging to the strategy pattern types of multiple non-linked operation tasks; according to the first cloud cluster dynamic analysis viewpoint, the target cloud cluster trend type or the target linkage operation task is determined from the multiple operation project keywords and the multiple linkage operation tasks, and the discrimination weight of the target cloud cluster trend type or the target linkage operation task is determined as the first cloud cluster dynamic analysis viewpoint; according to the second cloud cluster dynamic analysis viewpoint, the target non-linked operation task is determined from the multiple non-linked operation tasks, and the discrimination weight corresponding to the target non-linked operation task is determined as the second cloud cluster dynamic analysis viewpoint.

[0113] It can be understood that this optional embodiment focuses on deep mining and intelligent analysis of multi-dimensional meteorological element data, aiming to provide accurate and scientific operational guidance for artificial cloud dissipation and rain reduction operations.

[0114] After obtaining the multidimensional meteorological element data to be analyzed in depth, the artificial intelligence processing system can start the comprehensive warning analysis and processing process. The first step in this process is to extract the cross-meteorological element vectors of the multidimensional meteorological element data. The so-called cross-meteorological element vector refers to the combination and transformation of different meteorological element data to generate a new vector that can fully reflect the comprehensive characteristics of meteorological conditions. For example, the system may combine wind speed and wind direction into a wind vector, and combine temperature and humidity into a temperature and humidity vector. Through such extraction operations, the artificial intelligence processing system can obtain a cross-meteorological element vector set containing various comprehensive characteristics of the multidimensional meteorological element data to be analyzed.

[0115] Next, the AI ​​processing system can perform early warning identification based on the advanced AI model for this cross-meteorological element vector set. The AI ​​model is the core algorithm of the AI ​​processing system. After learning and training on massive historical meteorological data and operational event data, it has a deep understanding of the intrinsic connection between meteorological conditions and operational events. When a new cross-meteorological element vector set is input into the AI ​​model, the model can quickly identify the warning operational event label distribution data, warning operational event label discrimination weights, first cloud group dynamic analysis viewpoints, and second cloud group dynamic analysis viewpoints corresponding to these vectors.

[0116] Among them, the warning operation event label distribution data reveals the possibility of different warning operation events under current meteorological conditions. The warning operation event label discrimination weight further quantifies these possibilities, which indicates the probability that the warning operation event label corresponding to the meteorological monitoring representation vector is the target warning operation event label. Such a weight setting helps the artificial intelligence processing system to more accurately lock the target warning operation event in subsequent processing.

[0117] The first cloud cluster dynamic analysis viewpoint and the second cloud cluster dynamic analysis viewpoint are in-depth analyses of cloud cluster dynamic trends and operation strategy patterns by the artificial intelligence processing system. The first cloud cluster dynamic analysis viewpoint covers the discrimination weights of target warning operation events belonging to various operation project keywords, and the discrimination weights of target warning operation events' strategy pattern types belonging to various linkage operation tasks. This means that the artificial intelligence processing system not only pays attention to the current meteorological conditions, but also further predicts the operation events and operation strategies that may be triggered by these conditions. For example, if the current meteorological conditions indicate the appearance of rain clouds, then the first cloud cluster dynamic analysis viewpoint will give the probability of a rain operation event that may be triggered by this cloud cluster, as well as the probability of a linkage operation strategy that may be adopted by this operation event.

[0118] The second cloud cluster dynamic analysis perspective focuses on the discrimination weight of the strategy mode type of non-linked operation tasks. In artificial cloud dissipation and rain reduction operations, not all tasks need to be executed in a linked manner. Some tasks can be carried out independently, while others need to be decided whether to be executed based on the dynamic trend of the cloud cluster and other meteorological conditions. The second cloud cluster dynamic analysis perspective is the result of a probabilistic analysis of the strategy mode of such non-linked operation tasks.

[0119] After obtaining these rich analysis viewpoints, the artificial intelligence processing system can determine the target cloud group trend type or target linkage operation task from multiple operation project keywords and multiple linkage operation tasks based on the first cloud group dynamic analysis viewpoint. This determination process is based on the comparison and optimization of probability values. The artificial intelligence processing system can select the operation project keyword or linkage operation task with the highest probability value as the target cloud group trend type or target linkage operation task, and determine its corresponding discrimination weight as the final result of the first cloud group dynamic analysis viewpoint.

[0120] Similarly, the artificial intelligence processing system will determine the target non-linked operation task from a variety of non-linked operation tasks based on the second cloud group dynamic analysis point of view, and determine its corresponding judgment weight as the final result of the second cloud group dynamic analysis point of view.

[0121] After completing all the above steps, the artificial intelligence processing system has conducted a comprehensive and in-depth analysis of the current meteorological conditions. Next, the system can generate artificial cloud dissipation and rain reduction instruction information for the multi-dimensional meteorological element data to be analyzed based on these analysis results. This instruction information contains key information such as the warning operation event label distribution data of the target warning operation event label, the warning operation event label discrimination weight, the target cloud group trend type and the strategy mode type. This information provides scientific operation guidance for operators, helping them to carry out effective cloud dissipation and rain reduction operations at the best time and area.

[0122] In this way, the artificial intelligence processing system has achieved a comprehensive and in-depth analysis of meteorological conditions by introducing innovative technologies such as cross-meteorological element vector extraction and early warning identification based on AI models, and has generated accurate artificial cloud dissipation and rain reduction instruction information. In this way, the efficiency and effectiveness of cloud dissipation and rain reduction operations have been greatly improved, and a new solution has been provided for reducing the impact of meteorological disasters on human society.

[0123] Under an exemplary design idea, the target warning operation event label corresponding to each strategy pattern type is determined from the warning operation event labels corresponding to the multiple meteorological monitoring characterization vectors based on the warning operation event label distribution data, the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint and the second cloud group dynamic analysis viewpoint, including: determining multiple preliminary screening operation event labels and strategy pattern types corresponding to the multiple preliminary screening operation event labels from the warning operation event labels corresponding to the multiple meteorological monitoring characterization vectors based on the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint and the second cloud group dynamic analysis viewpoint; optimizing the preliminary screening operation event labels corresponding to each strategy pattern type based on the warning operation event label distribution data and the original cloud group dynamic analysis viewpoint to obtain the target warning operation event label corresponding to each strategy pattern type.

[0124] When elaborating on this exemplary design idea, the operating mechanism and implementation process of the design idea can be fully demonstrated in combination with the actual application scenarios of the artificial intelligence processing system.

[0125] 1. The artificial intelligence processing system is a comprehensive platform that integrates advanced algorithms, big data analysis capabilities and machine learning technologies. It aims to extract valuable information from massive data and make intelligent decisions or predictions based on it. In the field of meteorological monitoring and early warning, the artificial intelligence processing system analyzes the data collected by meteorological monitoring equipment, identifies potential meteorological disaster risks, and issues early warnings in a timely manner to protect public safety and production activities.

[0126] 2. Detailed explanation of design ideas

[0127] Data preparation and preprocessing: Before determining the labels of early warning events, the AI ​​processing system first needs to collect and organize the representation vector data of multiple meteorological monitoring stations. These data include but are not limited to meteorological elements such as temperature, humidity, air pressure, wind speed, wind direction, etc., which form the basis of system analysis.

[0128] Construction of warning operation event label distribution data: Based on historical meteorological data and corresponding disaster event records, the system constructs a warning operation event label distribution database. This database reflects the probability of occurrence of various warning events (such as heavy rain, typhoon, lightning, etc.) under different meteorological conditions, providing a statistical basis for subsequent label discrimination.

[0129] Introduction of cloud cluster dynamic analysis perspective: In order to more accurately judge the type of meteorological disaster that is about to occur, the system introduces the concept of "cloud cluster dynamic analysis perspective". It is divided into "first cloud cluster dynamic analysis perspective" and "second cloud cluster dynamic analysis perspective", which represent different angles of prediction of cloud cluster development trends. The first perspective focuses on the changes in the physical properties of clouds (such as temperature and humidity structure), while the second perspective considers more the kinematic characteristics of clouds (such as moving speed and direction).

[0130] Setting of discrimination weights for warning operation event labels: Based on past experience and learning results, the system sets discrimination weights for different types of warning operation event labels. These weights reflect the possibility and urgency of various types of warning events under specific meteorological conditions, and are an important basis for the system to make final judgments.

[0131] Determination of preliminary screening operation event labels: Next, the artificial intelligence processing system conducts a comprehensive analysis of multiple meteorological monitoring representation vectors based on the early warning operation event label discrimination weights, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint. The system calculates the score of each label and considers the weight influence to select several labels with higher scores from all possible early warning operation event labels as preliminary screening operation event labels. At the same time, the system will also associate these preliminary screening labels with the types of strategy patterns that they are most likely to correspond to.

[0132] Optimization of initial screening labels: In order to further improve the accuracy and pertinence of early warning, the system can optimize the initial screening event labels under each strategy mode type. This step is mainly based on the early warning operation event label distribution data and the original cloud group dynamic analysis point of view. The system corrects or confirms the initial screening labels by comparing them with actual events under similar meteorological conditions in the historical database, and considering the latest changes in cloud group dynamics, and finally determines the target early warning operation event labels corresponding to each strategy mode type.

[0133] Output and application of target warning operation event labels: After the above steps, the artificial intelligence processing system obtains target warning operation event labels for the current meteorological conditions and transmits this information to relevant departments and the public in real time. These labels not only indicate the types of disasters that may occur, but also imply recommended response strategy models, providing a scientific basis for disaster prevention and mitigation work.

[0134] In this way, first of all, the artificial intelligence processing system significantly improves the accuracy and timeliness of warning operation event labels by comprehensively considering various meteorological monitoring data, historical disaster records and multi-dimensional analysis of cloud group dynamics, providing more reliable protection for public safety and production activities. Secondly, the artificial intelligence processing system can output targeted warning information according to the types of strategy patterns under different meteorological conditions, which not only enhances the practicality of the warning, but also enables relevant departments to formulate and implement response measures more efficiently. In addition, through continuous learning and optimization, the artificial intelligence processing system can continuously adapt to changes in meteorological conditions, improve the generalization ability and long-term stability of the warning model, and provide strong technical support for the prevention and mitigation of meteorological disasters.

[0135] In a replaceable technical solution, the method of determining multiple preliminary screening operation event labels and the strategy pattern types corresponding to the multiple preliminary screening operation event labels from the warning operation event labels corresponding to the multiple meteorological monitoring characterization vectors based on the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint and the second cloud group dynamic analysis viewpoint includes: grouping the warning operation event label discrimination weight and the first cloud group dynamic analysis viewpoint of each warning operation event label respectively to obtain the original cloud group dynamic analysis viewpoint corresponding to each of the warning operation event labels; determining the warning operation event label corresponding to the original cloud group dynamic analysis viewpoint that meets the viewpoint discrimination requirements as the preliminary screening operation event label; if the first cloud group dynamic analysis viewpoint reflects that the strategy pattern of the target warning operation event is the target linkage operation task, determining the strategy pattern type of the target linkage operation task as the strategy pattern type corresponding to the preliminary screening operation event label.

[0136] Prior to this, we have introduced how the artificial intelligence processing system determines the warning operation event label and its corresponding strategy mode type from multiple meteorological monitoring representation vectors based on multiple factors such as the warning operation event label judgment weight and cloud dynamic analysis perspective.

[0137] In this technical solution, the artificial intelligence processing system first performs clustering on the warning operation event label discrimination weights and the first cloud cluster dynamic analysis viewpoints of each warning operation event label. The purpose of this step is to classify warning operation event labels with similar weights and dynamic analysis viewpoints into one category for subsequent unified processing and analysis. Through clustering, the system can obtain the original cloud cluster dynamic analysis viewpoints corresponding to each warning operation event label, which reflect the dynamic change characteristics of the cloud cluster under different meteorological conditions.

[0138] Next, the system needs to select the preliminary screening operation event labels from the early warning operation event labels after clustering according to certain discrimination requirements. These discrimination requirements may include the weight threshold of the label, the matching degree between the cloud cluster dynamic analysis viewpoint and the current meteorological conditions, etc. Through comparison and analysis, the system can determine the early warning operation event labels that meet the requirements as the preliminary screening operation event labels. These labels represent the most likely early warning events under the current meteorological conditions.

[0139] After determining the labels of the initial screening operation events, the artificial intelligence processing system needs to further determine the types of strategy patterns corresponding to these labels. The implementation of this step depends on an in-depth interpretation of the dynamic analysis viewpoints of the first cloud group. If the dynamic analysis viewpoint of the first cloud group reflects that the strategy pattern of the target warning operation event is a target linkage operation task, then the system will determine the strategy pattern type of the target linkage operation task as the strategy pattern type corresponding to the labels of the initial screening operation events. This means that the system not only identifies possible warning events, but also matches them with corresponding response strategies.

[0140] After completing the initial screening operation event labels and determining the types of strategy patterns, the artificial intelligence processing system can conduct a comprehensive analysis of this information. The system can further verify and adjust the initial screening labels by considering multiple factors such as current meteorological conditions, historical disaster records, and dynamic trends of cloud clusters. Ultimately, the system can output optimized warning operation event labels and their corresponding strategy pattern types, providing accurate and timely warning information to relevant departments and the public.

[0141] For example, the current meteorological monitoring equipment collects a set of characterization vector data, including meteorological elements such as temperature, humidity, and air pressure. The artificial intelligence processing system will first assign weights to each warning operation event label based on these data and the warning operation event label discrimination weight database. Next, the system can consider the first cloud group dynamic analysis viewpoints, which may indicate that the current cloud group has certain specific kinematic or physical characteristics. Through cluster processing, the system can classify warning operation event labels with similar weights and cloud group dynamic characteristics into one category. For example, labels with high weights and cloud group dynamic analysis views showing a strong convection development trend may be classified into one category. Next, the system can filter out the initial screening operation event labels according to the discrimination requirements, such as "heavy rain warning" and "lightning warning". Further, the system can analyze the association between the first cloud group dynamic analysis viewpoint and the target warning operation event. If the viewpoint indicates that the current cloud group dynamics may cause heavy rain and lightning to occur simultaneously, and this combination corresponds to a specific linkage operation task (such as starting the drainage system and lightning protection facilities at the same time), then the system will determine this strategy mode type as "heavy rain + lightning linkage response". Finally, the AI ​​processing system can output "rainstorm warning" and "lightning warning" as target warning operation event labels, and provide suggestions for the type of strategy mode of "rainstorm + lightning joint response". Such output not only provides warning information, but also provides a scientific basis for relevant departments to formulate response measures.

[0142] This design improves the accuracy and pertinence of warning operation event labels through cluster processing and comprehensive analysis. By considering multiple factors and performing weight assignment and cloud cluster dynamic analysis, the system can more accurately identify the warning events that are most likely to occur under current meteorological conditions. Secondly, the effective matching of warning operation event labels and strategy pattern types is achieved. The system not only provides warning information, but also formulates specific response measures for relevant departments. This greatly enhances the practicality and operability of warnings, allowing relevant departments to respond to meteorological disaster risks more quickly and effectively. In addition, it helps to improve the scientificity and intelligence level of meteorological warnings. Through continuous learning and optimization, the artificial intelligence processing system can continuously adapt to changes in meteorological conditions and the development trend of disaster risks. This enables the system to maintain high stability and accuracy in long-term operation, providing strong technical support for the prevention and mitigation of meteorological disasters.

[0143] In another alternative technical solution, the method also includes: if the first cloud cluster dynamic analysis viewpoint reflects that the target warning operation event is a target cloud cluster trend type, based on the second cloud cluster dynamic analysis viewpoint, the strategy pattern type of the target non-linked operation task is determined as the strategy pattern type corresponding to the initial screening operation event label.

[0144] The above content has introduced how to determine the warning operation event label and its corresponding strategy mode type from multiple meteorological monitoring representation vectors based on multiple factors such as the warning operation event label discrimination weight and cloud cluster dynamic analysis perspective.

[0145] In this technical solution, the artificial intelligence processing system will first analyze the first cloud dynamic analysis point of view. If this point of view reflects that the target warning operation event corresponds to a specific target cloud trend type, then the system will enter the next step of the judgment process. The target cloud trend type refers to the prediction of its possible future development trend based on the dynamic characteristics of the cloud, such as strengthening, weakening, and moving direction.

[0146] When the first cloud dynamic analysis viewpoint points to a specific target cloud trend type, the artificial intelligence processing system can turn to consider the second cloud dynamic analysis viewpoint. The second cloud dynamic analysis viewpoint may contain more detailed information about the cloud's internal structure, kinematic characteristics, or interaction with other meteorological elements. This information is crucial to determining the type of strategic mode for non-coordinated operation tasks.

[0147] Based on the second cloud dynamic analysis perspective, the AI ​​processing system can begin to determine the types of strategic patterns for the target non-coordinated tasks. Non-coordinated tasks are those that do not need to be performed in coordination with other tasks, but can be completed independently. In the context of meteorological warnings, a single response strategy for a specific cloud trend can be characterized.

[0148] For example, if the second cloud cluster dynamic analysis viewpoint indicates that the target cloud cluster is rapidly intensifying and may cause heavy rainfall, the system may select "heavy rainfall response" as the strategic mode type of the target non-linked operation task. This judgment process involves an in-depth analysis of the cloud cluster's dynamic characteristics and an assessment of the potential meteorological disaster risk.

[0149] Once the strategy pattern type of the target non-linked operation task is determined, the AI ​​processing system will correspond it to the preliminary screening operation event label. The preliminary screening operation event label is selected in the previous step by considering the warning operation event label discrimination weight and the first cloud group dynamic analysis point of view. This correspondence means that for each preliminary screening operation event label, the system assigns it a specific strategy pattern type.

[0150] After the strategic model type is determined, the artificial intelligence processing system can conduct a comprehensive analysis of all information. This includes considering current meteorological conditions, historical disaster records, cloud dynamics trends, and the applicability of strategic model types. Ultimately, the system can output optimized warning operation event labels and their corresponding strategic model types to provide accurate and timely warning information to relevant departments and the public.

[0151] For example, current meteorological monitoring equipment collects a set of characterization vector data that reflects the meteorological conditions in a specific area. The artificial intelligence processing system first assigns weights to each warning operation event label based on this data and the warning operation event label discrimination weight database. At the same time, the system can consider the first cloud group dynamic analysis viewpoint, which may indicate that the current cloud has certain specific kinematic or physical characteristics. Through the previous steps, the system has screened out the initial screening operation event labels, such as "heavy rainfall warning". Now, the system needs to further determine the type of strategy pattern corresponding to this label. The system analyzes the first cloud group dynamic analysis viewpoint and finds that it reflects that the target warning operation event corresponds to the target cloud trend type of "rapid cloud strengthening". Based on this information, the system turns to consider the second cloud group dynamic analysis viewpoint. These views may provide more details about the internal structure, movement speed, direction, and interaction with other meteorological elements (such as temperature and humidity) of the cloud. By analyzing the second cloud group dynamic analysis viewpoint, the system determines that the strategy pattern type of the target non-linked operation task is "rapid response to heavy rainfall". This strategy pattern type covers all response measures for heavy rainfall events that do not need to be coordinated with other operation tasks, such as starting the drainage system and issuing public warnings. Finally, the AI ​​processing system uses "heavy rainfall warning" as the target warning operation event label and outputs a strategy model type recommendation of "fast response to heavy rainfall". Such output not only provides warning information, but also provides a scientific basis for relevant departments to formulate response measures.

[0152] In this way, by deeply analyzing the association between the cloud group dynamic analysis viewpoint and the target cloud group trend type, the accuracy and pertinence of the warning operation event label are improved. This enables the system to more accurately identify the warning events that are most likely to occur under the current meteorological conditions, and provide more specific response strategy recommendations for relevant departments. Secondly, the effective determination of the strategy mode type of non-linked operation tasks is realized. By considering the second cloud group dynamic analysis viewpoint, the system can specify a specific strategy mode type for each preliminary screening operation event label, thereby enhancing the practicality and operability of the warning. This enables relevant departments to respond to meteorological disaster risks more quickly and effectively, and take corresponding response measures. In addition, through continuous learning and optimization, the artificial intelligence processing system can continuously adapt to changes in meteorological conditions and the development trend of disaster risks. This enables the system to maintain high stability and accuracy in long-term operation, providing strong technical support for the prevention and mitigation of meteorological disasters.

[0153] In a preferred technical solution, the preliminary screening operation event labels corresponding to each strategy pattern type are optimized based on the warning operation event label distribution data and the original cloud cluster dynamic analysis viewpoint to obtain the target warning operation event label corresponding to each strategy pattern type, including: based on the warning operation event label distribution data and the original cloud cluster dynamic analysis viewpoint, the preliminary screening operation event labels corresponding to each strategy pattern type are cleaned of redundant noise to filter out redundant warning operation event labels in the preliminary screening operation event labels to obtain the target warning operation event label corresponding to each strategy pattern type.

[0154] In the aforementioned embodiments, it has been introduced how the artificial intelligence processing system determines the warning operation event label and its corresponding strategy mode type from multiple meteorological monitoring characterization vectors based on various factors such as the warning operation event label discrimination weight and cloud group dynamic analysis perspective. Furthermore, the artificial intelligence processing system needs to optimize the initial screening operation event label corresponding to each strategy mode type based on the warning operation event label distribution data and the original cloud group dynamic analysis perspective to obtain a more accurate and refined target warning operation event label. This optimization process is intended to filter out redundant warning operation event labels in the initial screening operation event labels and improve the pertinence and effectiveness of the warning.

[0155] First, the warning operation event label distribution data reflects the frequency and distribution of various warning operation event labels in a specific time period. These data are important for understanding which types of warning events are more likely to occur under current meteorological conditions. The artificial intelligence processing system can first analyze these distribution data to identify potential warning event trends.

[0156] Secondly, the original cloud dynamic analysis viewpoints are based on the real-time data collected by meteorological monitoring equipment to analyze and interpret the dynamic characteristics of clouds (such as shape, movement speed, direction, etc.). These viewpoints are crucial for judging the development trend of clouds and the types of meteorological disasters that may be caused. The system can combine these viewpoints with the warning operation event label distribution data to guide the subsequent optimization process.

[0157] Then, after obtaining the preliminary screening operation event labels corresponding to each strategy mode type, the artificial intelligence processing system can start redundant noise cleaning. This process aims to filter out those warning operation event labels that do not match the current meteorological conditions, cloud dynamic characteristics, and strategy mode types. For example, if a preliminary screening operation event label corresponds to "heavy rain warning", but the current cloud dynamic analysis view shows that the cloud is weakening, and the warning operation event label distribution data shows that the frequency of "heavy rain warning" is low, then the system may regard this label as a redundant warning operation event label and filter it.

[0158] Furthermore, after redundant noise cleaning, the artificial intelligence processing system can obtain the target warning operation event labels corresponding to each strategy mode type. These labels are more refined and accurate, and can better reflect the warning events that are most likely to occur under current meteorological conditions. For example, for the "strong wind response" strategy mode type, the system may determine "strong wind warning" and "strong wind warning" as target warning operation event labels.

[0159] Finally, after the target warning operation event label is determined, the artificial intelligence processing system can conduct a comprehensive analysis of all information. This includes considering the current meteorological conditions, historical disaster records, cloud dynamic change trends, and the applicability of the target warning operation event label. Ultimately, the system can output optimized warning operation event labels and their corresponding strategy model types to provide accurate and timely warning information to relevant departments and the public.

[0160] For example, current meteorological monitoring equipment collects a set of representation vector data that reflects the meteorological conditions in a specific area. The artificial intelligence processing system first assigns weights to each warning operation event label based on this data and the warning operation event label discrimination weight database. At the same time, the system can consider the original cloud dynamic analysis viewpoints, which may indicate that the current cloud has certain specific kinematic or physical characteristics.

[0161] Through the previous steps, the system has screened out the preliminary screening operation event labels. For example, for the "strong wind response" strategy mode type, the preliminary screening operation event labels may include "strong wind warning", "high wind warning" and "storm warning". Now, the system needs to further optimize these preliminary screening operation event labels.

[0162] The system first analyzes the distribution data of warning operation event labels and finds that "strong wind warning" and "high wind warning" appear more frequently under current meteorological conditions, while "storm warning" appears relatively less frequently. Then, the system considers the original cloud cluster dynamic analysis points, which indicate that the current cloud cluster is strengthening and may cause strong wind events.

[0163] Based on this information, the system cleans redundant noise. It believes that "Storm Warning" does not match the current cloud dynamic characteristics and warning operation event label distribution data, so it filters it out. Finally, the system determines "Strong Wind Warning" and "High Wind Warning" as the target warning operation event labels for the "Strong Wind Response" strategy mode type.

[0164] Finally, the artificial intelligence processing system outputs the type of "strong wind response" strategy mode and its corresponding target warning operation event labels "strong wind warning" and "high wind warning". Such output not only provides warning information, but also provides a scientific basis for relevant departments to formulate response measures.

[0165] This design improves the accuracy and pertinence of the target warning operation event labels by deeply analyzing the warning operation event label distribution data and the original cloud cluster dynamic analysis viewpoint. This enables the system to more accurately identify the warning events that are most likely to occur under current meteorological conditions and provide more specific response strategy recommendations to relevant departments. Through redundant noise cleaning, the system can filter out those warning operation event labels that do not match the current meteorological conditions, cloud cluster dynamic characteristics, and strategy mode types, thereby improving the practicality and operability of the warning.

[0166] Under an optional technical idea, the full-time artificial cloud dissipation and rain reduction operation early warning analysis method is implemented based on a comprehensive early warning analysis algorithm model, and the method also includes: obtaining an algorithm model learning sample set, the algorithm model learning sample set includes a multidimensional meteorological element data sample and a priori early warning operation event annotations, a priori cloud trend type annotations and a priori strategy pattern type annotations corresponding to the multidimensional meteorological element data sample; inputting the multidimensional meteorological element data samples into the joint neural network to be debugged respectively, and obtaining early warning operation event label distribution data samples, early warning operation event label discrimination weight samples, first cloud group dynamic analysis viewpoint samples and second cloud group dynamic analysis viewpoint samples corresponding to multiple meteorological monitoring characterization vectors of the multidimensional meteorological element data samples; based on the early warning operation event label distribution data samples, the early warning operation event label discrimination weight samples, the first cloud group dynamic analysis viewpoint samples According to the sample of the early warning operation event label discrimination weight corresponding to each of the multiple meteorological monitoring characterization vectors and the sample of the prior early warning operation event annotation, a first training learning error is determined; according to the sample of the early warning operation event label distribution data corresponding to the benchmark early warning operation event label, the sample of the first cloud cluster dynamic analysis viewpoint and the sample of the second cloud cluster dynamic analysis viewpoint, as well as the prior early warning operation event annotation, the prior cloud cluster trend type annotation and the prior strategic mode type annotation, a second training learning error is determined; based on the first training learning error and the second training learning error, the joint neural network to be debugged is debugged to obtain the comprehensive early warning analysis algorithm model.

[0167] It is understandable that in the field of meteorology, artificial cloud and rain reduction operations are an important technical means to adjust and change local meteorological conditions in order to reduce or avoid meteorological disasters. In order to achieve this goal, the early warning analysis system plays a vital role.

[0168] With the continuous development of meteorological science and the continuous advancement of technical means, artificial cloud dissipation and rain reduction operations have become an important means of meteorological disaster prevention and control. However, it is not easy to accurately determine when, where and how to carry out cloud dissipation and rain reduction operations. This requires in-depth analysis of multi-dimensional meteorological elements and accurate prediction of cloud dynamic trends. Therefore, the development and application of early warning analysis systems are particularly important.

[0169] The early warning analysis system needs to have strong data processing capabilities and algorithm model support to extract valuable information from massive meteorological data and accurately predict meteorological trends. At the same time, the system also needs to be able to provide scientific decision-making support for artificial cloud dissipation and rain reduction operations based on the prediction results.

[0170] Therefore, in order to realize the early warning analysis of the full-time artificial cloud dissipation and rain reduction operation, the early warning analysis system adopts a comprehensive early warning analysis algorithm model. The learning and training process of the model includes the following steps.

[0171] (1) Obtaining a set of algorithm model learning examples: The early warning analysis system first needs to obtain a set of algorithm model learning examples. This set of examples includes multi-dimensional meteorological element data samples, as well as the prior warning operation event annotations, prior cloud trend type annotations, and prior strategy mode type annotations corresponding to these data samples. These annotations are obtained through in-depth analysis of historical meteorological data and artificial cloud dissipation and rain reduction operation records, and represent the warning operation events, cloud trend judgments, and strategy mode selections that should be taken under different meteorological conditions.

[0172] (2) Inputting multi-dimensional meteorological element data samples: The obtained multi-dimensional meteorological element data samples are input into the joint neural network to be debugged. This joint neural network is the core part of the comprehensive early warning and judgment algorithm model, which can perform deep learning and analysis on the input meteorological data.

[0173] (3) Obtaining multiple meteorological monitoring representation vectors: Through the processing of the joint neural network, the multi-dimensional meteorological element data samples are converted into multiple meteorological monitoring representation vectors. These representation vectors correspond to different meteorological monitoring indicators and characteristics, such as temperature, humidity, air pressure, etc. At the same time, the system will also output the warning operation event label distribution data samples, warning operation event label discrimination weight samples, first cloud group dynamic analysis viewpoint samples and second cloud group dynamic analysis viewpoint samples corresponding to these representation vectors.

[0174] (4) Determine the benchmark warning operation event label: Next, the warning analysis system can determine the benchmark warning operation event label corresponding to each strategy mode type from the warning operation event label samples corresponding to the multiple meteorological monitoring characterization vectors based on the output warning operation event label distribution data samples, warning operation event label discrimination weight samples, first cloud group dynamic analysis viewpoint samples, and second cloud group dynamic analysis viewpoint samples. This benchmark label represents the warning operation event that is most likely to occur under the current meteorological conditions.

[0175] (5) Calculation of training learning errors: In order to optimize the performance of the joint neural network, the early warning analysis system can calculate two training learning errors. The first error is determined based on the warning operation event label discrimination weight samples and prior warning operation event annotations corresponding to each of the multiple meteorological monitoring characterization vectors. It reflects the accuracy of the model in distinguishing warning operation events. The second error is determined based on the warning operation event label distribution data samples corresponding to the benchmark warning operation event label, the first cloud group dynamic analysis viewpoint sample and the second cloud group dynamic analysis viewpoint sample, as well as the prior warning operation event annotation, the prior cloud group trend type annotation and the prior strategy mode type annotation. It reflects the accuracy of the model in the overall warning judgment.

[0176] (6) Debugging the joint neural network: Finally, the early warning analysis system can debug the joint neural network based on the two calculated training learning errors. Through continuous iteration and optimization, the joint neural network can be more accurate and efficient in processing multi-dimensional meteorological element data. Finally, the optimized comprehensive early warning analysis algorithm model is obtained.

[0177] In order to more specifically illustrate how the above technical ideas are implemented, the following is an introduction through a specific application scenario example.

[0178] Heavy rain often occurs in a certain area, which has a serious impact on local production and life. In order to reduce the impact of heavy rain disasters, an early warning analysis system is introduced to conduct early warning analysis of full-time artificial cloud dissipation and rain reduction operations.

[0179] First, the system obtains multi-dimensional meteorological data of the region in the past few years, including temperature, humidity, air pressure, wind direction and speed, etc. At the same time, it also collects the corresponding prior warning operation event annotations, prior cloud trend type annotations, and prior strategy pattern type annotations.

[0180] Then, the system inputs these data samples into the joint neural network to be debugged, and obtains multiple meteorological monitoring characterization vectors and their corresponding warning operation event label distribution data samples, warning operation event label discrimination weight samples, first cloud group dynamic analysis viewpoint samples and second cloud group dynamic analysis viewpoint samples.

[0181] Next, the system determines the benchmark warning operation event label corresponding to each strategy mode type based on these output results. For example, under the rainstorm response strategy mode type, the system determines "rainstorm warning" as the benchmark warning operation event label.

[0182] Subsequently, the system calculated two training learning errors and debugged and optimized the joint neural network based on the error results. Finally, a comprehensive early warning and judgment algorithm model for heavy rain in the region was obtained.

[0183] In actual applications, when the meteorological conditions in the area change, the early warning analysis system can obtain the latest multi-dimensional meteorological element data in real time and input it into the comprehensive early warning analysis algorithm model for analysis and prediction. If the system determines that heavy rain is about to occur and meets the conditions for artificial cloud dissipation and rain reduction operations, it will immediately issue an early warning signal and recommend relevant departments to take corresponding cloud dissipation and rain reduction measures.

[0184] In this way, the early warning analysis system can effectively provide scientific early warning analysis and decision-making support for the full-time artificial cloud dissipation and rain reduction operations. This will not only help improve the prevention and control capabilities of meteorological disasters, but also bring more stable and reliable meteorological guarantees to local production and life.

[0185] In some other exemplary embodiments, the second training learning error includes a first local training learning error, a second local training learning error and a third local training learning error. According to the warning operation event label distribution data sample corresponding to the benchmark warning operation event label, the first cloud group dynamic analysis viewpoint sample and the second cloud group dynamic analysis viewpoint sample, and the prior warning operation event annotation, the prior cloud group trend type annotation and the prior strategy mode type annotation, determining the second training learning error includes: determining the first local training learning error according to the prior warning operation event annotation and the warning operation event label distribution data corresponding to the benchmark warning operation event label, the first local training learning error characterizing the difference between the benchmark warning operation event label and the warning operation event authentication label Event label identification offset; determining the second local training learning error based on the prior cloud trend type annotation, the prior strategy pattern type annotation and the first cloud dynamic analysis viewpoint sample corresponding to the benchmark early warning operation event label, the second local training learning error represents the discrimination offset between the first cloud dynamic analysis viewpoint sample corresponding to the benchmark early warning operation event label and the strategy pattern certification type; determining the third local training learning error based on the prior strategy pattern type annotation, and the first cloud dynamic analysis viewpoint sample and the second cloud dynamic analysis viewpoint sample corresponding to the benchmark early warning operation event label, the third local training learning error represents the discrimination offset between the second cloud dynamic analysis viewpoint sample corresponding to the benchmark early warning operation event label and the strategy pattern certification type.

[0186] In an embodiment of the present application, the following is a detailed introduction to three technical solutions for determining local training learning errors (a first local training learning error, a second local training learning error, and a third local training learning error).

[0187] 1) First local training learning error

[0188] Purpose: To determine the event label identification offset between the baseline warning operation event label and the warning operation event authentication label.

[0189] Input: Prior warning job event annotations: This is the warning job event that has been annotated in the historical data, which serves as the true label. Warning job event label distribution data corresponding to the baseline warning job event label: This is the probability distribution of each warning job event label output by the model.

[0190] Processing steps: Use the cross entropy loss function (or other suitable classification loss function) to calculate the difference between the warning job event label distribution data output by the model and the prior warning job event annotation. The cross entropy loss function can measure the difference between the probability distribution output by the model and the true label, thereby reflecting the accuracy of event label recognition.

[0191] Output: The first local training learning error (numeric value), which indicates the degree of deviation of the model in the recognition of warning operation event labels. The smaller the error, the more accurate the model recognition.

[0192] 2) Second local training learning error

[0193] Purpose: To determine the discrimination offset between the first cloud group dynamic analysis viewpoint sample and the strategy mode authentication category corresponding to the benchmark warning operation event label.

[0194] Input: Prior cloud trend category annotation: This is the cloud trend category that has been annotated in the historical data, which is used as the true label. Prior strategy pattern category annotation: This is the strategy pattern category that has been annotated in the historical data, which is also used as the true label. Example of the first cloud dynamic analysis view corresponding to the benchmark warning operation event label: This is the analysis view of the cloud dynamic output by the model.

[0195] Processing steps: Use a multi-classification loss function (such as softmax cross entropy loss) to calculate the difference between the first cloud dynamic analysis opinion sample output by the model and the prior cloud trend category annotation and prior strategy pattern category annotation. By comparing the cloud dynamic analysis opinion output by the model with the actual cloud trend and strategy pattern, the accuracy of the model in distinguishing cloud dynamic trends is evaluated.

[0196] Output: The second local training learning error (numerical value), which indicates the degree of deviation of the model in distinguishing the dynamic trend of cloud clusters. The smaller the error, the more accurate the model's judgment.

[0197] 3) Third local training learning error

[0198] Purpose: To determine the discrimination offset between the second cloud group dynamic analysis viewpoint sample and the strategy mode authentication category corresponding to the benchmark warning operation event label.

[0199] Input: Prior strategy pattern category annotation: This is the strategy pattern category that has been annotated in the historical data, which serves as the true label. The first cloud cluster dynamic analysis viewpoint example and the second cloud cluster dynamic analysis viewpoint example corresponding to the benchmark warning operation event label: These are two analysis viewpoints on cloud cluster dynamics output by the model.

[0200] Processing steps: Use a multi-task learning framework to consider both the first cloud dynamic analysis opinion sample and the second cloud dynamic analysis opinion sample. For each cloud dynamic analysis opinion sample, use a multi-classification loss function (such as softmax cross entropy loss) to calculate the difference between it and the prior policy pattern category annotation. The losses of the two cloud dynamic analysis opinion samples are weighted summed to obtain the total loss.

[0201] Output: The third local training learning error (numerical value), which indicates the comprehensive deviation of the model in distinguishing the dynamic trends of the two cloud clusters. The smaller the error, the more accurate the model is in distinguishing the dynamic trends of the two cloud clusters.

[0202] In summary, the above three technical solutions for local training learning errors are optimized and designed for the recognition of warning operation event labels, the discrimination of cloud dynamic trends, and the comprehensive discrimination of two cloud dynamic trends, respectively, to improve the accuracy and reliability of the model in the warning analysis of artificial cloud dissipation and rain reduction operations.

[0203] In some independent embodiments, after generating the artificial cloud dispersing and rain reduction indication information of the multidimensional meteorological element data to be analyzed based on the warning operation event label distribution data of the target warning operation event label, the warning operation event label discrimination weight, the target cloud trend type and the strategy mode type, the method also includes: visualizing the artificial cloud dispersing and rain reduction indication information to obtain user graphical interface information corresponding to the artificial cloud dispersing and rain reduction indication information.

[0204] Specifically, the artificial cloud dissipation and rain reduction indication information is visualized to obtain user graphical interface information corresponding to the artificial cloud dissipation and rain reduction indication information, including:

[0205] (1) extracting key data elements from the artificial cloud dissipation and rain reduction instruction information, wherein the key data elements include the geographical location, range boundary, and terrain characteristics of the operation area, the specific date, time period, and duration of the operation time, as well as the quantitative indicators of the expected effect and the scope of influence;

[0206] (2) formatting the key data elements, wherein the formatting is used to achieve consistency and readability of the key data elements, and the formatting includes converting the time data into a standard time format and converting the effect data into a chart form;

[0207] (3) according to a preset graphical interface template, filling the formatted key data elements into corresponding graphical elements to obtain a preliminarily generated user graphical interface, wherein the graphical interface template includes a map display area, a timeline control area, and an effect preview area;

[0208] (4) beautifying the initially generated user graphical interface, wherein the beautification includes selecting color matching, optimizing font size and style, and adjusting layout structure;

[0209] (5) Adding multiple interactive functions to the beautified user graphical interface to obtain target user graphical interface information;

[0210] (6) Outputting the target user graphical interface information in the form of a super-resolution image, an interactive web page, or a professional software plug-in.

[0211] First, key data elements are accurately extracted from the complex artificial cloud dissipation and rain reduction instruction information. These data elements comprehensively cover the geographical location, scope boundaries, terrain features of the operation area, the specific date, time period, duration of the operation time, as well as the quantitative indicators of the expected effect and the scope of influence. Subsequently, these extracted data are strictly formatted to ensure the accuracy, consistency and readability of the data. For example, the time data is uniformly converted into a standard time format so that users can more intuitively understand the arrangement of the operation time; at the same time, the effect data is converted into an intuitive chart form so that users can more clearly understand the quantitative indicators and scope of influence of the expected effect.

[0212] Next, according to the preset graphical interface template, the formatted key data is accurately filled into the corresponding graphic elements. This set of graphical interface templates is highly customized, taking into full account the professionalism and user needs of artificial cloud and rain reduction operations, and includes multiple functional modules such as map display area, timeline control area, and effect preview area. In the map display area, a high-resolution map base map is used to clearly display the topography of the operation area, and the key areas of the operation are highlighted through color gradients, icon annotations, etc.; in the timeline control area, users can easily view and select different operation time periods; in the effect preview area, intuitive charts and simulation animations are used to show the dynamic changes of the expected effects.

[0213] In order to further enhance the user's visual experience, the initially generated user graphical interface is beautified in all aspects. This includes carefully selecting color combinations to ensure that the interface colors are both beautiful and in line with the professional atmosphere; optimizing font size and style to ensure that text information is clear and readable; and fine-tuning the layout structure to ensure that each functional module is compact and does not interfere with each other. Through these beautification measures, the user graphical interface is more in line with the user's aesthetic habits and usage habits.

[0214] In addition, a variety of interactive functions have been innovatively added to the beautified user graphical interface. For example, when the user clicks on the operation area, a detailed information window will pop up to display the detailed operation plan and historical operation records of the area; when the user drags the timeline, the effect preview area will be dynamically updated to display the expected effect changes at different time points. The addition of these interactive functions has greatly enhanced the user's sense of participation and the convenience of operation.

[0215] Finally, the fully processed and optimized user graphical interface information is output in various forms such as high-resolution images, interactive web pages or professional software plug-ins. The output user graphical interface information can be smoothly displayed and operated on a variety of devices and platforms, such as desktop computers, mobile devices, large display screens, etc., to meet the usage needs of different users in different scenarios.

[0216] Further, Figure 2 Schematic diagram of the structure of an early warning analysis system 200 provided in an embodiment of the present application. Figure 2 The early warning analysis system 200 shown includes a processor 210, and the processor 210 can call and run a computer program from a memory to implement the method in the embodiment of the present application.

[0217] Alternatively, if Figure 2 As shown, the early warning analysis 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.

[0218] The memory 230 may be a separate device independent of the processor 210 , or may be integrated into the processor 210 .

[0219] Alternatively, if Figure 2 As shown, the early warning analysis 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.

[0220] Optionally, the early warning analysis 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.

[0221] It should be understood that the processor of the embodiment of the present application may be an integrated circuit chip with signal processing capabilities.

[0222] 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.

[0223] Based on the above, a readable storage medium is provided, on which a program or instruction is stored, and when the program or instruction is executed by a processor, the steps of the above method are implemented.

[0224] 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.

[0225] 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.

[0226] 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. A full-time artificial cloud dissipation and rain reduction operation early warning analysis method, characterized in that: Applied to an early warning analysis system, the method comprises: Obtain multi-dimensional meteorological element data to be analyzed; Performing comprehensive early warning analysis and judgment processing on the multidimensional meteorological element data to be analyzed, obtaining early warning operation event label distribution data, early warning operation event label discrimination weights, a first cloud group dynamic analysis viewpoint, and a second cloud group dynamic analysis viewpoint corresponding to each of the multiple meteorological monitoring characterization vectors of the multidimensional meteorological element data to be analyzed; wherein, the target early warning operation event includes at least one full-time manual operation project that activates a non-linked operation task or a linked operation task, the early warning operation event label discrimination weight represents the probability value that the early warning operation event label corresponding to the meteorological monitoring characterization vector is the target early warning operation event label, the first cloud group dynamic analysis viewpoint represents the probability value that the target early warning operation event belongs to the target cloud group trend type or the probability value that the strategy mode of the target early warning operation event belongs to the strategy mode type of the target linked operation task, and the second cloud group dynamic analysis viewpoint represents the probability value that the strategy mode of the target early warning operation event belongs to the strategy mode type of the target non-linked operation task; Determine a target warning operation event label corresponding to each strategy mode type from the warning operation event labels corresponding to the multiple meteorological monitoring characterization vectors according to the warning operation event label distribution data, the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint; The artificial cloud dissipation and rain reduction indication information of the multidimensional meteorological element data to be analyzed is generated based on the warning operation event label distribution data of the target warning operation event label, the warning operation event label discrimination weight, the target cloud group trend type and the strategy mode type.

2. The method according to claim 1, characterized in that The comprehensive early warning analysis and judgment processing of the multi-dimensional meteorological element data to be analyzed is performed to obtain early warning operation event label distribution data, early warning operation event label discrimination weights, first cloud group dynamic analysis viewpoints and second cloud group dynamic analysis viewpoints corresponding to each of the multiple meteorological monitoring characterization vectors of the multi-dimensional meteorological element data to be analyzed, including: Extracting cross meteorological element vectors from the multidimensional meteorological element data to be analyzed to obtain a cross meteorological element vector set; Performing early warning identification based on the AI ​​model on the cross meteorological element vector set, obtaining early warning operation event label distribution data, early warning operation event label discrimination weights, a first cloud group dynamic analysis viewpoint and a second cloud group dynamic analysis viewpoint corresponding to each of the multiple meteorological monitoring characterization vectors, wherein the first cloud group dynamic analysis viewpoint includes discrimination weights that the target early warning operation event belongs to multiple operation project keywords respectively and discrimination weights that the strategy mode of the target early warning operation event belongs to multiple strategy mode types of linkage operation tasks respectively, and the second cloud group dynamic analysis viewpoint includes discrimination weights that the strategy mode of the target early warning operation event belongs to multiple strategy mode types of non-linkage operation tasks respectively; Determine the target cloud cluster trend type or the target linkage operation task from the multiple operation project keywords and the multiple linkage operation tasks according to the first cloud cluster dynamic analysis viewpoint, and determine the discrimination weight of the target cloud cluster trend type or the target linkage operation task as the first cloud cluster dynamic analysis viewpoint; The target non-linked operation task is determined from the multiple non-linked operation tasks according to the second cloud group dynamic analysis viewpoint, and the discrimination weight corresponding to the target non-linked operation task is determined as the second cloud group dynamic analysis viewpoint.

3. The method according to claim 1, characterized in that The step of determining a target warning operation event label corresponding to each strategy mode type from the warning operation event labels corresponding to the plurality of meteorological monitoring characterization vectors based on the warning operation event label distribution data, the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint includes: Determine, from the warning operation event labels corresponding to the multiple meteorological monitoring characterization vectors, multiple preliminary screening operation event labels and the types of strategy patterns corresponding to the multiple preliminary screening operation event labels according to the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint, and the second cloud group dynamic analysis viewpoint; According to the warning operation event label distribution data and the original cloud cluster dynamic analysis viewpoint, the preliminary screening operation event label corresponding to each strategy pattern type is optimized to obtain the target warning operation event label corresponding to each strategy pattern type.

4. The method according to claim 3, characterized in that The determining of a plurality of preliminary screening operation event labels and the types of strategy patterns corresponding to the plurality of preliminary screening operation event labels from the warning operation event labels corresponding to the plurality of meteorological monitoring characterization vectors based on the warning operation event label discrimination weight, the first cloud group dynamic analysis viewpoint and the second cloud group dynamic analysis viewpoint includes: The warning operation event label discrimination weights and the first cloud cluster dynamic analysis viewpoints of each warning operation event label are respectively divided into groups to obtain the original cloud cluster dynamic analysis viewpoints corresponding to each warning operation event label; Determine the early warning operation event label corresponding to the original cloud dynamic analysis viewpoint that meets the viewpoint discrimination requirements as the preliminary screening operation event label; If the first cloud group dynamic analysis viewpoint reflects that the strategy mode of the target warning operation event is the target linkage operation task, the strategy mode type of the target linkage operation task is determined as the strategy mode type corresponding to the initial screening operation event label.

5. The method according to claim 4, characterized in that The method further comprises: If the first cloud cluster dynamic analysis viewpoint reflects that the target warning operation event is a target cloud cluster trend type, based on the second cloud cluster dynamic analysis viewpoint, the strategy mode type of the target non-linked operation task is determined as the strategy mode type corresponding to the initial screening operation event label.

6. The method according to claim 4, characterized in that According to the warning operation event label distribution data and the original cloud group dynamic analysis viewpoint, the preliminary screening operation event label corresponding to each strategy mode type is optimized respectively, and the target warning operation event label corresponding to each strategy mode type is obtained, including: According to the warning operation event label distribution data and the original cloud cluster dynamic analysis viewpoint, redundant noise cleaning is performed on each preliminary screening operation event label corresponding to each strategy pattern type to filter out redundant warning operation event labels in each preliminary screening operation event label, and obtain the target warning operation event label corresponding to each strategy pattern type.

7. The method according to any one of claims 1 to 6, characterized in that The full-time artificial cloud dissipation and rain reduction operation early warning analysis method is implemented based on a comprehensive early warning analysis algorithm model, and the method also includes: Acquire an algorithm model learning sample set, wherein the algorithm model learning sample set includes a multidimensional meteorological element data sample and a priori warning operation event annotation, a priori cloud trend type annotation, and a priori strategy mode type annotation corresponding to the multidimensional meteorological element data sample; Input the multi-dimensional meteorological element data samples into the joint neural network to be debugged respectively, and obtain early warning operation event label distribution data samples, early warning operation event label discrimination weight samples, first cloud group dynamic analysis viewpoint samples, and second cloud group dynamic analysis viewpoint samples corresponding to multiple meteorological monitoring characterization vectors of the multi-dimensional meteorological element data samples; Determine a benchmark warning operation event label corresponding to each strategy mode type from the warning operation event label samples corresponding to the multiple meteorological monitoring characterization vectors according to the warning operation event label distribution data sample, the warning operation event label discrimination weight sample, the first cloud group dynamic analysis viewpoint sample, and the second cloud group dynamic analysis viewpoint sample; Determine a first training learning error according to the warning operation event label discrimination weight samples corresponding to each of the plurality of meteorological monitoring characterization vectors and the priori warning operation event annotations; Determine a second training learning error according to the warning operation event label distribution data sample corresponding to the benchmark warning operation event label, the first cloud group dynamic analysis viewpoint sample and the second cloud group dynamic analysis viewpoint sample, and the prior warning operation event annotation, the prior cloud group trend type annotation and the prior strategy mode type annotation; The joint neural network to be debugged is debugged based on the first training learning error and the second training learning error to obtain the comprehensive early warning analysis algorithm model.

8. The method according to claim 7, characterized in that The second training learning error includes a first local training learning error, a second local training learning error and a third local training learning error. According to the warning operation event label distribution data sample corresponding to the benchmark warning operation event label, the first cloud group dynamic analysis viewpoint sample and the second cloud group dynamic analysis viewpoint sample, and the prior warning operation event annotation, the prior cloud group trend type annotation and the prior strategy mode type annotation, determining the second training learning error includes: Determining the first local training learning error according to the warning operation event label distribution data corresponding to the prior warning operation event annotation and the benchmark warning operation event label, wherein the first local training learning error represents the event label recognition offset between the benchmark warning operation event label and the warning operation event authentication label; Determine the second local training learning error according to the prior cloud trend type annotation, the prior strategy pattern type annotation and the first cloud dynamic analysis viewpoint example corresponding to the benchmark warning operation event label, wherein the second local training learning error represents the discrimination offset between the first cloud dynamic analysis viewpoint example corresponding to the benchmark warning operation event label and the strategy pattern authentication type; Based on the prior strategy pattern type annotation, and the first cloud cluster dynamic analysis viewpoint sample and the second cloud cluster dynamic analysis viewpoint sample corresponding to the benchmark warning operation event label, the third local training learning error is determined, and the third local training learning error characterizes the discrimination offset between the second cloud cluster dynamic analysis viewpoint sample corresponding to the benchmark warning operation event label and the strategy pattern authentication type.

9. An early warning analysis 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 8.

10. A readable storage medium, characterized in that: The readable storage medium stores a program or instruction, and when the program or instruction is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images