Airport meteorological service big data management method and system based on artificial intelligence large model

Abstract

The application relates to the technical field of big data management, in particular to an airport meteorological service big data management method and system based on an artificial intelligence big model, which comprises the following steps: S1, constructing and maintaining an airport exclusive meteorological-operation memory bank; the meteorological-operation memory bank is analyzed in depth by an artificial intelligence big model, historical data are associated, and a triad memory unit of a meteorological mode-operation response-effect evaluation is stored; S2, collecting real-time multidimensional data, calling the meteorological-operation memory bank, dynamically evaluating system vulnerability through the artificial intelligence big model, identifying weak disturbance element combinations, and generating an airport operation system vulnerability heat map; and S3, generating a hierarchical collaborative decision-making plan including a trigger condition, an execution subject and an operation sequence through the artificial intelligence big model according to the airport operation system vulnerability heat map. The application can expose potential bottleneck areas and implicit risks in advance, and effectively improve operation safety margin and scheduling foresight.

Description

Technical Field

[0001] This invention relates to the field of big data management technology, and in particular to a method and system for managing big data of airport meteorological services based on artificial intelligence big data models. Background Technology

[0002] As airport operations become increasingly complex and meteorological conditions become more uncertain, traditional airport meteorological service systems face multiple challenges. On the one hand, existing systems generally rely on rule templates, static thresholds, or single-step prediction models to process meteorological information, making it difficult to effectively reflect the dynamic impact of meteorological changes on operational aspects such as flight scheduling, ground services, and resource allocation. On the other hand, there are fragmented understandings of meteorological information among different business departments, and a lack of unified risk identification standards and collaborative response mechanisms. This often leads to delayed and fragmented operational decisions, making it difficult to cope with the cascading degradation problem under weak disturbances, especially under complex meteorological conditions such as thunderstorms, strong winds, and low visibility.

[0003] Current research has attempted to improve the accuracy of weather forecasts by introducing machine learning techniques, but most of these studies are limited to single models, static regression, or estimation of short-term wind and cloud field changes. They lack integration with deeper information such as flight operation status and resource scheduling rules, making it difficult to form a generalizable decision-making knowledge system. At the same time, existing contingency plan generation processes rely heavily on manual consultation, resulting in delayed updates to response strategies, inability to cover dynamic resource constraints and multi-department coordination needs, and a lack of closed-loop optimization capabilities. Therefore, there is an urgent need for a meteorological service method with the ability to understand large models, retain experience, and generate inferences to achieve intelligent support and collaborative response optimization for airport operation systems under complex weather scenarios. Summary of the Invention

[0004] This invention provides a method and system for big data management of airport meteorological services based on an artificial intelligence big data model.

[0005] The airport meteorological service big data management method based on artificial intelligence big data model includes the following steps: S1: Construct and maintain an airport-specific meteorological-operational memory bank. The meteorological-operational memory bank uses a large artificial intelligence model to perform deep correlation analysis on historical data and stores a triplet memory unit of meteorological pattern, operational response, and effect evaluation. S2: Collect real-time multi-dimensional data, call the meteorological-operation memory library, conduct dynamic assessment of system vulnerability through artificial intelligence big data model, identify combinations of weak disturbance elements and generate airport operation system vulnerability heat map; S3: Based on the vulnerability heat map of the airport operation system, generate a hierarchical collaborative decision-making plan including triggering conditions, executing entities, and operation sequences through an artificial intelligence big data model.

[0006] Optionally, S1 specifically includes: S11: Extract and integrate historical meteorological data, flight operation data, ground support resource data, and abnormal event report data from the business systems of airport air traffic control, airlines, ground services, and meteorological departments to form a correlated time-series dataset; S12: Use a multimodal large model to perform spatiotemporal alignment and deep feature extraction on the associated time series dataset, identify and structure the causal relationship between the combination of meteorological elements and the set of triggered operational adjustment actions; S13: Based on the actual execution result data of the operational response, quantitatively evaluate the impact of the operational response on flight punctuality rate, safety indicators and resource utilization efficiency, and generate an effect evaluation; S14: The meteorological model, operational response, and effect evaluation are encapsulated into triplet memory units, vectorized, and stored in a vector database to form a meteorological-operational memory bank.

[0007] Optionally, the maintenance of the meteorological-operational memory bank includes continuously generating triplet memory units based on newly generated data, and using the multimodal large model to perform timeliness assessment, conflict resolution, and knowledge distillation on the triplet memory units, thereby dynamically optimizing the knowledge density and decision relevance of the memory bank.

[0008] Optionally, S12 specifically includes: S121: Use a multimodal large model to perform spatiotemporal alignment processing on the meteorological data, operational data and event data in the associated time series dataset, unify the timestamp format and spatial identification method, and construct a feature sequence with synchronization and consistency; S122: Based on the completion of spatiotemporal alignment, deep feature extraction is performed through a multimodal large model to identify the causal relationship between the combination of meteorological elements and the set of triggered operational adjustment actions, and the data is stored in a structured format.

[0009] Optionally, S2 specifically includes: S21: Collect real-time multi-dimensional data, perform spatiotemporal alignment and feature encoding, and generate real-time feature vectors; S22: Using the real-time feature vector as the query vector, perform a similarity search in the vector database of the meteorological-operation memory bank to obtain a set of historical triplet memory units that are most relevant to it; S23: Using the aforementioned large-scale artificial intelligence model, compare the deviation of the real-time feature vector from the meteorological model vector in the retrieved historical triplet, and combine this with the effect evaluation of the corresponding operational response to predict the chain reaction and system performance degradation path that may be triggered by the current weak disturbance element combination. S24: Based on the prediction results of the chain reaction and performance degradation path, the potential risk levels of relevant nodes are spatially gridded and visualized to generate a vulnerability heat map of the airport operation system.

[0010] Optionally, S22 specifically includes: S221: Input the real-time feature vector as the query vector into the vector database constructed by the meteorological-operational memory, and perform efficient retrieval based on vector semantic similarity; S222: Extract several historical triplet memory units that are most relevant to the current feature vector from the retrieval results.

[0011] Optionally, S23 specifically includes: S231: Using a large artificial intelligence model, the current real-time feature vector is compared with the meteorological model vector in the retrieved historical triplet to calculate the element deviation. S232: Based on the effect evaluation results of the corresponding operating response in each triplet, predict the chain reaction that may be triggered by the current combination of weak disturbance elements and the corresponding system performance degradation path.

[0012] Optionally, S3 specifically includes: S31: Analyze the vulnerability heatmap, determine the spatial grids whose risk level exceeds the preset risk level threshold and their corresponding vulnerability types, and define the location, risk level and type of the spatial grids as the triggering conditions of the contingency plan; S32: Based on the vulnerability type and the predicted chain reaction path, the AI ​​big data model is used to match and determine the set of executive entities that must coordinate to deal with risks from the predefined airport operation organizational structure and responsibility knowledge graph. The set of executive entities includes at least two departments from air traffic control, airlines, ground services and airport command center. S33: Call the meteorological-operation memory bank, retrieve historical triplet memory units similar to the current triggering conditions, extract the operation response of the historical triplet memory units as a reference scheme, and optimize and serialize the reference scheme in combination with the real-time operation status and resource constraints to generate a specific and executable operation sequence; S34: The triggering conditions, the set of executing entities, and the operation sequence are encapsulated in a structured manner, and divided into three collaborative decision-making levels: early warning, response, and emergency, according to the risk level, to form a hierarchical collaborative decision-making plan; S35: Based on the level of the contingency plan and the implementing entity, the hierarchical collaborative decision-making contingency plan is automatically distributed to the corresponding terminals through the airport collaborative decision-making system, and the contingency plan execution tracking and feedback closed loop is initiated.

[0013] Optionally, S33 specifically includes: S331: Call the meteorological-operation memory library, perform a similarity search on the current triggering condition, obtain historical triplet memory units that are similar to it, and extract the corresponding operation response from the historical triplet memory units as a reference scheme; S332: Combining the current real-time operating status and resource constraints, the reference scheme is optimized and the operation sequence is reorganized through the artificial intelligence big model to generate a specific and executable operation sequence.

[0014] The airport meteorological service big data management system based on artificial intelligence big data models is used to implement the aforementioned airport meteorological service big data management method based on artificial intelligence big data models, and includes the following modules: Meteorological-Operational Memory Construction Module: Performs in-depth correlation analysis on historical meteorological data, flight operation data and operational result data, generates a triplet memory unit of meteorological model-operational response-effect evaluation and stores it in the memory bank; Real-time situational awareness and vulnerability assessment module: Receives real-time multi-dimensional data, combines it with meteorological and operational memory, identifies combinations of weak disturbance elements through a large artificial intelligence model, and generates a vulnerability heat map of the airport operation system; Collaborative decision-making plan generation module: Analyzes the vulnerability heatmap and generates a hierarchical collaborative decision-making plan that includes triggering conditions, executing entities, and operation sequences.

[0015] The beneficial effects of this invention are: This invention constructs an airport-specific meteorological-operational memory bank through a large-scale artificial intelligence model. It stores and vectorizes historical meteorological element combinations, corresponding operational responses, and actual operational effects in a ternary format, realizing the structured accumulation and continuous evolution of operational experience. This mechanism enables the system to quickly retrieve effective historical response strategies when facing similar or approximate meteorological scenarios, and optimize them based on effect evaluation, thereby improving the airport meteorological service's intelligent decision-making and generalization adaptability under complex and variable operational conditions.

[0016] This invention analyzes the deviation between real-time features and historical meteorological patterns using a large artificial intelligence model, identifies combinations of weak disturbances that may trigger chain reactions, and predicts the resulting system performance degradation paths. It then generates a vulnerability heat map of the airport operation system in a spatial grid format. This innovation elevates risk identification from single-point meteorological indicators to a dynamic perception of the overall vulnerable structure of the airport operation system, enabling early exposure of potential bottleneck areas and hidden risks, effectively improving operational safety margins and scheduling foresight. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method according to an embodiment of the present invention; Figure 2 This is a system block diagram of an embodiment of the present invention. Detailed Implementation

[0019] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. Those skilled in the art may employ other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0020] like Figure 1 As shown, the airport meteorological service big data management method based on artificial intelligence big data model includes the following steps: S1: Construct and maintain an airport-specific meteorological-operations memory bank. The meteorological-operations memory bank uses a large artificial intelligence model to perform in-depth correlation analysis on historical data and stores triplet memory units of meteorological patterns, operational responses, and effect evaluations. S1 specifically includes: S11: Extract and integrate the following historical data from the business systems of airport air traffic control, airlines, ground services, and meteorological departments to form a correlated time-series dataset, including historical meteorological data, flight operation data, ground support resource data, and abnormal event report data, forming a unified format multi-source heterogeneous time-series dataset, denoted as: ; in, Indicates at time The Meteorological element vectors, covering common extreme weather conditions in airport operations, include wind speed. value range ,wind direction The range of values ​​is ,visibility The range of values ​​is Precipitation intensity The range of values ​​is Thunderstorm level Discrete values ; Indicates at time The set of execution response actions, Indicates the performance evaluation metrics after the operation response, describing the operation response. The actual impact on the system status after implementation, including average flight delay time. Safety risk change indicators Resource utilization efficiency , The range of values ​​is minute, The range of values ​​is Negative values ​​indicate reduced risk. The range of values ​​is , This indicates the total time duration, which is determined by the historical data duration of the business system.

[0021] S12: Utilizing large-scale multimodal artificial intelligence models on datasets Processing is performed to align different data sources in both time and space dimensions, and meteorological elements are extracted and combined. Corresponding running response The causal relationship between them can be expressed as: ; The results are stored in a structured format as meteorological model-operational response pairs. For typical meteorological element combinations identified by the large model, a traceable and generalizable mapping relationship is established between them and the operational response. For the set of execution response actions, it is the corresponding Response strategy; Operational response behaviors are extracted from a large number of similar meteorological scenarios in historical datasets, and stable, high-performing standard response patterns are extracted through causal analysis models to obtain the final set of operational response actions. Specifically, it includes: S121: In historical time-series data, find all data related to a specific weather pattern. Similar moments are represented as: ; in, The similarity function is cosine similarity. The similarity threshold ranges from 0.85 to 0.98. A higher similarity threshold results in fewer but more similar matched samples, suitable for fine-grained backtracking strategies. A lower similarity threshold results in more but lower-quality samples, which can easily introduce noisy memory units from different classes. For the set of similar sample times; S122: Collect the actual response actions at all these moments, represented as: ; in, This is the initial response action sample set; S123: Yes Perform K-means clustering analysis to remove noisy responses and retain the most common, optimal, or average-effective response strategies to obtain the final inductive response. ; S124: Introduce the DoWhy causal reasoning framework for evaluation, expressed as: ; in, To arrive at the final inductive response pattern, i.e., the final set of operational response actions, the centroids and mode responses are obtained through K-means clustering. As an effectiveness evaluation indicator, the range of values ​​for delay variation is: The safety score change ranges from [minutes] to [value]. The range of values ​​for the resource efficiency improvement ratio is: , The causal effect threshold, with a value ranging from 0.05 to 0.2, is used to filter out policy responses with ineffective or even negative effects. S13: Response-based The results after actual implementation, quantified and evaluated using a large-scale artificial intelligence model, are expressed as follows: ; in, For the effect evaluation results, The evaluation function is modeled based on causal regression. Triple data including meteorological models, corresponding operational responses, and final operational results indicators are extracted from historical data. By comparing the differences in operational effects of taking and not taking a certain response action under similar meteorological conditions, the causal relationship between meteorological models and response actions is established. S14: Encapsulate each of the above data sets into a triplet memory unit, that is: ; And then embed them into vectors using a pre-trained vector generation model: ; all Stored in a vector database, forming an airport meteorological-operations memory, represented as: ; in, The triplet memory unit facilitates storage, retrieval, and vectorization, forming an interpretable knowledge base structure. It is a vector representation of memory units, with values ​​ranging from 128 to 1024 dimensions. For airport meteorological and operational memory; S15: For newly generated data, new triples are continuously generated by performing similarity matching and causal structure verification with existing memory units, and the database content is updated using the following strategies: Timeliness assessment: Assess whether the old triplet is obsolete based on the latest operational trends; Conflict resolution: When multiple response strategies exist under the same weather pattern, the strategy is selected for retention based on its effectiveness. Knowledge distillation: This involves compressing repetitive or redundant memory units into more compact representations through a representation learning mechanism in a large model. The maintenance function can be expressed as: ; To ensure that the content of the memory always has high decision relevance and knowledge density, a maintenance function is needed. This function is used to update and optimize the vectorized memory in real time. It merges the newly generated memory units with the original memory and performs filtering, merging and compression processing on the overall content.

[0022] S2: Collect real-time multi-dimensional data, call the meteorological-operational memory bank, conduct dynamic assessment of system vulnerability through artificial intelligence big data model, identify combinations of weak disturbance elements and generate airport operation system vulnerability heat map; S2 specifically includes: S21: Obtain the current time point from various airport business systems. The real-time multi-dimensional data is used to construct a unified input vector, represented as follows: ; in, This represents the multi-source raw data collected at the current moment. Indicates the spatiotemporal alignment and encoding function. The generated real-time feature vector is a high-dimensional vector after unified spatiotemporal alignment and encoding. S22: Using real-time feature vectors To retrieve the query vector, a semantic similarity search is performed in the vectorized memory to obtain the set of historical triplet memory units most similar to the current context, represented as: ; in, For all vector representations in the memory bank, This is a Top-K retrieval function based on vector similarity, which retrieves the top vectors in the memory that are most similar to the query vector. A vector, , The most similar A set of historical triples, each including ; S23: Utilizing large-scale artificial intelligence models to analyze current real-time feature vectors With the retrieved historical meteorological model vector Performing element-level skewness comparison is represented as follows: ; Combining historical triples with corresponding response strategies Effect evaluation results A model for predicting the chain evolution of system operating states is constructed, represented as: ; in, The element skewness between the current context and historical patterns. For historical performance evaluation results, This is to infer the potential chain reaction sequence and the system performance degradation path; It is a prediction model driven by historical memory triples. It takes the real-time feature vector extracted at the current moment, the deviation information of the most similar historical meteorological pattern and its corresponding historical performance evaluation as input, and performs multi-layer correlation modeling through artificial intelligence big model to output the dynamic degradation prediction result of system performance. S24: Predict the results Mapping to the airport's operational space to construct a spatial grid The risk level value for each grid is calculated, and a vulnerability heatmap is generated, represented as follows: ; ; in, Indicates the first Grid cells, This represents the risk level of the unit at the current moment. A heatmap of vulnerability of the airport operation system after visualization rendering.

[0023] S3: Based on the vulnerability heat map of the airport operation system, generate a hierarchical collaborative decision-making plan that includes triggering conditions, implementing entities, and operation sequences through a large artificial intelligence model; S3 specifically includes: S31: Analysis of Vulnerability Heatmap To identify risk level values Exceeding the preset risk threshold All spatial grid cells , is represented as: ; For each high-risk grid Analyze its spatial location Risk level and the vulnerability types marked by the system tags. The combination of these three factors constitutes the triggering condition for the contingency plan, expressed as follows: ; in, Indicates the first The risk intensity of the grid cell, with a value range of: It is used to screen high-risk areas and serve as a risk benchmark for subsequent strategy generation. The risk level threshold is set to an integer level of 2 to balance the sensitivity of the contingency plan with response resources and prevent false alarms or delayed responses. S32: Vulnerability type extracted based on triggering conditions and predicted chain reaction pathways The knowledge graph of airport operations organization is retrieved and matched using a large artificial intelligence model. It identifies the execution responsibility nodes associated with the handling of this type of risk, and the output is a set of entities that must coordinate to respond, represented as: ; in, It must include at least two of the following departments: air traffic control, airlines, ground services, and airport command center; S33: Using the trigger condition as the query vector, call the meteorological-runtime memory, represented as: ; Extract operational response strategies from historically similar scenarios. As a reference scheme, combined with the current real-time running state vector Resource status The strategy content is optimized and serialized using a large model, outputting a structured operation sequence, represented as follows: ; in, The trigger condition for the contingency plan is a tripartite combination of location, risk level, and vulnerability type, which serves as the basis for generating the contingency plan. This serves as a reference memory unit set, providing empirical support samples for contingency plans and avoiding the need to generate strategies from scratch. For reference response strategies, This is the current real-time running state vector. For resource status, For the generated operation sequence; S34: Encapsulate the above three elements into a structured contingency plan object, represented as: ; And according to risk level They are classified and categorized into three levels, and finally stored according to the following structure: ; in, For the set of contingency plan levels, A warning level indicates that a potential risk has just emerged. The response level indicates that the operating state has deviated. An emergency level indicates a significant trend of system degradation. S35: Based on the level attributes of the contingency plan and the set of implementing entities The contingency plan is automatically distributed to corresponding terminal systems, including the control tower, operations control center, dispatching positions, and duty terminals, through the airport collaborative decision-making system. Simultaneously, a closed-loop feedback mechanism is activated to monitor the plan's execution status in real time and transmit results back, supporting automatic updates to the triplet effect evaluation units in the memory bank based on actual feedback.

[0024] like Figure 2 As shown, the airport meteorological service big data management system based on an artificial intelligence big data model is used to implement the aforementioned airport meteorological service big data management method based on an artificial intelligence big data model, and includes the following modules: Meteorological-Operational Memory Construction Module: Performs in-depth correlation analysis on historical meteorological data, flight operation data and operational result data, generates a triplet memory unit of meteorological model-operational response-effect evaluation and stores it in the memory bank; Real-time situational awareness and vulnerability assessment module: Receives real-time multi-dimensional data, combines it with meteorological and operational memory, identifies combinations of weak disturbance elements through a large artificial intelligence model, and generates a vulnerability heat map of the airport operation system; Collaborative Decision-Making Plan Generation Module: Analyzes vulnerability heatmaps to generate hierarchical collaborative decision-making plans that include triggering conditions, implementing entities, and operation sequences.

[0025] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0026] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A big data management method for airport meteorological services based on an artificial intelligence big data model, characterized in that: Includes the following steps: S1: Construct and maintain an airport-specific meteorological-operational memory bank. The meteorological-operational memory bank uses a large artificial intelligence model to perform deep correlation analysis on historical data and stores a triplet memory unit of meteorological pattern, operational response, and effect evaluation. S2: Collect real-time multi-dimensional data, call the meteorological-operation memory library, conduct dynamic assessment of system vulnerability through artificial intelligence big data model, identify combinations of weak disturbance elements and generate airport operation system vulnerability heat map; S3: Based on the vulnerability heat map of the airport operation system, generate a hierarchical collaborative decision-making plan including triggering conditions, executing entities, and operation sequences through an artificial intelligence big data model.

2. The airport meteorological service big data management method based on artificial intelligence big data model according to claim 1, characterized in that, S1 specifically includes: S11: Extract and integrate historical meteorological data, flight operation data, ground support resource data, and abnormal event report data from the business systems of airport air traffic control, airlines, ground services, and meteorological departments to form a correlated time-series dataset; S12: Use a multimodal large model to perform spatiotemporal alignment and deep feature extraction on the associated time series dataset, identify and structure the causal relationship between the combination of meteorological elements and the set of triggered operational adjustment actions; S13: Based on the actual execution result data of the operational response, quantitatively evaluate the impact of the operational response on flight punctuality rate, safety indicators and resource utilization efficiency, and generate an effect evaluation; S14: The meteorological model, operational response, and effect evaluation are encapsulated into triplet memory units, vectorized, and stored in a vector database to form a meteorological-operational memory bank.

3. The airport meteorological service big data management method based on artificial intelligence big data model according to claim 2, characterized in that, The maintenance of the meteorological-operational memory bank includes continuously generating triplet memory units based on newly generated data, and using the multimodal large model to perform timeliness evaluation, conflict resolution, and knowledge distillation on the triplet memory units, thereby dynamically optimizing the knowledge density and decision relevance of the memory bank.

4. The airport meteorological service big data management method based on artificial intelligence big data model according to claim 2, characterized in that, S12 specifically includes: S121: Use a multimodal large model to perform spatiotemporal alignment processing on the meteorological data, operational data and event data in the associated time series dataset, unify the timestamp format and spatial identification method, and construct a feature sequence with synchronization and consistency; S122: Based on the completion of spatiotemporal alignment, deep feature extraction is performed through a multimodal large model to identify the causal relationship between the combination of meteorological elements and the set of triggered operational adjustment actions, and the data is stored in a structured format.

5. The airport meteorological service big data management method based on artificial intelligence big data model according to claim 1, characterized in that, S2 specifically includes: S21: Collect real-time multi-dimensional data, perform spatiotemporal alignment and feature encoding, and generate real-time feature vectors; S22: Using the real-time feature vector as the query vector, perform a similarity search in the vector database of the meteorological-operation memory bank to obtain a set of historical triplet memory units that are most relevant to it; S23: Using the aforementioned large-scale artificial intelligence model, compare the deviation of the real-time feature vector from the meteorological model vector in the retrieved historical triplet, and combine this with the effect evaluation of the corresponding operational response to predict the chain reaction and system performance degradation path that may be triggered by the current weak disturbance element combination. S24: Based on the prediction results of the chain reaction and performance degradation path, the potential risk levels of relevant nodes are spatially gridded and visualized to generate a vulnerability heat map of the airport operation system.

6. The airport meteorological service big data management method based on an artificial intelligence big data model according to claim 5, characterized in that, S22 specifically includes: S221: Input the real-time feature vector as the query vector into the vector database constructed by the meteorological-operational memory, and perform efficient retrieval based on vector semantic similarity; S222: Extract several historical triplet memory units that are most relevant to the current feature vector from the retrieval results.

7. The airport meteorological service big data management method based on artificial intelligence big data model according to claim 5, characterized in that, S23 specifically includes: S231: Using a large artificial intelligence model, the current real-time feature vector is compared with the meteorological model vector in the retrieved historical triplet to calculate the element deviation. S232: Based on the effect evaluation results of the corresponding operating response in each triplet, predict the chain reaction that may be triggered by the current combination of weak disturbance elements and the corresponding system performance degradation path.

8. The airport meteorological service big data management method based on artificial intelligence big data model according to claim 1, characterized in that, S3 specifically includes: S31: Analyze the vulnerability heatmap, determine the spatial grids whose risk level exceeds the preset risk level threshold and their corresponding vulnerability types, and define the location, risk level and type of the spatial grids as the triggering conditions of the contingency plan; S32: Based on the vulnerability type and the predicted chain reaction path, the AI ​​big data model is used to match and determine the set of executive entities that must coordinate to deal with risks from the predefined airport operation organizational structure and responsibility knowledge graph. The set of executive entities includes at least two departments from air traffic control, airlines, ground services and airport command center. S33: Call the meteorological-operation memory bank, retrieve historical triplet memory units similar to the current triggering conditions, extract the operation response of the historical triplet memory units as a reference scheme, and optimize and serialize the reference scheme in combination with the real-time operation status and resource constraints to generate a specific and executable operation sequence; S34: The triggering conditions, the set of executing entities, and the operation sequence are encapsulated in a structured manner, and divided into three collaborative decision-making levels: early warning, response, and emergency, according to the risk level, to form a hierarchical collaborative decision-making plan; S35: Based on the level of the contingency plan and the implementing entity, the hierarchical collaborative decision-making contingency plan is automatically distributed to the corresponding terminals through the airport collaborative decision-making system, and the contingency plan execution tracking and feedback closed loop is initiated.

9. The airport meteorological service big data management method based on artificial intelligence big data model according to claim 8, characterized in that, Specifically, S33 includes: S331: Call the meteorological-operation memory library, perform a similarity search on the current triggering condition, obtain historical triplet memory units that are similar to it, and extract the corresponding operation response from the historical triplet memory units as a reference scheme; S332: Combining the current real-time operating status and resource constraints, the reference scheme is optimized and the operation sequence is reorganized through the artificial intelligence big model to generate a specific and executable operation sequence.

10. An airport meteorological service big data management system based on an artificial intelligence big data model, used to implement the airport meteorological service big data management method based on an artificial intelligence big data model as described in any one of claims 1-9, characterized in that, Includes the following modules: Meteorological-Operational Memory Construction Module: Performs in-depth correlation analysis on historical meteorological data, flight operation data and operational result data, generates a triplet memory unit of meteorological model-operational response-effect evaluation and stores it in the memory bank; Real-time situational awareness and vulnerability assessment module: Receives real-time multi-dimensional data, combines it with meteorological and operational memory, identifies combinations of weak disturbance elements through a large artificial intelligence model, and generates a vulnerability heat map of the airport operation system; Collaborative decision-making plan generation module: Analyzes the vulnerability heatmap and generates a hierarchical collaborative decision-making plan that includes triggering conditions, executing entities, and operation sequences.

Images