Intelligent airspace data optimization storage method and system

Abstract

This invention discloses an intelligent method and system for optimizing and storing airspace data. The method includes airspace data acquisition, airspace feature pre-selection, final optimization of airspace features, robustness quantification of airspace data, airspace data clustering, and optimized storage of airspace data. This invention belongs to the field of data storage, specifically referring to an intelligent method and system for optimizing and storing airspace data. This scheme achieves Pareto optimality of value and cost by introducing time weights to construct feature interaction weights and combining them with dynamically decreasing inertial weights; it achieves accurate removal of abnormal airspace data by constructing inherent robustness confidence values; it introduces anomaly-driven dynamic decay factors to construct inherent confidence values ​​and designs collaborative confidence values ​​endorsed by neighboring airspace data to avoid being misled by low-robustness airspace data; it introduces dynamic boundaries and superimposes robustness difference weights; and it designs an anomaly score removal mechanism based on cluster-internal confidence value distribution, thereby improving the robustness of optimized storage of airspace data.

Description

Technical Field

[0001] This invention relates to the field of data storage, specifically to an intelligent method and system for optimizing and storing airspace data. Background Technology

[0002] Airspace data optimization storage methods are a class of data management techniques that balance storage costs and data availability while ensuring the core application value of airspace data. These methods utilize features, redundancy removal, and storage strategy adaptation. However, general airspace data optimization storage methods often neglect time value and feature co-constitutive redundancy, leading to storage resource mismatch, an imbalance between storage value and cost, and ultimately poor optimization performance. Furthermore, these methods often fail to quantify the impact of abnormal airspace data, employ coarse attenuation logic, ignore the co-constitutive endorsement of neighboring airspace data, and suffer from distorted judgments of single airspace data, resulting in storage latency and poor robustness. Summary of the Invention

[0003] To address the aforementioned issues and overcome the shortcomings of existing technologies, this invention provides an intelligent airspace data optimization and storage method and system. Addressing the problem that general airspace data optimization and storage methods neglect time value and feature co-redundancy, leading to storage resource misallocation, an imbalance between storage value and cost, and consequently poor optimization performance, this solution introduces time weights to construct feature interaction weights, ensuring resources are tilted towards high real-time airspace data, thus enabling accurate removal of redundant features. Combined with dynamically decreasing inertial weights and the introduction of a dual-objective fitness function, it achieves Pareto optimality for value and cost. Furthermore, it constructs inherent robustness confidence values ​​to accurately remove abnormal airspace data, thereby improving the efficiency of airspace data optimization and storage. To optimize storage performance, this solution addresses several issues with conventional spatial data optimization methods. These include unquantified impact of anomalous spatial data, coarse attenuation logic, neglect of collaborative endorsement from neighboring spatial data, and distorted judgments based on single spatial data, leading to storage latency and poor robustness. This solution addresses these problems by introducing anomaly-driven dynamic attenuation factors to construct inherent confidence values, quantifying the impact of anomalies and preventing confidence value distortion. It also designs collaborative confidence values ​​backed by neighboring spatial data to avoid being misled by low-robustness spatial data; introduces dynamic boundaries and superimposes robustness difference weights; and designs an anomaly score removal mechanism based on cluster-internal confidence value distribution to accurately quantify the contribution of spatial data. Ultimately, this improves the robustness of optimized spatial data storage.

[0004] The technical solution adopted by this invention is as follows: This invention provides an intelligent airspace data optimization and storage method, which includes the following steps:

[0005] Step S1: Spatial data acquisition;

[0006] Step S2: Pre-selection of spatial features;

[0007] Step S3: Final optimization of spatial features;

[0008] Step S4: Robustness quantification of spatial data;

[0009] Step S5: Spatial data clustering;

[0010] Step S6: Optimize storage of spatial data.

[0011] Further, in step S1, the airspace data acquisition involves obtaining airspace data, including airspace monitoring data, flight operation data, and environmental protection data; the acquired airspace data undergoes initial filtering and standardization processing; the initial filtering involves removing airspace data with transmission errors and filtering physically invalid airspace data.

[0012] Further, in step S2, the pre-selection of spatial features specifically includes:

[0013] Construct spatial feature interaction weights to measure the value of feature storage;

[0014] Spatial feature pre-selection involves dividing features into redundant groups and core groups using the average value of all feature interaction weights as a threshold; then, features are selected from the core groups as a subset of pre-selected stored features.

[0015] Furthermore, in step S3, the final optimization of the spatial features specifically includes:

[0016] Initialization: Initialize the selection method for different pre-selected storage feature subsets; the selection method for each pre-selected storage feature subset is treated as an optimization unit;

[0017] Decreasing inertia weight design, dynamically adjusting the optimization direction;

[0018] Optimize individual location updates to ensure the discreteness of feature selection;

[0019] The dual-objective fitness function design outputs the spatial data corresponding to the best individual in the population when the maximum number of optimization attempts is reached or the fitness of the best individual in the population converges, thus completing the final optimization of the spatial features.

[0020] Furthermore, in step S4, the robust quantization of spatial data specifically includes:

[0021] Inherent robustness confidence values ​​are initialized to quantify the fundamental robustness of spatial domain data.

[0022] The robustness confidence value of the airspace data is updated by introducing a decay factor to drive robustness decay through anomalous airspace data; then the inherent robustness confidence value is updated.

[0023] Design collaborative robustness confidence values, backed by adjacent spatial data;

[0024] The final comprehensive robustness confidence value is obtained.

[0025] Furthermore, in step S5, the spatial data clustering specifically includes:

[0026] For the selection of spatial data cluster centers, a storage priority cluster center decision value is constructed based on a comprehensive robustness confidence value; the spatial data with the highest decision value is selected as the initial cluster center.

[0027] The spatial domain data storage priority allocation introduces a dynamic cluster boundary, which, based on the feature distance, superimposes the robustness confidence value difference between the spatial domain data and the cluster center to allocate storage priority;

[0028] Anomaly detection of spatial data within a cluster is performed by calculating anomaly scores for each spatial data point based on the statistical distribution of the comprehensive confidence value within the cluster.

[0029] The storage priority cluster center is updated, and the spatial data update weight is calculated based on the anomaly score to quantify the contribution of the spatial data.

[0030] The overall confidence value of the cluster centers is updated by updating the weights based on spatial data.

[0031] Furthermore, in step S6, the optimized storage of spatial data is based on converged spatial data, and priorities are assigned according to the comprehensive robustness confidence value of the cluster center; thereby achieving optimized storage of spatial data.

[0032] This invention provides an intelligent airspace data optimization and storage system, comprising an airspace data acquisition module, an airspace feature pre-selection module, an airspace feature final optimization module, an airspace data robustness quantification module, an airspace data clustering module, and an airspace data optimization and storage module;

[0033] The airspace data acquisition module acquires airspace data and performs initial filtering and standardization processing.

[0034] The spatial feature pre-selection module quantifies the value of spatial data storage through feature interaction weights and performs spatial feature pre-selection.

[0035] The final optimization module for spatial features is based on pre-selected spatial features and designs a decreasing inertial weight to optimize feature selection, thereby obtaining the optimal combination of stored features.

[0036] The spatial data robustness quantification module obtains a comprehensive confidence value based on inherent and collaborative robustness confidence values, and performs dynamic robustness updates.

[0037] The spatial data clustering module selects the initial cluster centers of the spatial data based on robustness, and allocates storage priorities by combining feature distance and confidence value differences to achieve spatial data clustering.

[0038] The spatial data optimization and storage module achieves optimized storage of spatial data based on the spatial data clustering results.

[0039] The beneficial effects achieved by the present invention using the above solution are as follows:

[0040] (1) In view of the problem that general spatial data optimization storage methods ignore time value and feature co-redundancy, resulting in misallocation of storage resources, imbalance between storage value and cost, and thus poor optimization storage effect, this solution introduces time weight to construct feature interaction weight to ensure that resources are tilted towards high real-time spatial data, so as to accurately remove redundant features of spatial data; combined with dynamically decreasing inertial weight and introducing a dual-objective fitness function, Pareto optimality of value and cost is achieved; and abnormal spatial data is accurately removed by constructing inherent robustness confidence value; thereby improving the optimization storage effect of spatial data.

[0041] (2) To address the issues of unquantified impact of abnormal spatial data, coarse attenuation logic, neglect of collaborative endorsement from neighboring spatial data, and distorted judgment of single spatial data in general spatial data optimization storage methods, which lead to storage delays and poor robustness, this solution introduces an anomaly-driven dynamic attenuation factor to construct an inherent confidence value, quantifies the impact of anomalies, and avoids confidence value distortion; designs a collaborative confidence value endorsed by neighboring spatial data to avoid being misled by low-robustness spatial data; introduces a dynamic boundary and superimposes robustness difference weights; and designs an anomaly score elimination mechanism based on cluster-embedded confidence value distribution to accurately quantify the contribution of spatial data; thereby improving the robustness of spatial data optimization storage. Attached Figure Description

[0042] Figure 1 A schematic diagram illustrating an intelligent airspace data optimization and storage method provided by the present invention;

[0043] Figure 2 A schematic diagram of an intelligent airspace data optimization and storage system provided by the present invention;

[0044] Figure 3 This is a flowchart illustrating step S4.

[0045] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. Detailed Implementation

[0046] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0047] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the system or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0048] Example 1, see Figure 1 The present invention provides an intelligent airspace data optimization and storage method, which includes the following steps:

[0049] Step S1: Spatial data acquisition, acquiring spatial data and performing initial filtering and standardization processing;

[0050] Step S2: Spatial feature pre-selection, quantifying the value of spatial data storage through feature interaction weights, and performing spatial feature pre-selection;

[0051] Step S3: Final optimization of spatial features. Based on the pre-selected spatial features, a decreasing inertial weight is designed to optimize feature selection and obtain the optimal combination of storage features.

[0052] Step S4: Spatial data robustness quantification, obtaining a comprehensive confidence value based on inherent and collaborative robustness confidence values, and dynamically updating the robustness;

[0053] Step S5: Spatial data clustering. Based on robustness, the initial cluster centers of the spatial data are selected, and storage priorities are allocated by combining feature distance and confidence value differences to achieve spatial data clustering.

[0054] Step S6: Optimize the storage of spatial data, and realize the optimized storage of spatial data based on the spatial data clustering results.

[0055] Example 2, see Figure 1This embodiment is based on the above embodiment. In step S1, airspace data acquisition involves obtaining airspace data, including airspace monitoring data, flight operation data, and environmental support data. The airspace monitoring data includes radar data, ADS-B data, and secondary radar data. The flight operation data includes flight plan data and real-time operation data. The environmental support data includes meteorological data and airspace environment data. The acquired airspace data undergoes initial filtering and standardization processing. The initial filtering removes airspace data with transmission errors (airspace data frame verification failure, field missing ≥30%) and filters physically invalid airspace data (radar coordinates exceeding the airspace sector range).

[0056] Example 3, see Figure 1 This embodiment is based on the above embodiment. In step S2, the pre-selection of airspace features is inherently determined by the real-time nature of airspace data, which determines its storage value (recent radar data has a much greater reference value for airspace scheduling decisions than historically expired airspace data). Simultaneously, it is necessary to quantify the collaborative redundancy relationships between features (the synergy between flight speed and airspace traffic density, and the redundancy of repeated radar sampling points). Therefore, an S-shaped time weight is introduced to assign higher weights to recent airspace data features. At the same time, an interaction weight is constructed to determine the necessity of storing features. The airspace data is then grouped according to the interaction weights, and redundant features are eliminated. The specific operation is as follows:

[0057] Construct spatial feature interaction weights to measure the value of feature storage, represented as follows: ;in, and It represents any two spatial features; S is the spatial data storage value category, divided into high, medium, and low priorities; CMI(·|·) is conditional mutual information; MI(·) is mutual information; It is time weight; ; This is the time weighting adjustment coefficient, ranging from 0.6 to 1.2; t is the current airspace data acquisition time; T is the airspace data storage period; when t > T / 2 (recent airspace data), Approaching 1 (higher weight, stored first);

[0058] Spatial feature pre-selection is based on the average of all feature interaction weights. The threshold is used to divide the features into redundant groups ( (low storage value) and core group ( (High storage value); then select U (values ​​3 to 5) features from the core group as a pre-selected storage feature subset, which reduces redundancy while retaining core value features.

[0059] Example 4, see Figure 1This embodiment is based on the above embodiment. In step S3, the final optimization of spatial features is to address the problem that the pre-selected storage features may still have excessively high storage costs (due to the huge amount of spatial data for some high-value features) or local optima. This is addressed by designing a decreasing inertia weight to explore high-value feature combinations globally in the early stage and then converge to a low-cost storage scheme locally in the later stage. A dual-objective fitness function is also constructed, considering both storage value and storage cost, to achieve the optimal combination of high-value and low-cost storage features. The specific operation is as follows:

[0060] Initialization: Initialize the selection method of different pre-selected storage feature subsets; each selection method of pre-selected storage feature subset is treated as an optimization individual, and the corresponding dimension of the individual is assigned a value of 1 when a feature is selected; for the moved individual position, if the dimension value is not less than 0.5, it is assigned a value of 1, otherwise it is assigned a value of 0; if the dimension of the individual position assigned a value of 1 is higher than U, only the first U items of the dimension value are retained and assigned a dimension value of 1.

[0061] Decreasing inertia weight design, dynamically adjusting the optimization direction to avoid local optima, is expressed as:

[0062] Where K is a scaling factor that controls the weight range, with a value of 0.8 to 1.2. and These are the maximum and minimum values ​​of the inertia weight, respectively, ranging from 0.8 to 0.9 and from 0.3 to 0.4. It is the fitness value of the currently optimized individual; It is the fitness value of the best individual in the population; It is a smoothing term, taking the value 1e. -5 ~1e -6 ; is the rate factor, which controls the decay rate, and its value ranges from 0.5 to 1.0; k is the current optimization count; This is the maximum number of optimization attempts, ranging from 50 to 100. It is the inertial weight of the j-th individual;

[0063] Optimizing individual location updates while ensuring the discreteness of feature selection can be represented as:

[0064] ;

[0065] ;

[0066] in, It represents the position of the j-th individual at the corresponding number of optimization iterations; Sigmoid(·) is the Sigmoid function; It is the speed at which the j-th individual corresponds to the number of optimizations; It is the position of the optimal individual; It represents the individual's fitness value; rand is a random number between 0 and 1.

[0067] The bi-objective fitness function design is expressed as: ;in, is the weighting factor, with a value of 0.5 to 0.7; G is the storage value clustering profile coefficient based on the current features (classifying the storage value of historical spatial data through k-means clustering; the larger G is, the higher the feature's ability to distinguish value and the stronger the storage value). It is the total amount of spatial data for the pre-selected features; It is the total amount of spatial data for all original features; The smaller the value, the lower the storage cost; when the maximum number of optimization iterations (values ​​from 5 to 10) are reached or the fitness of the optimal individual in the population converges (convergence threshold is 1e), the storage cost is lower. -4 ~1e -3 Output the spatial data corresponding to the best individual in the population, and complete the final optimization of the spatial features.

[0068] Example 5, see Figure 1 and Figure 3 This embodiment is based on the above embodiment. In step S4, the robustness quantification of airspace data requires that airspace data be stored hierarchically according to scheduling urgency and reuse frequency. However, some airspace data has anomalies (invalid airspace data generated by radar failure, false flight trajectories). If normal airspace data is mixed in, it will lead to a waste of storage resources. Therefore, the robustness confidence value of airspace data is the core, and a hierarchical storage process of confidence value modeling - dynamic clustering - anomaly removal - priority update is implemented. Specifically, it includes:

[0069] The inherent robustness confidence value is initialized, quantifying the fundamental robustness of the spatial domain data, and is expressed as follows: Where c represents the number of normal spatial data acquisitions; d represents the number of abnormal spatial data acquisitions. It is the local density of spatial data m; it quantifies the degree of spatial core. Where m and n are any two spatial data samples; L mn It is the spatial feature distance between spatial data samples (using Euclidean distance); L th It is the cutoff distance, which selects the maximum spatial feature distance between spatial data samples; It is the inherent robustness confidence value of the initialization of spatial data m.

[0070] By performing the above operations, this solution addresses the problem that general spatial data optimization storage methods neglect time value and feature co-redundancy, leading to storage resource misallocation, an imbalance between storage value and cost, and consequently poor optimization storage performance. This solution introduces time weights to construct feature interaction weights, ensuring resources are tilted towards high real-time spatial data, thus accurately removing redundant features from the spatial data. Combined with dynamically decreasing inertial weights and a dual-objective fitness function, it achieves Pareto optimality for value and cost. Furthermore, it constructs inherent robustness confidence values ​​to accurately remove abnormal spatial data, thereby improving the spatial data optimization storage performance.

[0071] Example 6, see Figure 1 and Figure 3 This embodiment is based on the above embodiment. In step S4, the robust quantization of spatial data further includes:

[0072] Spatial data robustness confidence value update, introducing a decay factor A robust decay is driven by anomalous spatial data. The higher the proportion of anomalous spatial data, the faster the confidence value decays, as expressed as: ;in, It is the attenuation coefficient, with a value of 0.7 to 0.9, which controls the degree to which historical confidence values ​​are retained; It is the decay factor from the previous acquisition cycle; then, the inherent robustness confidence value is updated, expressed as: ;in, , These are the intrinsic confidence values ​​for the current period and the previous period of spatial data m, respectively. The initial intrinsic confidence value is... V represents the effective percentage of airspace data, calculated as the number of normal airspace data collections divided by the total number of collections, with a statistical period of 1-2 hours. It is the minimum distance from spatial data m to higher-density spatial data; It is the global maximum distance;

[0073] Design collaborative robustness confidence value The robustness is enhanced by the support of adjacent airspace data, as shown below: ; ;in, It is the set of neighboring spatial data of spatial data m, where the neighbor number M takes the value 10~20, and p is the index of the neighboring spatial data. It is the nearest neighbor weight; It is the updated inherent robustness confidence value of the neighboring spatial data p; It is the initial inherent robustness confidence value of the neighboring spatial data p;

[0074] Finally, the overall robustness confidence value is obtained. , represented as: If the airspace data contains serious anomalies (false flight trajectories, airspace data from completely faulty radar), then the collaborative confidence values ​​are removed, and only the data from the airspace data is retained. This is to prevent abnormal spatial data from obtaining high confidence values ​​through neighbor endorsement, which could lead to incorrect storage.

[0075] Example 7, see Figure 1 This embodiment is based on the above embodiment. In step S5, the spatial data clustering specifically includes:

[0076] For spatial data cluster center selection, highly robust and core spatial data are chosen as initial cluster centers, and priority cluster center decision values ​​are stored. express: Select the S spatial data points with the highest decision values ​​as the initial cluster centers.

[0077] Spatial domain data storage priority allocation introduces a dynamic clustering boundary. Based on feature distance, it superimposes the robustness confidence value difference between spatial domain data and cluster centers. Less robust spatial domain data is more difficult to be assigned to high-priority clusters, ensuring that high-priority storage resources are used only for highly robust spatial domain data. The storage priority allocation formula is expressed as: ;in, It is the final storage priority cluster for spatial data m; It is the comprehensive robustness confidence value of the cluster centers c in the spatial data; C is the set of cluster centers; It is a moderating factor, ranging from 0.5 to 0.8, which controls the degree of influence of robustness differences; It is the feature distance (using Euclidean distance) between the spatial data m and the spatial data cluster center c.

[0078] Intra-cluster anomaly spatial data detection is performed based on the statistical distribution of the comprehensive confidence value within the cluster. Anomaly scores are calculated for each spatial data point to ensure high robustness of spatial data within each priority cluster and avoid wasting storage resources. The intra-cluster spatial data anomaly score is represented as follows: ;in, It is the anomaly score of the spatial data m; and It is the mean and standard deviation of the combined confidence values ​​of all spatial data within the cluster to which m belongs;

[0079] The storage priority cluster center update calculates the spatial data update weight, quantifies the contribution of spatial data, and is represented as follows: ;in, It is the update weight of the spatial data m; It is the maximum decision value of the storage priority cluster center. A higher decision value means higher weight and stronger core.

[0080] The overall confidence value of cluster centers is updated based on the spatial data update weights, and is expressed as follows: ;in, It is the updated overall confidence value of the cluster centers; It is the set of all spatial data in cluster center c;

[0081] Storage priority clustering determination: To avoid storage latency caused by infinite clustering iterations, a maximum number of clustering iterations is preset (10-15 times). When any of the following conditions are met, clustering terminates and the final storage priority scheme is output. The conditions include: the rate of change of the comprehensive confidence value of the cluster centers in two consecutive iterations is less than a threshold (0.03-0.07), indicating that the centers are stable and the priority division has not changed significantly; the maximum number of clustering iterations is reached; if the conditions are not met, the process returns to spatial data storage priority allocation to redistribute priorities until convergence.

[0082] By performing the above operations, this solution addresses the problems of unquantified impact of abnormal spatial data, coarse attenuation logic, neglect of collaborative endorsement from neighboring spatial data, and distorted judgment of single spatial data in general spatial data optimization storage methods, leading to storage latency and poor robustness. This solution addresses these issues by introducing an anomaly-driven dynamic attenuation factor to construct an inherent confidence value, quantifying the impact of anomalies and avoiding confidence value distortion; designing a collaborative confidence value endorsed by neighboring spatial data to avoid being misled by low-robustness spatial data; introducing dynamic boundaries and superimposing robustness difference weights; and designing an anomaly score removal mechanism based on cluster-internal confidence value distribution to accurately quantify the contribution of spatial data; thereby improving the robustness of spatial data optimization storage.

[0083] Example 8, see Figure 1 This embodiment is based on the above embodiment. In step S6, the spatial data optimization storage is based on the converged spatial data and is divided into high, medium and low priorities according to the comprehensive robustness confidence value of the cluster center. The highest comprehensive robustness confidence value is the high priority, and hierarchical storage is performed. The high priority cluster spatial data is stored in the SSD solid-state drive, the medium priority is stored in the SAS hard drive, and the low priority is stored in the tape library.

[0084] Example 9, see Figure 2 Based on the above embodiments, this embodiment provides an intelligent airspace data optimization and storage system, including an airspace data acquisition module, an airspace feature pre-selection module, an airspace feature final optimization module, an airspace data robustness quantification module, an airspace data clustering module, and an airspace data optimization and storage module.

[0085] The airspace data acquisition module acquires airspace data and performs initial filtering and standardization processing.

[0086] The spatial feature pre-selection module quantifies the value of spatial data storage through feature interaction weights and performs spatial feature pre-selection.

[0087] The final optimization module for spatial features is based on pre-selected spatial features and designs a decreasing inertial weight to optimize feature selection, thereby obtaining the optimal combination of stored features.

[0088] The spatial data robustness quantification module obtains a comprehensive confidence value based on inherent and collaborative robustness confidence values, and performs dynamic robustness updates.

[0089] The spatial data clustering module selects the initial cluster centers of the spatial data based on robustness, and allocates storage priorities by combining feature distance and confidence value differences to achieve spatial data clustering.

[0090] The spatial data optimization and storage module achieves optimized storage of spatial data based on the spatial data clustering results.

[0091] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

[0092] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. An intelligent method for optimizing and storing airspace data, characterized in that: The method includes the following steps: Step S1: Spatial data acquisition, acquiring spatial data and performing initial filtering and standardization processing; Step S2: Spatial feature pre-selection, quantifying the value of spatial data storage through feature interaction weights, and performing spatial feature pre-selection; Step S3: Final optimization of spatial features. Based on the pre-selected spatial features, a decreasing inertial weight is designed to optimize feature selection and obtain the optimal combination of storage features. Step S4: Spatial data robustness quantification, obtaining a comprehensive confidence value based on inherent and collaborative robustness confidence values, and dynamically updating the robustness; Step S5: Spatial data clustering. Based on robustness, the initial cluster centers of the spatial data are selected, and storage priorities are allocated by combining feature distance and confidence value differences to achieve spatial data clustering. Step S6: Optimize the storage of spatial data, and realize the optimized storage of spatial data based on the spatial data clustering results; In step S3, the final optimization of the spatial features specifically includes: Initialization: Initialize the selection method for different pre-selected storage feature subsets; the selection method for each pre-selected storage feature subset is treated as an optimization unit; Decreasing inertia weight design, dynamically adjusting the optimization direction; Optimize individual location updates to ensure the discreteness of feature selection; The dual-objective fitness function design outputs the spatial data corresponding to the best individual in the population when the maximum number of optimizations is reached or the fitness of the best individual in the population converges, thus completing the final optimization of the spatial features. In step S4, the robust quantization of spatial data specifically includes: Inherent robustness confidence values ​​are initialized to quantify the fundamental robustness of spatial domain data. The robustness confidence value of the airspace data is updated by introducing a decay factor to drive robustness decay through anomalous airspace data; then the inherent robustness confidence value is updated. Design collaborative robustness confidence values, backed by adjacent spatial data; The final comprehensive robustness confidence value is obtained.

2. The intelligent airspace data optimization and storage method according to claim 1, characterized in that: In step S2, the spatial feature pre-selection specifically includes: Construct spatial feature interaction weights to measure the value of feature storage; Spatial feature pre-selection involves dividing features into redundant groups and core groups using the average value of all feature interaction weights as a threshold; then, features are selected from the core groups as a pre-selected storage feature subset.

3. The intelligent airspace data optimization and storage method according to claim 2, characterized in that: In step S5, the spatial data clustering specifically includes: For the selection of spatial data cluster centers, a storage priority cluster center decision value is constructed based on a comprehensive robustness confidence value; the spatial data with the highest decision value is selected as the initial cluster center. Priority allocation for airspace data storage; Intra-cluster anomaly spatial data detection: Based on the statistical distribution of the comprehensive confidence value within the cluster, anomaly score is calculated for each spatial data. The storage priority cluster center is updated, and the spatial data update weight is calculated based on the anomaly score to quantify the contribution of the spatial data. The overall confidence value of the cluster centers is updated by updating the weights based on spatial data.

4. The intelligent airspace data optimization and storage method according to claim 3, characterized in that: In step S5, the spatial data storage priority allocation introduces a dynamic clustering boundary. Based on the feature distance, the difference in robustness confidence values ​​between the spatial data and the cluster center is superimposed to allocate storage priority.

5. The intelligent airspace data optimization and storage method according to claim 4, characterized in that: In step S1, the airspace data acquisition involves obtaining airspace data, including airspace monitoring data, flight operation data, and environmental protection data; and performing initial filtering and standardization processing on the acquired airspace data. The initial filtering process involves removing spatial data with transmission errors and filtering physically invalid spatial data.

6. The intelligent airspace data optimization and storage method according to claim 5, characterized in that: In step S6, the optimized storage of spatial data is based on converged spatial data, and priorities are assigned according to the comprehensive robustness confidence value of the cluster center; thereby achieving optimized storage of spatial data.

7. An intelligent airspace data optimization and storage system, used to implement the intelligent airspace data optimization and storage method as described in any one of claims 1-6, characterized in that: It includes a spatial data acquisition module, a spatial feature pre-selection module, a spatial feature final optimization module, a spatial data robustness quantification module, a spatial data clustering module, and a spatial data optimization and storage module. The airspace data acquisition module acquires airspace data and performs initial filtering and standardization processing. The spatial feature pre-selection module quantifies the value of spatial data storage through feature interaction weights and performs spatial feature pre-selection. The final optimization module for spatial features is based on pre-selected spatial features and designs a decreasing inertial weight to optimize feature selection, thereby obtaining the optimal combination of stored features. The spatial data robustness quantification module obtains a comprehensive confidence value based on inherent and collaborative robustness confidence values, and performs dynamic robustness updates. The spatial data clustering module selects the initial cluster centers of the spatial data based on robustness, and allocates storage priorities by combining feature distance and confidence value differences to achieve spatial data clustering. The spatial data optimization and storage module achieves optimized storage of spatial data based on the spatial data clustering results.

Images