Multi-source data driven low-altitude operation intelligent scheduling method and system

Abstract

The invention relates to the technical field of big data analysis, and discloses a multi-source data driven low-altitude operation intelligent scheduling method and system, and the method comprises the steps: firstly obtaining multi-source heterogeneous data through the system, and carrying out the standardization processing, and forming a unified multi-source data set; and on this basis, dynamic identification and density evaluation are executed to generate a traffic density distribution map. And then performing clustering analysis on the distribution map, extracting a boundary range of a key region, updating a fence boundary according to the boundary range, and generating a new fence constraint condition. And the system corrects the flight parameters according to the constraint to obtain an adjusted running state, and carries out data acquisition in combination with traffic judgment to form regional traffic dynamic data with higher precision. And finally, the system executes load balancing scheduling and abnormal dynamic updating based on the data to form a newest low-altitude traffic management state. The method can solve the problem of edge resource scheduling staticization.

Description

Technical Field

[0001] The present invention relates to the field of big data analysis technology, and in particular to a low-altitude operation intelligent scheduling method and system driven by multi-source data. Background Art

[0002] The present invention relates to the field of big data analysis technology. Big data analysis refers to a key technology for collecting, processing, fusing, and modeling massive amounts of heterogeneous structured and unstructured data to explore potential correlation patterns and support intelligent decision-making. With the development of infrastructure such as the Internet of Things, cloud computing, and edge computing, big data analysis technology has been widely used in finance, healthcare, transportation, security, energy, and other fields, promoting the transition from a static rule-driven to a data-driven management model. This technology emphasizes the ability to efficiently process real-time, spatial, and multi-dimensional dynamic information, and shows significant potential in scenarios that require rapid perception and response to complex environmental changes.

[0003] Existing low-altitude operation management technologies generally use a static allocation mechanism based on preset rules for regional resource scheduling. This mechanism maps edge computing tasks to nodes based on geographic region or node performance, lacking dynamic awareness and response to real-time data status. Because existing technologies do not collect real-time operational pressures on current nodes or analyze fluctuations in traffic density, they struggle to meet the practical needs for flexible edge node scheduling in high-density mission scenarios.

[0004] In summary, the existing technology has the problem of static edge resource scheduling. Summary of the Invention

[0005] The present invention provides a multi-source data-driven low-altitude operation intelligent scheduling method and system to solve the problem of static edge resource scheduling.

[0006] In a first aspect, in order to solve the above technical problems, the present invention provides a multi-source data-driven low-altitude operation intelligent scheduling method, comprising:

[0007] Acquire multi-source heterogeneous data;

[0008] Performing data standardization processing on the multi-source heterogeneous data to obtain a multi-source data set;

[0009] Performing dynamic identification and density assessment based on the multi-source data set to obtain a traffic density distribution map;

[0010] Perform cluster identification analysis based on the traffic density distribution map to obtain the boundary range of the key area;

[0011] According to the boundary range of the key area, the fence boundary is updated to obtain the updated fence constraint condition;

[0012] Correcting flight parameters according to the updated fence constraint conditions to obtain an adjusted operating state;

[0013] According to the adjusted operating status, traffic determination and data collection are performed to obtain more accurate regional traffic dynamic data;

[0014] Perform resource load balancing scheduling based on the higher-precision regional traffic dynamic data to obtain a final resource allocation plan;

[0015] According to the final resource allocation plan, abnormal monitoring and dynamic adjustment are carried out to obtain the latest low-altitude traffic management status.

[0016] Preferably, data standardization is performed on the multi-source heterogeneous data to obtain a multi-source data set, including:

[0017] The multi-source heterogeneous data includes: UAV sensor data, ground monitoring data and meteorological information data;

[0018] Performing field mapping and structure conversion on the drone sensing data to obtain a sensor standard data set;

[0019] Extracting fields and regularizing formats based on the ground monitoring data to obtain a monitoring standard data set;

[0020] The sensor standard data set and the monitoring standard data set are integrated to obtain a multi-source data set in a unified format.

[0021] Preferably, dynamic identification and density assessment are performed based on the multi-source data set to obtain a traffic density distribution map, including:

[0022] Performing unit division and data extraction based on the multi-source data set to obtain a unit-level preliminary traffic data set, wherein the unit-level preliminary traffic data set includes speed data, height data, location information, and target identification;

[0023] Calculating the magnitude of change within each unit based on the speed data and the height data in the traffic data set;

[0024] If the change amplitude within the unit is greater than or equal to a preset amplitude threshold, the corresponding unit is marked as a high dynamic area;

[0025] If the amplitude of the change in the unit is less than a preset amplitude threshold, the unit is marked as a low dynamic area;

[0026] Integrating the high dynamic area and the low dynamic area to generate a dynamic distribution feature map;

[0027] A density distribution map is generated based on the dynamic distribution characteristic map, the target identifier and the location information to obtain a traffic density distribution map.

[0028] Preferably, cluster identification analysis is performed based on the traffic density distribution map to obtain the boundary range of the key area, including:

[0029] Perform density screening according to the traffic density distribution map to obtain a high-density unit set;

[0030] Perform cluster identification based on the high-density unit set to obtain a key area identification set;

[0031] Based on the key area identification set, boundary extraction and expansion are performed to obtain the boundary range of the key area.

[0032] Preferably, the fence boundary is updated according to the boundary range of the key area to obtain the updated fence constraint conditions, including:

[0033] Obtaining boundary coordinate data of the key area;

[0034] Generate a vector fence based on the boundary range coordinate data of the key area to obtain a preliminary fence boundary shape;

[0035] Performing boundary expansion processing according to the preliminary fence boundary shape to obtain an expanded fence boundary shape;

[0036] Based on the expanded fence boundary shape, boundary confirmation synchronization is performed to obtain updated fence constraint conditions.

[0037] Preferably, traffic determination and data collection are performed based on the adjusted operating state to obtain more accurate regional traffic dynamic data, including:

[0038] Performing safety parameter comparison and traffic determination according to the adjusted operating state to obtain a traffic determination result;

[0039] According to the access determination result, permission mapping and allocation operations are performed to obtain effective access permissions;

[0040] Based on the effective traffic rights, priority road sections are screened and nodes are planned to obtain a node deployment plan;

[0041] According to the node deployment plan, data integration and frequency evaluation are performed to obtain more accurate regional traffic dynamic data.

[0042] Preferably, resource load balancing scheduling is performed based on the higher-precision regional traffic dynamic data to obtain a final resource allocation solution, including:

[0043] Perform load detection and list generation based on the higher-precision regional traffic dynamic data to obtain a node load status list;

[0044] Perform task migration scheduling according to the node load status list to obtain load distribution status;

[0045] According to the load distribution status, the scheme is synchronously deployed to obtain the final resource allocation scheme.

[0046] Preferably, according to the final resource allocation plan, abnormal monitoring and dynamic adjustment are performed to obtain the latest low-altitude traffic management status, including:

[0047] According to the final resource allocation plan, data collection and density calculation are performed to obtain the traffic density of each area;

[0048] If the traffic density is greater than or equal to a preset density threshold, the abnormal time point is recorded and the corresponding area location information is obtained to obtain the initial distribution information of the abnormal fluctuation;

[0049] If the traffic density is less than the preset density threshold, the corresponding area is judged to be in a normal state and fluctuation detection is continued;

[0050] According to the initial distribution information of the abnormal fluctuation, dynamic range identification and update instruction generation are performed to obtain updated traffic situation map and fence boundary data;

[0051] According to the updated traffic situation map and fence boundary data, synchronous publishing and status confirmation are performed to obtain the latest low-altitude traffic management status.

[0052] In a second aspect, the present invention provides a multi-source data-driven low-altitude operation intelligent scheduling system, comprising:

[0053] Data acquisition module, used to acquire multi-source heterogeneous data;

[0054] A standardization module, configured to perform data standardization processing on the multi-source heterogeneous data to obtain a multi-source data set;

[0055] A traffic density module is used to perform dynamic identification and density assessment based on the multi-source data set to obtain a traffic density distribution map;

[0056] A boundary range module is used to perform cluster identification analysis based on the traffic density distribution map to obtain the boundary range of the key area;

[0057] A constraint condition module, configured to update the fence boundary according to the boundary range of the key area to obtain an updated fence constraint condition;

[0058] a correction module, configured to correct flight parameters according to the updated fence constraint conditions to obtain an adjusted operating state;

[0059] A dynamic data module is used to perform traffic determination and data collection based on the adjusted operating status to obtain more accurate regional traffic dynamic data;

[0060] An allocation plan module is used to perform resource load balancing scheduling based on the higher-precision regional traffic dynamic data to obtain a final resource allocation plan;

[0061] The management status module is used to perform abnormal monitoring and dynamic adjustment according to the final resource allocation plan to obtain the latest low-altitude traffic management status.

[0062] In a third aspect, the present invention also provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the multi-source data-driven low-altitude operation intelligent scheduling method described above is implemented.

[0063] In a fourth aspect, the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the multi-source data-driven low-altitude operation intelligent scheduling methods described above.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] (1) The present invention acquires multi-source heterogeneous data and performs standardized processing, which can integrate UAV sensor data, ground monitoring data and meteorological information data to achieve data integration with unified format and consistent structure, improve the compatibility and information integrity of downstream processing, and solve the problem of fragmented data sources and difficulty in unified analysis in traditional methods.

[0066] (2) The present invention can accurately distinguish high-dynamic and low-dynamic areas by dividing multi-source data sets into units, calculating the change amplitude, and generating density distribution maps, and generate traffic density distribution maps with spatial distribution characteristics, effectively improving the accuracy of target recognition and the precision of dynamic perception in the area, and supporting the precise formulation of subsequent management strategies.

[0067] (3) The present invention performs task migration and synchronous deployment based on traffic density and operating status data, combined with real-time load information of edge nodes. It can flexibly adjust computing resource allocation according to the current node status, effectively alleviate the overload problem of some nodes, and improve the overall resource utilization efficiency and responsiveness of the system.

[0068] (4) The present invention performs load detection, task migration and synchronous deployment operations based on traffic dynamic data, and can dynamically adjust the resource allocation structure according to the real-time load status of the edge node, breaking through the limitations of the existing static edge resource scheduling, and improving the system's scheduling flexibility and processing capabilities in high-concurrency or sudden task scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 This is a flowchart of a multi-source data-driven low-altitude operation intelligent scheduling method provided by the first embodiment of the present invention;

[0070] Figure 2 This is a structural diagram of a low-altitude operation intelligent scheduling system driven by multi-source data provided in the second embodiment of the present invention. DETAILED DESCRIPTION

[0071] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0072] Reference Figure 1 The first embodiment of the present invention provides a multi-source data-driven low-altitude operation intelligent scheduling method, comprising the following steps:

[0073] S11, acquiring multi-source heterogeneous data;

[0074] S12, performing data standardization processing on the multi-source heterogeneous data to obtain a multi-source data set;

[0075] S13, performing dynamic identification and density assessment based on the multi-source data set to obtain a traffic density distribution map;

[0076] S14, performing cluster identification analysis based on the traffic density distribution map to obtain the boundary range of the key area;

[0077] S15, updating the fence boundary according to the boundary range of the key area to obtain an updated fence constraint condition;

[0078] S16, performing flight parameter correction according to the updated fence constraint condition to obtain an adjusted operating state;

[0079] S17, performing traffic determination and data collection based on the adjusted operating status to obtain more accurate regional traffic dynamic data;

[0080] S18, performing resource load balancing scheduling based on the higher-precision regional traffic dynamic data to obtain a final resource allocation plan;

[0081] S19: Based on the final resource allocation plan, perform abnormal monitoring and dynamic adjustment to obtain the latest low-altitude traffic management status.

[0082] In step S11, multi-source heterogeneous data is obtained;

[0083] It's worth noting that the system acquires multi-source heterogeneous data, primarily consisting of two data streams: drone sensor data and ground monitoring data, originating from the drone platform and ground monitoring equipment. Within the drone sensor data, velocity data is acquired through the accelerometer and gyroscope in the drone's onboard inertial navigation unit. The system continuously reads velocity changes at a fixed sampling frequency (e.g., 50 Hz) and uses time integration to obtain the average velocity over a short period. Altitude data is measured using a laser altimeter or a barometric altimeter. For example, when flying in urban areas, a laser ranging module is used to determine the vertical distance to ground obstacles, thereby marking the drone's vertical altitude. Ground monitoring data originates from image acquisition devices deployed in traffic corridors, take-off and landing points, or low-altitude monitoring posts. The system captures video streams using cameras with a frame rate of at least 25 frames per second. Image recognition is achieved using methods based on edge detection and contour matching. The specific process is as follows: the system first uses a ground-based surveillance camera to capture continuous images, converts each frame into a grayscale image, and extracts image edges using the Canny algorithm. Based on the resulting edge image, the system further uses a contour extraction algorithm, such as the findContours function in OpenCV, to obtain closed boundary areas. Subsequently, the system pre-creates a boundary contour template for the aircraft based on typical aircraft features, such as the "rotor + fuselage" geometry. Template matching is performed on the contour results in the real-time image. By comparing indicators such as contour area, aspect ratio, and boundary shape, it determines whether an aircraft target exists in the current image and assigns a unique identification code to it as a target identifier. This is used to uniquely identify a specific aircraft or ground object. Position information is obtained by connecting the device to a GNSS receiver or a fused positioning device (such as an RTK module). The system synchronously extracts the latitude and longitude or plane coordinates of the identified target in each frame according to the timestamp.

[0084] In step S12, data standardization is performed on the multi-source heterogeneous data to obtain a multi-source data set, including:

[0085] The multi-source heterogeneous data includes: UAV sensor data and ground monitoring data;

[0086] Performing field mapping and structure conversion on the drone sensing data to obtain a sensor standard data set;

[0087] Extracting fields and regularizing formats based on the ground monitoring data to obtain a monitoring standard data set;

[0088] The sensor standard data set and the monitoring standard data set are integrated to obtain a multi-source data set in a unified format.

[0089] It is worth noting that the system first performs field mapping and structure conversion on the acquired drone sensor data to form a unified and standardized sensor standard data set. UAV sensor data is generally reported in the form of raw data packets, and the data structure contains multiple irregular fields, such as V x 、V y and Z alt And other physical quantities with units or different naming formats. In order to achieve data consistency and readability, the system pre-establishes a field mapping table to map these original fields to standard field names, such as "V x ” and “V y " is combined and calculated into the total speed field "speed", which is calculated by taking the square root of the sum of the speeds in both directions, i.e. speed is equal to V x The square of V y The square root of the sum of the squares. alt " field is directly mapped to the unified altitude field "altitude" and unit conversion is performed, for example, converting feet to meters. The structural conversion part integrates the timestamps, location numbers, and sensor numbers recorded separately in the original data into a unified data record row, with time as the primary index. Each row of records contains the speed, altitude, sensor number, and data source label at the corresponding moment. Taking a practical application as an example, when the drone device numbered U01 reports a speed of 15 meters per second and an altitude of 120 meters at 10:20 on May 1, 2024, the system converts this data into a row of structured records with the following fields: time 202405011020, number U01, speed 15, and altitude 120.

[0090] The system first extracts fields and formats the received ground monitoring data to construct a unified monitoring standard data set. Ground monitoring data comes from fixed cameras, radars, and image sensors. The original records contain time-indexed location information and target identification information. To ensure a clear data structure, the system must first parse the metadata attached to the monitoring image frame. For example, the original location information field format is "gps x and GPS y ", the target identifier is "track id", but the naming and format of these fields are often not uniform. The system will create a field mapping table and convert "gps x and GPS y " is merged and converted into a unified two-dimensional coordinate field "position", whose format is a floating point array of the form [120.154234,30.245678]. At the same time, "track id " field is renamed to "target id ", ensuring that all monitored objects have unique identification in the entire dataset. The system then aligns each record with a unified timestamp, such as sampling once per second, and organizes it into a standard structured record containing the acquisition time, location coordinates and target number.

[0091] Finally, the system aligns the sensor standard data group and the monitoring standard data group according to the timestamp field, matches the data records at the same time point one by one through traversal, and merges the speed, altitude, position and target number into one row to generate a unified format data set with consistent structure.

[0092] In step S13, dynamic identification and density assessment are performed based on the multi-source data set to obtain a traffic density distribution map, including:

[0093] Performing unit division and data extraction based on the multi-source data set to obtain a unit-level preliminary traffic data set, wherein the unit-level preliminary traffic data set includes speed data, height data, location information, and target identification;

[0094] Calculating the magnitude of change within each unit based on the speed data and the height data in the traffic data set;

[0095] If the change amplitude within the unit is greater than or equal to a preset amplitude threshold, the corresponding unit is marked as a high dynamic area;

[0096] If the amplitude of the change in the unit is less than a preset amplitude threshold, the unit is marked as a low dynamic area;

[0097] Integrating the high dynamic area and the low dynamic area to generate a dynamic distribution feature map;

[0098] A density distribution map is generated based on the dynamic distribution characteristic map, the target identifier and the location information to obtain a traffic density distribution map.

[0099] It is worth noting that the system first divides the monitoring area into a two-dimensional grid based on the geographic space range to construct a regular set of cells. The division granularity is generally based on a square with a length of 100 meters on each side, ensuring that each cell can cover multiple target points but is not too large to cause ambiguity. After the division is completed, the system uses each cell as the search area, traverses all records in the multi-source dataset, and extracts records that fall into the cell based on their location information field. Each record contains a speed value, an altitude value, a target number, and corresponding latitude and longitude coordinates.

[0100] The system then needs to calculate the magnitude of change within each unit, first calculating the speed change and the height change:

[0101]

[0102] ΔV is the change in speed, ΔH is the change in height, is the velocity value at time point t1, is the velocity value at time point t2, is the height value at time point t1, is the height value at time point t2.

[0103] The normalized weighted average method is used to calculate the change range:

[0104]

[0105] Among them, Δ u nit is the amplitude of change, α and β are weight coefficients, α=0.7, β=0.3, V max is the maximum velocity of all targets in all units in the current evaluation cycle, V min is the minimum speed value of all targets in all units in the current evaluation cycle; H max is the maximum height value that occurs in the same period, H min The two maximum and two minimum values ​​are obtained by traversing every record in the traffic dataset and comparing the speed and altitude fields. They can effectively reflect the global speed and altitude limits within this cycle.

[0106] The α and β weights are set so that changes in speed directly reflect the aircraft's propulsion state and maneuvering behavior, such as acceleration, deceleration, or sudden turns. These changes are accompanied by control commands or path adjustments, making them more sensitive to the dynamic nature of the flight area. Altitude changes, on the other hand, are limited by airspace control and terrain constraints in most low-altitude operating environments, resulting in a smaller range of variation. They are often used to assist in determining flight trends rather than directly reflecting dynamic behavior.

[0107] The preset amplitude threshold set here is 0.1, which effectively distinguishes between flight states with gentle altitude changes and those with drastic speed fluctuations. When the amplitude is greater than or equal to 0.1, the system marks the cell as a high-dynamic area, indicating that there is drastic movement. Conversely, if the amplitude is less than 0.1, the cell is considered a low-dynamic area with stable movement characteristics.

[0108] The system uniformly marks all units that have completed the change amplitude calculation, and assigns them "high dynamic" and "low dynamic" signs respectively, and then summarizes the marking information of high dynamic and low dynamic units to obtain a dynamic distribution feature map.

[0109] To further identify traffic aggregation, the system uses the units marked as highly dynamic in the dynamic distribution map as key statistical areas and performs target aggregation analysis in a fixed time window.

[0110] Specifically, the system extracts all target identifiers within each marked dynamic cell between 10:00:00 and 10:00:10, filtering out duplicates to count the number of active targets within the cell. For example, if target_ids A101, A102, A103, A104, and A105 are recorded within a block, the number of active targets for that cell during this time period is 5, which serves as the initial density value for that cell.

[0111] The system then performs a unified spatial mapping operation based on the location coordinates of each target marker. The system divides the two-dimensional area into equal-width grid cells using a preset spatial grid step size (e.g., 50 meters) and counts the number of target markers falling within each cell. This statistical value is the traffic density value; a higher density indicates a higher concentration of targets within the area and a more complex traffic situation.

[0112] In step S14, cluster identification analysis is performed based on the traffic density distribution map to obtain the boundary range of the key area, including:

[0113] Perform density screening according to the traffic density distribution map to obtain a high-density unit set;

[0114] Perform cluster identification based on the high-density unit set to obtain a key area identification set;

[0115] Based on the key area identification set, boundary extraction and expansion are performed to obtain the boundary range of the key area.

[0116] It is worth noting that the system first performs a density screening operation on all cells in the traffic density distribution map to identify high-density areas and provide an input basis for subsequent cluster analysis. The specific operation process is as follows: the system traverses all cells in the density distribution map, extracts their corresponding density values, and records them as a density sequence. Subsequently, the average density and standard deviation of all cells are calculated based on the density sequence. The average density represents the overall level of traffic activity in the entire map, and the standard deviation describes the degree of discreteness of the density of each cell. Based on this, the system sets the density screening threshold, specifically the average density plus one standard deviation, to define the "high-density" interval. This screening condition can effectively eliminate cells within the normal fluctuation range, ensuring that the retained cells have statistically significant density characteristics.

[0117] For example, if the average density of the entire map is 12 and the standard deviation is 3, the screening threshold is 15. The system marks cells with a density value greater than or equal to 15 as "high-density cells," while the remaining cells are marked as normal areas and excluded from subsequent steps. All cells that meet the screening criteria are grouped together into a high-density cell set, each of which contains its location coordinates in the original distribution map and its actual density value.

[0118] Based on the high-density cell set identified in the previous stage, the system performs cluster recognition, identifying clusters of areas with close proximity and continuous traffic flow. The specific operation is as follows: First, the system calculates the spatial distance between each high-density cell and its other high-density cells based on the center coordinates of each cell, and groups cells within a certain distance threshold into one group. The distance threshold can be determined based on the grid scale. For example, if the system uses a cell side length of 50 meters, the neighboring cell distance threshold is set to 70 meters to ensure that diagonally adjacent cells are included.

[0119] The system uses a one-by-one traversal method to retrieve other high-density units in its vicinity for each unmarked high-density unit, and uniformly assigns a cluster number to all units that meet the conditions. This process continues to iterate until all high-density units are classified into a cluster. Taking a certain area as an example, if the coordinates of unit A are (100,100) and unit B is (130,120), the Euclidean distance between the two is 36 meters, which is less than the threshold of 70 meters, so they are classified into the same cluster. Through this operation, the system eventually identifies several spatially continuous and densely populated areas, and each cluster represents a key area.

[0120] Each cluster corresponds to a key area identifier, which the system numbers as Z1, Z2, Z3, etc., and records the coordinate range of the cells it contains. This set of key area identifiers provides a clear geographical scope basis for subsequent boundary extraction and scheduling decisions.

[0121] Based on the identified key area identification set, the system extracts and expands the boundaries of the cell set in each key area to clarify the spatial boundary range of the key area. Specifically, the system first reads the coordinate data of all member cells in each key area and constructs its boundaries using the minimum enclosing rectangle method. This method determines a closed rectangular box by finding the minimum horizontal and vertical coordinate values ​​and the maximum horizontal and vertical coordinate values ​​that contain all cells. For example, if the horizontal coordinate range of the cells contained in a key area is 100 to 150 and the vertical coordinate range is 200 to 240, then its initial bounding box is defined as the upper left corner of the rectangle is (100,240) and the lower right corner is (150,200).

[0122] Subsequently, the system expands the boundary range. The expansion operation refers to extending the original minimum circumscribed rectangle outward by a certain distance to cover the possible interference units or critical units at the edge of the area, to avoid the area recognition range being too narrow due to instantaneous density fluctuations. For example, if the expansion boundary is set to 10 meters, the width of the unit will be expanded to the left, right, top, and bottom based on the above rectangle. After the expansion, the boundary range of the key area is updated to 90 to 160 on the horizontal coordinate and 190 to 250 on the vertical coordinate.

[0123] Ultimately, each key area is defined as a closed area with a clear coordinate range. The system records and numbers its boundary range in the form of a coordinate set. For example, the boundary range of key area Z1 is 90 to 160 horizontally and 190 to 250 vertically.

[0124] In step S15, the fence boundary is updated according to the boundary range of the key area to obtain updated fence constraint conditions, including:

[0125] Obtaining boundary coordinate data of the key area;

[0126] Generate a vector fence based on the boundary range coordinate data of the key area to obtain a preliminary fence boundary shape;

[0127] Performing boundary expansion processing according to the preliminary fence boundary shape to obtain an expanded fence boundary shape;

[0128] Based on the expanded fence boundary shape, boundary confirmation synchronization is performed to obtain updated fence constraint conditions.

[0129] It is worth noting that the system retrieves the corresponding cell position coordinates one by one based on the key area identification set extracted in the previous step, and combines them to form the boundary range coordinate data. The boundary range of the key area refers to a closed area composed of multiple adjacent high-density cells, and its spatial distribution is described by the coordinate points of the outermost cells. The system first extracts the two-dimensional position index information of each key area cell, for example, expressed as an X and Y coordinate pair in the form of the coordinates of the upper left corner of the grid. Subsequently, the system sequentially connects this set of coordinate points to obtain a set of boundary coordinate points. Taking a key area containing five adjacent cells as an example, its boundary coordinate points are [(100,200), (100,250), (150,250), (150,200)], and this data set is the boundary range coordinate data of the current key area.

[0130] To convert the boundary coordinate data into a continuous, closed boundary line, the system uses a contour stitching algorithm to sequentially connect the coordinate points to form one or more closed polygons. This algorithm determines the line relationships between adjacent points based on the order of the coordinate points and constructs a polygon path when the end-to-end closure conditions are met.

[0131] Once constructed, the system stores the resulting boundary line data in the fence boundary array. Each boundary line is recorded as a sequence of line segments, representing the initial physical blockade of the area. For example, if the boundary coordinate point set is [(100, 200), (100, 250), (150, 250), (150, 200)], the system generates a rectangular boundary wireframe, corresponding to a preliminary fence shape of 50 meters in width and 50 meters in height.

[0132] Based on the generated preliminary fence boundary shape, the system performs boundary extension processing. This aims to maintain safety margins in critical areas and prevent traffic fluctuations near the boundary from affecting the original boundary. The expansion process operates as follows: the system first reads all boundary coordinates of the preliminary fence and performs a buffer calculation on each coordinate point in turn. The buffer calculation involves shifting each point outward a certain distance along its normal direction to form a new extended edge point. The extension distance is preset by the system and is set to ten meters based on the mission requirements.

[0133] The system then reconnects all the outward expansion points into a closed path and, through geometric Boolean operations, merges the area between the original boundary polygon and the outward expansion path to form a new polygonal boundary, referred to as the expanded fence boundary shape. For example, if the original boundary is a rectangle with four corner points ABCD, the system will translate each of these four points outward by ten meters to obtain A'B'C'D', and then reconstruct the expanded bounding box with these new points. This expansion result spatially covers the initial critical area and leaves a buffer zone, effectively improving the fence's tolerance and control capabilities for sudden traffic expansion. The final output of the expanded boundary is still a sequence of polygonal coordinate points.

[0134] After extracting and organizing the historical data for the fenced area, the system calculates and sets specific operating thresholds for each expanded fenced boundary area. Thresholds are important reference benchmarks used to limit the target's operating behavior within the area, and include speed and altitude thresholds. The system first extracts the operating speed and altitude of each target within the area based on all historical trajectory records and normalizes each record to ensure a consistent data structure. The system then uses statistical analysis methods to sort all speed data and selects the 90th percentile of speed values ​​as the maximum speed threshold, ensuring that this threshold covers most normally operating targets while excluding extreme outliers.

[0135] For example, if the extracted speed data for a certain area is an array containing 300 records, and the 270th speed value after sorting is 13.6 meters per second, the speed threshold for that area is set to 13.6 meters per second. Similarly, the system performs the same processing on the altitude data, setting the altitude threshold to the 90th percentile altitude value of the target in that area over the past period, for example, 115 meters. This approach ensures that the resulting threshold is both statistically reasonable and operationally safe, avoiding overly restrictive or overly permissive behavior. Finally, each updated fence constraint binds the complete control structure of the speed and altitude thresholds.

[0136] In step S16, the flight parameters are modified according to the updated fence constraint conditions to obtain an adjusted operating state, including:

[0137] Acquiring speed parameters and altitude parameters of the sensor standard data set;

[0138] The updated fence constraint conditions include: a speed threshold and a height threshold;

[0139] According to the speed parameter and the altitude parameter, the speed threshold and the altitude threshold are compared and analyzed respectively and an instruction is generated, thereby obtaining an aircraft control instruction;

[0140] According to the aircraft control instructions, parameters are issued and operation corrections are made to obtain an adjusted operation state.

[0141] It's worth noting that the system first extracts the aircraft's current speed and altitude parameters from a standard sensor data set. These parameters are typically provided by the drone's inertial measurement unit and barometric altimeter, sampled once per second. The data fields are speed and altitude. This real-time data serves as the basis for determining the aircraft's current operating status.

[0142] The system then loads the fence constraints generated during the previous boundary verification synchronization phase. These constraints include speed and height thresholds for a specific area. For example, a fenced area may have a speed threshold of 15 meters per second and a height threshold of 120 meters.

[0143] Next, the system performs a numerical comparison and analysis of the current speed and altitude parameters. If the current speed is greater than the threshold, the aircraft may be at risk of rapid crossing, and the system will generate a "reduce speed" control command. If the speed is below the threshold, no adjustment is required. If the current altitude is greater than the set threshold, it indicates that the aircraft is flying too high and there is a potential risk of exceeding airspace restrictions. The system will generate a "lower altitude" command. If the altitude is below the threshold, it may also indicate that the aircraft is flying too low and entering a dangerous area, and the system will also generate a "raise altitude" command.

[0144] Ultimately, the system generates comprehensive aircraft control instructions based on the above judgment results. For example: if the speed is 18 meters per second and the altitude is 130 meters, both exceeding the thresholds at the same time, the system will simultaneously issue deceleration and altitude reduction instructions to correct the flight status and return it to a safe operating range.

[0145] The correction process operates within a closed feedback loop. The system monitors the aircraft's speed and altitude in real time to determine whether the control effect meets the target state. For example, if, within one second of issuing a command, the system detects a decrease in speed from 18 meters per second to 14.8 meters per second and a decrease in altitude from 130 meters to 121 meters, the flight state is considered to have been effectively corrected.

[0146] The adjusted operating status means that all parameters of the aircraft are within the fence setting range, specifically the speed does not exceed 15 meters per second, the altitude is stable below 120 meters, and there is no sudden change trend.

[0147] In step S17, traffic determination and data collection are performed based on the adjusted operating state to obtain more accurate regional traffic dynamic data, including:

[0148] Performing safety parameter comparison and traffic determination according to the adjusted operating state to obtain a traffic determination result;

[0149] According to the access determination result, permission mapping and allocation operations are performed to obtain effective access permissions;

[0150] Based on the effective traffic rights, priority road sections are screened and nodes are planned to obtain a node deployment plan;

[0151] According to the node deployment plan, data integration and frequency evaluation are performed to obtain more accurate regional traffic dynamic data.

[0152] It is worth noting that based on the adjusted operating status obtained in the previous stage, the system verifies the safety of the aircraft's current speed, altitude and other key operating parameters, and judges whether it has the conditions to continue to pass. Specifically, the system first retrieves the speed threshold and altitude threshold set for each aircraft in the fenced area, and compares them item by item with the real-time speed and altitude values ​​reported by the aircraft sensors. The comparison process is carried out in the form of difference judgment: if the current speed is less than the set speed limit and the altitude is within the allowed altitude range, the system will preliminarily determine that the operating status is "passable". Otherwise, if any parameter exceeds the preset amplitude threshold, the system will mark it as "restricted passage".

[0153] For example, if a vehicle's current speed is 12 meters per second and the fenced area's upper speed limit is 15 meters per second, its speed parameters meet the pass criteria. If its altitude is 90 meters, and the area's altitude range is 80 to 100 meters, it also meets the altitude requirement, and the system will generate a "pass" result. Conversely, if the altitude exceeds the upper limit or the speed is too high, a "no pass" result will be returned.

[0154] Based on the access determination results generated in the previous phase, the system performs permission mapping and allocation operations to ensure that each aircraft has legal access to the path during actual operation. The access determination result is a Boolean judgment information, represented by two types of labels: "Access allowed" or "Access prohibited", which indicates whether the system determines whether the aircraft's current operating status meets the dynamic control conditions of the fenced area.

[0155] The specific operation process is as follows: the system first reads the access determination result and establishes a mapping relationship between the determination result and the permission level. If the determination result is "access allowed", the system assigns the aircraft "Level 1 access permission", allowing it to execute tasks within the current area according to the default path or scheduling strategy. If the determination result is "access denied", the system downgrades its permission to "restricted access" or "pending scheduling" and restricts its access to certain sensitive nodes or high-density areas at the mission level to prevent resource conflicts or the spread of flight risks.

[0156] During the permission mapping process, the system generates a permission mapping table using the aircraft ID as an index. It then assigns specific access ranges based on the current fence position, flight time period, and mission type. For example, if aircraft U01 is determined to be allowed access in a certain area, it will be granted valid access within "key_area_A, 10:00–10:10." This permission information is synchronously recorded in the mission scheduling table and edge node cache for subsequent path planning and data reporting.

[0157] During the priority road section screening process, the system will perform multi-dimensional indicator calculations on each grid channel in the traffic area, mainly including three core indicators: historical traffic frequency, node density and task concentration. First, the system counts the target number of passes of each grid within the set time window to obtain the traffic frequency parameter. The higher the frequency, the more mature the channel is and the more stable its traffic capacity. Secondly, the system calculates the number of perception and communication nodes in the coverage area of ​​each channel, and uses the ratio of the number of nodes to the coverage area as the node density indicator to reflect its support capacity. Thirdly, the system counts the number of scheduling tasks carried by each grid in the current period to form a task concentration, which is used to describe the spatial aggregation of tasks.

[0158] After obtaining the above parameters, the system integrates the metrics of the grids spanned by each candidate path, calculating the mean frequency of travel, mean node density, and total task concentration. These three metrics are then normalized and integrated using set weights to output the path's overall traffic load. The system also analyzes the number and density of congestion records for the grids within the path, combining historical trajectory records, to calculate the congestion probability of the path. Ultimately, the system prioritizes the paths based on load and congestion probability, selecting the top paths with the lowest load, balanced node distribution, and reasonable task concentration as priority passages.

[0159] Taking aircraft U01 as an example, if it is authorized to pass through area R5, the system will prioritize screening channels within R5 with low historical congestion rates, high mission target density, and low path overlap rates. For example, it will select three paths, R5-A1, R5-A3, and R5-B2, to construct a preferred route with low interference and high traffic efficiency.

[0160] Node planning involves setting up physical nodes for data collection, status synchronization, or communication relay for each preferred path. Based on the path length and the aircraft's communication radius, the system calculates the minimum number of nodes required to cover the entire path and evenly distributes them based on geographic coordinates. For example, if the aircraft's communication radius is 60 meters for a 300-meter road section, the system will deploy five communication nodes at intervals along the section to ensure uninterrupted information flow.

[0161] The final output node deployment plan is a set of binding relationships between paths and node numbers, in the form of: {path: R5-A1, node: [N21, N22, N23]}, which provides key support for subsequent data collection scheduling and communication collaboration.

[0162] Based on the previously generated node deployment plan, the system integrates and processes the data collected by each node and performs traffic frequency analysis based on the sampling frequency. The goal is to obtain more fine-grained, updated, and spatially comprehensive regional traffic dynamics data. Each monitoring node included in the node deployment plan has clear location information and collection range. After deployment, it will continuously collect information such as target trajectory, speed, direction, and identification tags within the area. The system first unifies the structure of this data, including field format standardization, timestamp alignment, and multi-source data merging, to form a clearly structured and temporally continuous data stream collection.

[0163] During the frequency assessment process, the system counts the number of target detections at each node within a set time window (for example, 1 second or 2 seconds), and further calculates the frequency and duration of the target within the perception range of the node to reflect the fluctuation of local traffic density. For example, if a node captures the target trajectory 20 times continuously from 10:00:00 to 10:00:10, and each interval is less than 1 second, it means that the node is in a high-traffic active state, and its data is highly timely and representative. The system further integrates the statistical results of adjacent nodes, and based on position interpolation and time interpolation methods, fills in coverage blind spots or sampling gaps to generate a continuous and complete dynamic data grid.

[0164] The final "higher-precision regional traffic dynamics data" refers to gridded traffic situation data with a sampling period of seconds and a spatial resolution of fifty meters. It not only includes the number of targets, average speed, and center of gravity in each time slice, but also includes path distribution density and movement direction trends.

[0165] In step S18, resource load balancing scheduling is performed based on the higher-precision regional traffic dynamic data to obtain a final resource allocation plan, including:

[0166] Perform load detection and list generation based on the higher-precision regional traffic dynamic data to obtain a node load status list;

[0167] Perform task migration scheduling according to the node load status list to obtain load distribution status;

[0168] According to the load distribution status, the scheme is synchronously deployed to obtain the final resource allocation scheme.

[0169] It's worth noting that the system performs load detection and list generation based on already acquired, higher-precision regional traffic dynamics data. This aims to identify the current load intensity of each operating node, providing fundamental support for subsequent task scheduling. This higher-precision traffic dynamics data refers to gridded target status data, collected with a sampling period of seconds and a spatial resolution of 50 meters. This data covers the number of targets, average speed, concentration, and path direction within each time slice. The system first analyzes the spatial grid covered by each node, extracting target frequency, average dwell time, and task accumulation indicators for that node within the current time window. The system then normalizes this extracted data, normalizing each indicator to a range of 0 to 1, and then performs a weighted fusion based on pre-set weights. Specifically, the system sets a weight of 0.5 for target number, 0.3 for dwell time, and 0.2 for task density, respectively, to represent their contribution to different dimensions of the node's load composition. Weighted fusion involves multiplying these three normalized indicators by their corresponding weights and summing the results to calculate the node's overall load score. The higher the score, the more tasks the node currently carries, the higher the target concentration, and the heavier the load. Ultimately, the system sets a grading threshold based on the score value, divides all running nodes into three categories: high load, medium load, and low load, and generates a structured node load status list. The list clearly indicates the number, spatial location, load level, and score value of each node to guide subsequent task migration and resource allocation operations. For example, the node numbered N105 received 12 independent targets within 60 seconds, with an average residence time of 35 seconds and a task density index higher than 0.8. The system identifies it as a high-load node and records its detailed indicators and location coordinates in the list.

[0170] Based on the location information and task attributes of each node recorded in the list, the system selects multiple candidate target nodes within a set distance radius around the source node. The screening rules prioritize nodes with low current loads, while also evaluating whether their communication distance is acceptable and whether they have the resource capacity to receive the corresponding task type. For each pair of task source and target nodes, the system performs a preliminary matching of tasks based on the principle of shortest path priority and the node's remaining receiving capacity. It then updates the task list of each node to determine the load distribution status.

[0171] During synchronous deployment, the system first reads the task acceptance records for all target nodes in the state table and generates scheduling instructions based on the scheduling policy for each path. These instructions include the task source node number, target node number, task summary, and planned execution time. The system then issues these scheduling instructions one by one to the node control units, following the task scheduling time sequence and path order, to implement synchronous deployment. The resulting resource allocation plan includes the task number, source path, expected execution time period, and required resource amount for each node.

[0172] In step S19, according to the final resource allocation plan, abnormal monitoring and dynamic adjustment are performed to obtain the latest low-altitude traffic management status, including:

[0173] According to the final resource allocation plan, data collection and density calculation are performed to obtain the traffic density of each area;

[0174] If the traffic density is greater than or equal to a preset density threshold, the abnormal time point is recorded and the corresponding area location information is obtained to obtain the initial distribution information of the abnormal fluctuation;

[0175] If the traffic density is less than the preset density threshold, the corresponding area is judged to be in a normal state and fluctuation detection is continued;

[0176] According to the initial distribution information of the abnormal fluctuation, dynamic range identification and update instruction generation are performed to obtain updated traffic situation map and fence boundary data;

[0177] According to the updated traffic situation map and fence boundary data, synchronous publishing and status confirmation are performed to obtain the latest low-altitude traffic management status.

[0178] It's worth noting that data collection focuses on the number of target appearances, target locations, and corresponding spatial grid numbers within each area within a unit time period. This collection frequency is performed on a per-second basis, with a 50-meter grid resolution. The number of targets appearing in each grid is counted once per second. The system accumulates the target appearance frequency within each grid within the current time period and records the statistical time period and grid area.

[0179] Traffic density is calculated using the following formula:

[0180]

[0181] Where ρ is the traffic density, in units of targets / km² / minute, N is the cumulative number of targets, A is the spatial area of ​​the corresponding grid, in units of square meters, and T is the length of the statistical period, in seconds.

[0182] Taking the R3 area as an example, if the area is 2500 square meters and a total of 2 targets are detected within a continuous statistical period of 5 seconds, the traffic density of the area can be calculated to be about 9600 / km2·min. In order to effectively identify abnormal traffic concentration areas, the system sets the traffic density judgment threshold to 8000 / km 2 min. This threshold is based on the average traffic density of typical routes during peak hours, and is highly representative and practical. When the density value of an area is greater than or equal to this threshold, the system identifies it as an area of ​​excessive traffic density, automatically triggering the subsequent dynamic identification, boundary extraction, and fence update processes. Conversely, if the density value is below this threshold, the area is considered stable and no adjustments are performed.

[0183] When the traffic density value in a certain area is greater than or equal to the threshold, the system immediately records the time of the anomaly and extracts the two-dimensional spatial coordinates corresponding to the grid. Through the spatial index structure, the system can quickly locate the specific position of the area in the entire low-altitude grid. The system then combines this time point with the corresponding location information to form an initial distribution record of the abnormal fluctuation, which serves as the input basis for subsequent dynamic identification and boundary updates. This information is the "abnormal fluctuation initial distribution information" and includes the unique number of the abnormal area, the time of occurrence, the traffic density value, and the spatial location coordinates.

[0184] If the traffic density value in a certain area is less than the preset density threshold, it indicates that the current traffic status in that area is within the normal range. The system will not immediately adjust the boundaries of the area or reallocate resources, but will continue to monitor the traffic density changes at a sampling interval of 5 seconds. If the fluctuation threshold is not reached within multiple consecutive sampling periods, the area will remain labeled "Normal Area" to avoid misjudgment due to short-term occasional changes.

[0185] After reading the initial distribution information of abnormal fluctuations, the system first constructs an analysis window around the abnormal area, sets the spatial buffer range to 50 meters to 100 meters, and retrieves the adjacent grid nodes within this range. Then, a traffic density review operation is performed on each adjacent grid to determine whether it has reached or approached the fluctuation threshold within the same time window. For example, taking area R7 as the initial point of abnormality, the system will sequentially detect the eight adjacent grids R6, R8, R12, etc. around it. If there are grid nodes with a density value greater than 8000, they will be included in the same dynamic area, eventually forming a continuous or nearly continuous set of abnormal areas. This set is the "dynamic identification range."

[0186] During the update of fence boundary data, the system uses a contour extraction algorithm to detect closed contours on the edge grids of the dynamically identified area. This process involves three steps: First, the system constructs a binary matrix based on the identified high-density abnormal grid set, marking abnormal areas as 1 and other areas as 0. Second, it scans row by row, starting from the upper left corner, to find the boundary grid, and then traverses the adjacent grids point by point in a clockwise direction, recording the coordinates of the boundary intersections. Third, when the trajectory returns to the starting point and forms a complete closed loop, it outputs the coordinate sequence of the boundary grid, completing the contour extraction.

[0187] The coordinate sequence generated by the contour extraction operation will serve as the core component of the "fence boundary data". Its data content includes: first, the two-dimensional spatial coordinates of the boundary grid points, with the upper left corner as the origin, and the unit is meters; second, the boundary direction information, which is used to indicate the topological continuity of the fence boundary and facilitate subsequent area closure processing; third, the unique identification code of the abnormal area, such as Zone_202405_T103, which is used to mark the source and timestamp of the anomaly detection corresponding to the fence.

[0188] In the process of generating the latest traffic situation map, the system performs grid-level situation fusion and layer rendering operations based on the initial distribution information of abnormal fluctuations and the updated traffic density data. First, the system integrates the data such as the number of targets, average speed and movement direction of each grid in the latest time slice, and overwrites the information of the corresponding grid in the updated history map, where high-density areas are represented by darker color gradients, the movement direction is marked with arrows, and abnormal fluctuation areas are highlighted with red frames. Subsequently, the system completes the layer rendering through the map service interface, superimposing the fused traffic density matrix with the fence boundary, and dividing it into a geographic base map, a normal situation layer and an abnormal area annotation layer according to the layer structure. The final traffic situation map is connected to the user end in the form of a web page, allowing dispatchers to view the traffic conditions, abnormal changes and fence adjustment results of each area in real time.

[0189] After the traffic situation map and fence boundary data are updated, the system immediately performs synchronization and status confirmation to ensure the timely delivery and adoption of the latest low-altitude traffic control information. First, the system packages the updated traffic situation map and corresponding fence boundary data into a structured configuration file. The format includes grid numbers, real-time traffic density values, a sequence of fence boundary coordinate points, and a fluctuation zone identification field. The system then pushes this configuration file to the dispatch terminal, command terminal, and mission execution devices via the data distribution interface. Each node then parses and completes the local policy replacement. During the status confirmation phase, the system receives confirmation signals and configuration checksums from each device to determine whether the new data has been loaded and implemented. If any nodes have inconsistent configurations or delayed responses, a secondary synchronization command will be triggered. For example, if a drone execution device does not return a configuration update confirmation message within the set five seconds, the system will automatically switch to a backup channel to retransmit the fence coordinates and situation data for the mission area. Finally, based on the feedback from all nodes, the system confirms that the current operating environment for the entire area has been updated, thereby obtaining the latest low-altitude traffic management status. This state refers to the unified operating configuration adapted by all flight equipment, monitoring nodes and scheduling strategies in the system, ensuring consistency in command distribution, fence restrictions and access strategies.

[0190] Reference Figure 2 The second embodiment of the present invention provides a multi-source data-driven low-altitude operation intelligent scheduling system, including:

[0191] Data acquisition module, used to acquire multi-source heterogeneous data;

[0192] A standardization module, configured to perform data standardization processing on the multi-source heterogeneous data to obtain a multi-source data set;

[0193] A traffic density module is used to perform dynamic identification and density assessment based on the multi-source data set to obtain a traffic density distribution map;

[0194] A boundary range module is used to perform cluster identification analysis based on the traffic density distribution map to obtain the boundary range of the key area;

[0195] A constraint condition module, configured to update the fence boundary according to the boundary range of the key area to obtain an updated fence constraint condition;

[0196] a correction module, configured to correct flight parameters according to the updated fence constraint conditions to obtain an adjusted operating state;

[0197] A dynamic data module is used to perform traffic determination and data collection based on the adjusted operating status to obtain more accurate regional traffic dynamic data;

[0198] An allocation plan module is used to perform resource load balancing scheduling based on the higher-precision regional traffic dynamic data to obtain a final resource allocation plan;

[0199] The management status module is used to perform abnormal monitoring and dynamic adjustment according to the final resource allocation plan to obtain the latest low-altitude traffic management status.

[0200] It should be noted that the multi-source data-driven low-altitude operation intelligent scheduling system provided in an embodiment of the present invention is used to execute all the process steps of the multi-source data-driven low-altitude operation intelligent scheduling method of the above embodiment. The working principles and beneficial effects of the two correspond one to one, so they will not be repeated here.

[0201] An embodiment of the present invention further provides an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a management status program. When the processor executes the computer program, the steps in the above-mentioned embodiments of the multi-source data-driven low-altitude operation intelligent scheduling method are implemented, such as Figure 1 Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are realized, such as the management status module.

[0202] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of the computer program in the electronic device.

[0203] The electronic device may be a computing device such as a desktop computer, notebook, PDA, or smart tablet. The electronic device may include, but is not limited to, a processor and memory. Those skilled in the art will appreciate that the aforementioned components are merely examples of electronic devices and do not constitute a limitation of the electronic device. The electronic device may include more or fewer components than those described above, or a combination of certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, and the like.

[0204] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the electronic device, connecting various parts of the entire electronic device using various interfaces and lines.

[0205] The memory can be used to store the computer programs and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created based on the use of the mobile phone (such as audio data, a phone book, etc.). In addition, the memory can include a high-speed random access memory and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0206] Wherein, if the module / unit integrated in the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by a computer program to instruct the relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, it can implement the steps of each of the above-mentioned method embodiments. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0207] It should be noted that the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive effort.

[0208] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A multi-source data-driven low-altitude operation intelligent scheduling method, characterized in that: include: Acquire multi-source heterogeneous data; Performing data standardization processing on the multi-source heterogeneous data to obtain a multi-source data set; Performing dynamic identification and density assessment based on the multi-source data set to obtain a traffic density distribution map; Perform cluster identification analysis based on the traffic density distribution map to obtain the boundary range of the key area; According to the boundary range of the key area, the fence boundary is updated to obtain the updated fence constraint condition; Correcting flight parameters according to the updated fence constraint conditions to obtain an adjusted operating state; According to the adjusted operating status, traffic determination and data collection are performed to obtain more accurate regional traffic dynamic data; Perform resource load balancing scheduling based on the higher-precision regional traffic dynamic data to obtain a final resource allocation plan; According to the final resource allocation plan, abnormal monitoring and dynamic adjustment are carried out to obtain the latest low-altitude traffic management status.

2. The multi-source data-driven low-altitude operation intelligent scheduling method according to claim 1 is characterized in that: The step of performing data standardization processing on the multi-source heterogeneous data to obtain a multi-source data set includes: The multi-source heterogeneous data includes: UAV sensor data and ground monitoring data; Performing field mapping and structure conversion on the drone sensing data to obtain a sensor standard data set; Extracting fields and regularizing formats based on the ground monitoring data to obtain a monitoring standard data set; The sensor standard data set and the monitoring standard data set are integrated to obtain a multi-source data set in a unified format.

3. The multi-source data-driven low-altitude operation intelligent scheduling method according to claim 1 is characterized in that: The method of performing dynamic identification and density assessment based on the multi-source data set to obtain a traffic density distribution map includes: Performing unit division and data extraction based on the multi-source data set to obtain a unit-level preliminary traffic data set, wherein the unit-level preliminary traffic data set includes speed data, height data, location information, and target identification; Calculating the magnitude of change within each unit based on the speed data and the height data in the traffic data set; If the change amplitude within the unit is greater than or equal to a preset amplitude threshold, the corresponding unit is marked as a high dynamic area; If the amplitude of the change in the unit is less than a preset amplitude threshold, the unit is marked as a low dynamic area; Integrating the high dynamic area and the low dynamic area to generate a dynamic distribution feature map; A density distribution map is generated based on the dynamic distribution characteristic map, the target identifier and the location information to obtain a traffic density distribution map.

4. The multi-source data-driven low-altitude operation intelligent scheduling method according to claim 1 is characterized in that: The cluster identification analysis is performed based on the traffic density distribution map to obtain the boundary range of the key area, including: Perform density screening according to the traffic density distribution map to obtain a high-density unit set; Perform cluster identification based on the high-density unit set to obtain a key area identification set; Based on the key area identification set, boundary extraction and expansion are performed to obtain the boundary range of the key area.

5. The multi-source data-driven low-altitude operation intelligent scheduling method according to claim 1 is characterized in that: The updating of the fence boundary according to the boundary range of the key area to obtain the fence constraint condition includes: Obtaining boundary coordinate data of the key area; Generate a vector fence based on the boundary range coordinate data of the key area to obtain a preliminary fence boundary shape; Performing boundary expansion processing according to the preliminary fence boundary shape to obtain an expanded fence boundary shape; Based on the expanded fence boundary shape, boundary confirmation synchronization is performed to obtain updated fence constraint conditions.

6. The multi-source data-driven low-altitude operation intelligent scheduling method according to claim 2 is characterized in that: The step of modifying the flight parameters according to the updated fence constraint conditions to obtain an adjusted operating state includes: Acquiring speed parameters and altitude parameters of the sensor standard data set; The updated fence constraint conditions include: a speed threshold and a height threshold; According to the speed parameter and the altitude parameter, the speed threshold and the altitude threshold are compared and analyzed respectively and an instruction is generated, thereby obtaining an aircraft control instruction; According to the aircraft control instructions, parameters are issued and operation corrections are made to obtain an adjusted operation state.

7. The multi-source data-driven low-altitude operation intelligent scheduling method according to claim 1 is characterized in that: The traffic determination and data collection are performed according to the adjusted operating state to obtain more accurate regional traffic dynamic data, including: Performing safety parameter comparison and traffic determination according to the adjusted operating state to obtain a traffic determination result; According to the access determination result, permission mapping and allocation operations are performed to obtain effective access permissions; Based on the effective traffic rights, priority road sections are screened and nodes are planned to obtain a node deployment plan; According to the node deployment plan, data integration and frequency evaluation are performed to obtain more accurate regional traffic dynamic data.

8. The multi-source data-driven low-altitude operation intelligent scheduling method according to claim 1 is characterized in that: The method of performing resource load balancing scheduling based on the higher-precision regional traffic dynamic data to obtain a final resource allocation solution includes: Perform load detection and list generation based on the higher-precision regional traffic dynamic data to obtain a node load status list; Perform task migration scheduling according to the node load status list to obtain load distribution status; According to the load distribution status, the scheme is synchronously deployed to obtain the final resource allocation scheme.

9. The multi-source data-driven low-altitude operation intelligent scheduling method according to claim 1 is characterized in that: According to the final resource allocation plan, abnormal monitoring and dynamic adjustment are performed to obtain the latest low-altitude traffic management status, including: According to the final resource allocation plan, data collection and density calculation are performed to obtain the traffic density of each area; If the traffic density is greater than or equal to a preset density threshold, the abnormal time point is recorded and the corresponding area location information is obtained to obtain the initial distribution information of the abnormal fluctuation; If the traffic density is less than the preset density threshold, the corresponding area is judged to be in a normal state and fluctuation detection is continued; According to the initial distribution information of the abnormal fluctuation, dynamic range identification and update instruction generation are performed to obtain updated traffic situation map and fence boundary data; According to the updated traffic situation map and fence boundary data, synchronous publishing and status confirmation are performed to obtain the latest low-altitude traffic management status.

10. A multi-source data-driven low-altitude operation intelligent scheduling system, characterized in that: include: Data acquisition module, used to acquire multi-source heterogeneous data; A standardization module, configured to perform data standardization processing on the multi-source heterogeneous data to obtain a multi-source data set; A traffic density module is used to perform dynamic identification and density assessment based on the multi-source data set to obtain a traffic density distribution map; A boundary range module is used to perform cluster identification analysis based on the traffic density distribution map to obtain the boundary range of the key area; A constraint condition module, configured to update the fence boundary according to the boundary range of the key area to obtain an updated fence constraint condition; a correction module, configured to correct flight parameters according to the updated fence constraint conditions to obtain an adjusted operating state; A dynamic data module is used to perform traffic determination and data collection based on the adjusted operating status to obtain more accurate regional traffic dynamic data; An allocation plan module is used to perform resource load balancing scheduling based on the higher-precision regional traffic dynamic data to obtain a final resource allocation plan; The management status module is used to perform abnormal monitoring and dynamic adjustment according to the final resource allocation plan to obtain the latest low-altitude traffic management status.

Images