A low-altitude economic flight data management method and system based on edge computing

Abstract

The present invention belongs to the technical field of data management, and discloses a low-altitude economic flight data management method and system based on edge computing; the method comprises: establishing a distributed edge computing network, the distributed edge computing network is composed of a plurality of edge nodes and a central server, each edge node is responsible for collecting low-altitude economic flight data from an aircraft; the low-altitude economic flight data in all edge nodes are divided into n data blocks; each data block in the edge node is integrally encoded to obtain an encoded data block, and a flight data integral value corresponding to the encoded data block is calculated; the edge node uploads the encoded data block and the flight data integral value to the central server, and the central server manages and schedules the encoded data block of the edge node according to the flight data integral value, thereby providing strong support for the planning, scheduling and management of low-altitude economic flight.

Description

Technical Field

[0001] The present invention relates to the field of data management technology, and more specifically, to a low-altitude economic flight data management method and system based on edge computing. Background Art

[0002] At present, with the vigorous development of the low-altitude economic flight industry, a large number of drones, cargo drones and other aircraft have been put into use, generating massive amounts of low-altitude economic flight data; however, the existing data management methods have many drawbacks and shortcomings, which seriously restrict the development of low-altitude economic flight.

[0003] First, the traditional centralized data management method is inefficient. Due to the huge amount of low-altitude economic flight data, all of them are aggregated to the central server for processing, which not only increases the burden on the central server, but also leads to high latency and bandwidth occupancy during data transmission, which cannot meet the urgent real-time needs of low-altitude economic flight. Secondly, the direct transmission and storage of massive amounts of raw data not only wastes precious network bandwidth resources, but also increases storage pressure, resulting in high operating costs. Furthermore, the existing methods fail to distinguish the importance of data well, and cannot reasonably allocate computing resources, which may lead to delays in key data processing. In addition, the lack of a scientific and reasonable edge node deployment strategy makes it impossible to achieve complete coverage of the target area. Low-altitude economic flight covers a wide area, and the number of aircraft, flight duration, geographical environment, etc. in the area are unevenly distributed. A single deployment method is difficult to take into account, resulting in the existence of blind spots in data collection, affecting the integrity and accuracy of the data.

[0004] In view of this, the present invention proposes a low-altitude economic flight data management method and system based on edge computing to solve the above problems. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solutions: A low-altitude economic flight data management method based on edge computing, comprising: S1, establishing a distributed edge computing network, the distributed edge computing network consists of a number of edge nodes and a central server, each edge node is responsible for collecting low-altitude economic flight data from an aircraft;

[0006] S2, divide the low-altitude economic flight data in all edge nodes into n data blocks;

[0007] S3. Perform integral encoding on each data block in the edge node to obtain an encoded data block, and calculate the flight data integral value corresponding to the encoded data block; the edge node uploads the encoded data block and the flight data integral value to the central server, and the central server manages and schedules the encoded data block of the edge node according to the flight data integral value.

[0008] Furthermore, the method of establishing a distributed edge computing network includes:

[0009] Determine the data collection area, and determine the number of edge nodes to be deployed based on the size of the data collection area and the number of aircraft; determine the geographic boundaries of the data collection area, and divide the entire data collection area into regular grids, with each grid representing the coverage of a potential edge node; collect the flight trajectory data of all aircraft in a fixed time period, and project it onto the grid, and fit the flight trajectory data through the spline interpolation method to obtain the flight trajectory; for each flight trajectory, divide it into a series of trajectory points in chronological order; traverse each trajectory point and determine the grid to which it belongs based on its latitude and longitude coordinates; for each grid, count the number of all trajectory points that appear in the grid as the number of aircraft in the corresponding grid; for each grid, calculate the cumulative flight time of all trajectory points in the corresponding grid as the total flight time;

[0010] The weight value of the corresponding grid is calculated based on the number of aircraft and the total flight time; the calculation formula of the weight value is:

[0011] ;

[0012] in, , and is a grid weight coefficient between 0 and 1; For the grid in time The weight value when ; For the grid in time The standardized number of aircraft at the time; For the grid in time Normalized total flight time For the grid in time The geographical location weight coefficient at time For the grid in time The time weight coefficient of ;

[0013] ;

[0014] in, , and is the weight of the corresponding item; For the grid in time The Euclidean distance or great circle distance from the geographic coordinates at the time to the city center, is the reference distance; For the grid in time The average altitude of the location at the time; is the reference altitude; For the grid in time The ratio of the population to the area of ​​the region at that time is called population density. is the maximum population density; is the standard deviation;

[0015] Based on the weight values ​​of the grids, the grids are classified into several grid groups; for each grid group, the weight values ​​of all the grids therein are obtained, and the grid with the largest weight value is used as the representative grid of the corresponding grid group; the geographical location of the representative grid is determined as the preliminary deployment location of the edge node, and all the grids in the corresponding grid group are within the coverage area of ​​the edge node; repeat until all grid groups have determined the preliminary deployment location of the corresponding edge nodes;

[0016] The initial deployment positions of the edge nodes are comprehensively adjusted to obtain the final deployment positions of several edge nodes; a communication link is established between the edge nodes and the aircraft, and a central server is deployed as the control center of the entire distributed edge computing network. A wired network or a wireless network is used to establish a communication link between the central server and each edge node, thus completing the construction of the distributed edge computing network.

[0017] Furthermore, the method of classifying the grids into a plurality of grid groups includes:

[0018] The longitude and latitude coordinates of each grid and the corresponding weight value are combined as the input data point; the desired number of groups k, the average number of dimensions l and the neighborhood radius r are defined; m input data points are randomly selected from all the input data points as the initial seed points; the n1-l data points closest to the corresponding seed point are selected from the circular area constructed with the seed point as the center and the neighborhood radius r as the radius, where n1 is the dimension of the input data point; PCA principal component analysis is performed on these nearest n1-l data points to find l principal component dimensions; the distance from each input data point to the seed point is calculated on these l principal component dimensions, and the input data points whose distance is less than the preset distance threshold are divided into a grid group.

[0019] Furthermore, the method of comprehensively adjusting the initial deployment position of the edge node includes:

[0020] The coverage areas of all edge nodes are represented as polygons, and a boundary table and a seed set are created; the polygons are placed in a pre-constructed two-dimensional coordinate system, the edges of all polygons are sorted from small to large according to their vertical coordinates in the two-dimensional coordinate system, and a scanning queue is constructed;

[0021] Take out the edges with the smallest vertical coordinate from the scan queue in turn and add them to the boundary table. In the two-dimensional coordinate system, move a horizontal scan line from top to bottom or from bottom to top. For each edge in the boundary table, calculate the horizontal coordinate of the intersection with the horizontal scan line, sort them from small to large according to the horizontal coordinate, and add the line segments between the intersections to the seed set. If two adjacent intersections belong to the same polygon, the line segment between them will not be added to the seed set; remove the processed edges from the boundary table and add new edges; repeat until the boundary table is empty; reallocate the edge nodes based on the seed set to obtain the final deployment position of the edge nodes.

[0022] Furthermore, the method of reallocating edge nodes includes:

[0023] Discretize the data collection area and divide it into a large number of small rectangular cells, each of which is equivalent to a pixel.

[0024] For each line segment in the seed set, its two endpoints are used as seed points. Starting from the seed point, the seed filling algorithm is used to fill the surrounding pixels. During the filling process, if the boundary of other polygons is encountered, the filling is stopped. For each filled area, the edge node to which it belongs is determined.

[0025] For each filled area, if it is covered by several edge nodes, the corresponding area is an overlapping area; for each overlapping area, calculate the optimization index value when assigning the overlapping area to different edge nodes; the calculation formula of the optimization index value is:

[0026] ;

[0027] in, is the balance coefficient, To assign the overlapping area to The optimization index value when there are edge nodes, is the fixed energy consumption of the corresponding edge node, is the variable energy consumption of the corresponding edge node, is the area of ​​the original coverage area of ​​the corresponding edge node; is the area of ​​the corresponding overlapping region, is the sum of the delays in the original coverage area of ​​the corresponding edge node, is the sum of the delays in the overlapping areas; is the maximum coverage area of ​​the edge node;

[0028] The edge node with the smallest optimization index value is selected as the representative node of the corresponding overlapping area, and the edge nodes except the representative node in the corresponding overlapping area are removed from the overlapping area; the edge nodes except the representative node are recorded as non-representative nodes; all pixels in the overlapping area are traversed, and for each pixel, if it is covered by the coverage area of ​​the non-representative node, the pixel is removed from the coverage area of ​​the non-representative node; for each edge node, the polygon of its coverage area is recalculated according to the removed pixels; and its centroid is calculated as the final deployment position of the corresponding edge node.

[0029] Further, the low-altitude economic flight data includes aircraft information, flight trajectory data, mission information, environmental data and communication data;

[0030] Aircraft information includes aircraft ID, aircraft type, payload, maximum flight time and battery power; flight trajectory data includes timestamp, latitude and longitude coordinates, altitude and speed; mission information includes mission type, take-off time, landing time, route planning and cargo information; environmental data includes temperature, humidity, wind speed, wind direction and visibility; communication data includes signal strength, latency and packet loss rate.

[0031] Furthermore, the data blocks are divided into:

[0032] The low-altitude economic flight data is constructed into a vector representation to obtain a flight data vector; the similarity between any two edge nodes is calculated and represented by the cosine similarity between the corresponding flight data vectors; a similarity threshold is defined, all edge nodes are traversed, and two edge nodes with a similarity greater than the similarity threshold are connected with a virtual undirected edge, and the corresponding similarity is used as the weight of the undirected edge; then a complete weighted undirected graph is constructed as a data network graph; data blocks are divided based on the data network graph to obtain n data blocks.

[0033] Furthermore, the step of dividing the data blocks based on the data network diagram includes:

[0034] Step 1: The edge nodes in the data network graph are recorded as graph nodes; initially, each graph node is initialized as an independent community;

[0035] Step 2: Traverse each graph node , for graph nodes Each graph node connected ; Calculate the graph nodes Move to a graph node from the current community The network modularity increment of the community ;

[0036] ;

[0037] in, For graph nodes Move to graph node The sum of the weights of the undirected edges within the community; For graph nodes With graph nodes The sum of the weights of the undirected edges between other nodes in the community; is the total number of undirected edges in the data network graph, is the sum of the weights of the undirected edges of the entire data network graph; For graph nodes The sum of the weights of the undirected edges between all other graph nodes in the entire data network graph; For graph nodes Move to graph node After the community, the graph node With graph nodes The average time distance between other graph nodes in the community; is the maximum time distance between any two graph nodes in the data network graph; is a positive real number;

[0038] like If it is greater than 0, the graph node Move to a graph node from the current community Community; repeat until Less than or equal to 0; get several middle communities;

[0039] Step 3: Treat each mid-segment community as a new node; the weight of the undirected edges between the new nodes is the sum of the weights of the undirected edges between the nodes in the original community internal graph; and obtain a new community network;

[0040] Step 4: Repeat steps 2 and 3 on the new community network until the network modularity increment no longer increases; finally, assign the graph nodes in the data network graph to the corresponding communities, and each community is a data block.

[0041] Furthermore, the method of performing integral encoding includes:

[0042] The flight trajectory data in each data block are fitted into a flight trajectory curve, and the flight trajectory curve is divided into several trajectory segments, each trajectory segment corresponds to a continuous flight process, and the attached attributes of each trajectory segment are defined as low-altitude economic flight data in addition to the flight trajectory data; each point on the trajectory segment corresponds to a set of low-altitude economic flight data in addition to the flight trajectory data; the trajectory segment is projected onto a two-dimensional plane; the two-dimensional plane is divided into regular grids; for each grid, the integral value of the trajectory segment in the corresponding grid is calculated; for each grid, its integral weight is defined ;in, is the distance adjustment parameter; , and is the periodic parameter; is the current time period; is the distance between the center of the corresponding grid and the two-dimensional plane;

[0043] The integral values ​​of all grids are multiplied by the corresponding integral weights and concatenated in a certain order to obtain the encoding vector of the corresponding trajectory segment; for each grid, the encoding vectors of all trajectory segments within it are concatenated to obtain an encoding data block;

[0044] Flight data points ;in, is the total length of all track segments in the coded data block, is the average speed of all trajectory points in the encoded data block, is the total time of all track segments in the encoding data block, , and is the weight coefficient of the corresponding item.

[0045] A low-altitude economic flight data management system based on edge computing, which is used to implement a low-altitude economic flight data management method based on edge computing, including: a network establishment module, used to establish a distributed edge computing network, the distributed edge computing network is composed of a plurality of edge nodes and a central server, each edge node is responsible for collecting low-altitude economic flight data from an aircraft;

[0046] A data partitioning module is used to divide the low-altitude economic flight data in all edge nodes into n data blocks;

[0047] The coding management and scheduling module is used to perform integral coding on each data block in the edge node to obtain a coded data block and calculate the flight data integral value corresponding to the coded data block; the edge node uploads the coded data block and the flight data integral value to the central server, and the central server manages and schedules the coded data block of the edge node according to the flight data integral value; each module is connected by wired and / or wireless means.

[0048] The technical effects and advantages of the low-altitude economic flight data management method and system based on edge computing of the present invention are as follows:

[0049] The present invention constructs a distributed edge computing network to disperse data processing tasks to edge nodes, which greatly improves data processing capabilities, effectively alleviates the pressure on central servers, and realizes efficient and real-time data collection and processing; at the same time, it reduces the demand for network bandwidth, reduces the delay in data transmission, and ensures the real-time requirements of low-altitude economic flight; secondly, it highly compresses massive amounts of low-altitude economic flight data, greatly reduces the bandwidth and space required for data transmission and storage, and reduces system operating costs; at the same time, the encoding process retains the main features of the data, laying the foundation for subsequent data analysis and decision support; in addition, it dynamically adjusts data processing priorities according to the importance of flight data, reasonably allocates computing resources, ensures that key data can be processed first, improves the flexibility and adaptability of the system, and meets the urgent real-time requirements of low-altitude economic flight; it provides strong support for the planning, scheduling and management of low-altitude economic flight.

[0050] The low-altitude economic flight data management method proposed in the present invention establishes a secure communication link between the edge node and the central server. The encryption algorithm of the secure communication link can strengthen data security and protection, establish data security management, and prevent data leakage and abuse. It provides network security, functional safety, and data security functions that are highly reliable, highly trustworthy, and highly available. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic diagram of a low-altitude economic flight data management method based on edge computing of the present invention;

[0052] Figure 2 This is a schematic diagram of a low-altitude economic flight data management system based on edge computing of the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] Example 1

[0055] See also Figure 1 As shown, this embodiment provides a low-altitude economic flight data management method based on edge computing, including:

[0056] S1. Establish a distributed edge computing network, which consists of several edge nodes and a central server. Each edge node is responsible for collecting low-altitude economic flight data from aircraft.

[0057] S2. Divide the low-altitude economic flight data in all edge nodes into n data blocks; n is an integer greater than 1;

[0058] S3. Perform integral encoding on each data block in the edge node to obtain an encoded data block, and calculate the flight data integral value corresponding to the encoded data block; the edge node uploads the encoded data block and the flight data integral value to the central server, and the central server manages and schedules the encoded data block of the edge node according to the flight data integral value.

[0059] Ways to build a distributed edge computing network include:

[0060] Determine the data collection area, and determine the number of edge nodes to be deployed based on the size of the data collection area and the number of aircraft; specifically, determine the geographic boundaries of the data collection area, using latitude and longitude coordinates or other geographic information system (GIS) data to represent; divide the entire data collection area into regular grids, each grid representing the coverage of a potential edge node.

[0061] The flight trajectory data of all aircraft within a fixed time period (for example, a week or a month) are collected and projected onto a grid; specifically, the flight trajectory data includes timestamp, longitude and latitude coordinates, altitude and speed; the flight trajectory data is fitted by a spline interpolation method (such as cubic spline, B-spline) to obtain the flight trajectory; for each flight trajectory, it is divided into a series of trajectory points in chronological order; each trajectory point is traversed and the grid to which it belongs is determined according to its longitude and latitude coordinates; for each grid, the number of all trajectory points appearing in the grid is counted as the number of aircraft in the corresponding grid; for each grid, the cumulative flight time of all trajectory points in the corresponding grid is calculated as the total flight time.

[0062] The weight value of the corresponding grid is calculated based on the number of aircraft and the total flight time; the calculation formula of the weight value is:

[0063] ;

[0064] in, , and is a grid weight coefficient between 0 and 1, which is adjusted and optimized according to the actual situation; For the grid in time The weight value when ; For the grid in time The standardized number of aircraft at the time; For the grid in time Normalized total flight time For the grid in time The geographical location weight coefficient at that time reflects the influence of the geographical location and environment of the grid on data collection. For the grid in time The time weight coefficient reflects the importance of data in different time periods.

[0065] ;

[0066] in, , and is the weight of the corresponding item, and the sum of the three is 1; For the grid in time The Euclidean distance or great circle distance from the geographic coordinates at the time to the city center, The reference distance is a preset constant used to control the influence of distance on the weight. For example, half of the city radius can be taken as the reference distance. For the grid in time The average altitude of the location at the time; The reference altitude is a preset constant used to represent the ideal altitude for data collection; for example, for some scenes, a lower altitude may be more suitable for data collection; For the grid in time The ratio of the population to the area of ​​the region at that time is called population density. The maximum population density is used to normalize the population density to between 0 and 1. You can take the maximum population density value of the area, or take an empirical value as the upper limit; is the standard deviation, which describes the distribution of altitude deviation from the reference altitude. It is calculated based on the actual data obtained. For the altitude distribution characteristics of the region, a reasonable standard deviation value can also be set directly based on experience.

[0067] For the time weight coefficient, according to the actual situation, define a time importance function g(t), normalize it to get the time weight coefficient; define it according to specific needs, for example, the weights of weekdays and holidays are different: if the time is a weekday, then g(t)=w1; if the time is a holiday, g(t)=w2; different weights for different time periods: g(t)=w(h) (h is the number of hours in a day, w(h) is the weight of the corresponding hour); consider seasonal changes: g(t)=w(m) (m is the month, w(m) is the weight of the corresponding month).

[0068] Based on the weight values ​​of the grids, the grids are classified into several grid groups; each grid group represents the coverage of an edge node; specifically, the longitude and latitude coordinates of each grid and the corresponding weight value are combined as input data points, and the combination method can be vector-level splicing; define the expected number of groups k, the average number of dimensions l and the neighborhood radius r; randomly select m input data points from all input data points as initial seed points; select n1-l data points closest to the corresponding seed point in a circular area constructed with the seed point as the center and the neighborhood radius r as the radius, where n1 is the dimension of the input data point; perform PCA principal component analysis on these nearest n1-l data points to find the l principal component dimensions with the largest variance after projection; calculate the distance from each input data point to the seed point on these l principal component dimensions, and divide the input data points with a distance less than a preset distance threshold into a grid group.

[0069] For each grid group, the weight values ​​of all the grids therein are obtained, and the grid with the largest weight value is used as the representative grid of the corresponding grid group; the geographical location of the representative grid is determined as the preliminary deployment position of the edge node, and all grids in the corresponding grid group are within the coverage area of ​​the edge node; repeat until all grid groups have determined the preliminary deployment positions of the corresponding edge nodes; the preliminary deployment positions of the edge nodes are comprehensively adjusted to obtain the final deployment positions of several edge nodes.

[0070] Specifically, the coverage areas of all edge nodes are represented as polygons, and a boundary table and a seed set are created; the polygons are placed in a pre-constructed two-dimensional coordinate system, the edges of all polygons are sorted from small to large according to their vertical coordinates in the two-dimensional coordinate system, and a scanning queue is constructed;

[0071] Take out the edges with the smallest ordinate from the scan queue in turn and add them to the boundary table. In the two-dimensional coordinate system, move a horizontal scan line from top to bottom (or from bottom to top). For each edge in the boundary table, calculate the horizontal coordinate of the intersection with the horizontal scan line, sort them from small to large according to the horizontal coordinate, and add the line segments between the intersections to the seed set. If two adjacent intersections belong to the same polygon, the line segment between them will not be added to the seed set; remove the processed edges from the boundary table and add new edges (if any); repeat until the boundary table is empty; reallocate the edge nodes based on the seed set to obtain the final deployment position of the edge nodes.

[0072] The data collection area is discretized and divided into a large number of small rectangular cells, each of which is equivalent to a pixel point; for each line segment in the seed set, its two endpoints are used as seed points, starting from the seed point, the seed filling algorithm is used to fill the surrounding pixels; during the filling process, if the boundary of other polygons is encountered, the filling is stopped; for each filled area, the edge node to which it belongs is determined.

[0073] For each filled area, if it is covered by several edge nodes, the corresponding area is an overlapping area; for each overlapping area, calculate the optimization index value when assigning the overlapping area to different edge nodes; the calculation formula of the optimization index value is:

[0074] ;

[0075] in, is the balance coefficient, which is used to balance the weights of factors affecting node energy consumption and delay. To assign the overlapping area to The optimization index value when there are edge nodes, is the fixed energy consumption of the corresponding edge node, is the variable energy consumption of the corresponding edge node, is the area of ​​the original coverage area of ​​the corresponding edge node; is the area of ​​the corresponding overlapping region, is the sum of the delays in the original coverage area of ​​the corresponding edge node, is the sum of the delays in the overlapping areas; is the maximum coverage area of ​​the edge node.

[0076] It should be noted that fixed energy consumption refers to the basic energy consumption of edge nodes during operation, including hardware energy consumption (CPU, memory, network, etc.) and software energy consumption (operating system, middleware, etc.); it is obtained through testing and monitoring, such as measuring the energy consumption of edge nodes under no-load conditions; variable energy consumption refers to the energy consumption related to the workload of edge nodes, which usually increases with the increase of coverage area or amount of data processed; the maximum coverage area of ​​edge nodes refers to the maximum coverage capacity of edge nodes, and the edge nodes may not be able to operate efficiently beyond this area; it is estimated based on the hardware configuration and software capabilities of edge nodes, and can also be obtained through stress testing; the sum of delays refers to the sum of delays from all pixels to edge nodes within the original coverage area of ​​the edge nodes.

[0077] The edge node with the smallest optimization index value is selected as the representative node of the corresponding overlapping area, and the edge nodes except the representative node in the corresponding overlapping area are removed from the overlapping area; the edge nodes except the representative node are recorded as non-representative nodes; all pixels in the overlapping area are traversed, and for each pixel, if it is covered by the coverage area of ​​the non-representative node, the pixel is removed from the coverage area of ​​the non-representative node; for each edge node, the polygon of its coverage area is recalculated according to the removed pixels; and its centroid is calculated as the final deployment position of the corresponding edge node.

[0078] To establish a communication link between the edge node and the aircraft, wireless communication technologies such as 5G and LoRa can be used to ensure that the edge node can receive the low-altitude economic flight data sent by the aircraft in real time; deploy a central server as the control center of the entire distributed edge computing network; the central server has sufficient computing power and storage space to process data from all edge nodes.

[0079] Use wired or wireless networks to establish communication links between the central server and each edge node to ensure efficient data transmission, thus completing the construction of a distributed edge computing network. Build a distributed edge computing network covering the entire target area to achieve real-time collection, processing and management of low-altitude economic flight data.

[0080] The low-altitude economic flight data management method proposed in the present invention establishes a secure communication link between the edge node and the central server. The encryption algorithm of the secure communication link can strengthen data security and protection, establish data security management, and prevent data leakage and abuse. It provides network security, functional safety, and data security functions that are highly reliable, highly trustworthy, and highly available.

[0081] Low-altitude economic flight data includes aircraft information, flight trajectory data, mission information, environmental data and communication data; aircraft information includes aircraft ID, aircraft type, payload, maximum flight time and battery power; flight trajectory data includes timestamp, latitude and longitude coordinates, altitude and speed; mission information includes mission type (cargo, monitoring, etc.), take-off time, landing time, route planning and cargo information; environmental data includes temperature, humidity, wind speed, wind direction and visibility; communication data includes signal strength, latency and packet loss rate.

[0082] The data blocks are divided into the following ways:

[0083] The low-altitude economic flight data is constructed into a vector representation, that is, the aircraft information, flight trajectory data, mission information, environmental data and communication data are spliced ​​at the vector level to obtain the flight data vector; the similarity between any two edge nodes is calculated, and it is represented by the cosine similarity between the corresponding flight data vectors; a similarity threshold is defined, all edge nodes are traversed, and two edge nodes with a similarity greater than the similarity threshold are connected with a virtual undirected edge, and the corresponding similarity is used as the weight of the undirected edge; then a complete weighted undirected graph is constructed as a data network graph; data blocks are divided based on the data network graph to obtain n data blocks.

[0084] Specifically, the steps include:

[0085] Step 1: The edge nodes in the data network graph are recorded as graph nodes; initially, each graph node is initialized as an independent community;

[0086] Step 2: Traverse each graph node , for graph nodes Each graph node connected ; Calculate the graph nodes Move to a graph node from the current community The network modularity increment of the community ;

[0087] ;

[0088] in, For graph nodes Move to graph node The sum of the weights of the undirected edges within the community; For graph nodes With graph nodes The sum of the weights of the undirected edges between other nodes in the community; is the total number of undirected edges in the data network graph, is the sum of the weights of the undirected edges of the entire data network graph; For graph nodes The sum of the weights of the undirected edges between all other graph nodes in the entire data network graph; For graph nodes Move to graph node After the community, the graph node With graph nodes The average time distance between other graph nodes in the community (the difference between the timestamps of two graph nodes); is the maximum time distance between any two graph nodes in the data network graph; is a positive real number used to adjust relative importance;

[0089] like If it is greater than 0, the graph node Move to a graph node from the current community Community; repeat until Less than or equal to 0; get several middle communities;

[0090] Step 3: Treat each mid-segment community as a new node; the weight of the undirected edges between the new nodes is the sum of the weights of the undirected edges between the nodes in the original community internal graph; and obtain a new community network.

[0091] Step 4: Repeat steps 2 and 3 on the new community network until the network modularity increment no longer increases; finally, assign the graph nodes in the data network graph to the corresponding communities, and each community is a data block.

[0092] Ways to encode points include:

[0093] The flight trajectory data in each data block are fitted into a flight trajectory curve, and the flight trajectory curve is divided into several trajectory segments, each trajectory segment corresponds to a continuous flight process, and the attached attributes of each trajectory segment are defined as low-altitude economic flight data in addition to the flight trajectory data; each point on the trajectory segment corresponds to a set of low-altitude economic flight data in addition to the flight trajectory data; the trajectory segment is projected onto a two-dimensional plane; the two-dimensional plane is divided into regular grids; for each grid, the integral value of the trajectory segment in the corresponding grid is calculated; specifically, the two adjacent sides of the grid are used as the two axes of the two-dimensional coordinate system, and the area formed by the corresponding trajectory segment and the two axes is the integral value.

[0094] For each grid, define its integral weight:

[0095] ;

[0096] in, It is a distance adjustment parameter, usually ranging from 0.01 to 0.1, which is used to control the influence of distance; , and is a periodic parameter, determined according to actual conditions; for example, , , ; is the current time period, for example, the hour of the day; is the distance between the center of the corresponding grid and the two-dimensional plane.

[0097] The integral values ​​of all grids are multiplied by the corresponding integral weights and concatenated in a certain order to obtain the encoding vector of the corresponding trajectory segment; for each grid, the encoding vectors of all the trajectory segments inside it are concatenated to obtain the encoding data block of the grid (with attached attributes).

[0098] Calculate the flight data integral value corresponding to the coded data block:

[0099] ;

[0100] in, is the total length of all track segments in the coded data block, is the average speed of all trajectory points in the encoded data block, is the total time of all track segments in the encoding data block, , and is the weight coefficient of the corresponding item, which will be adjusted according to actual conditions.

[0101] These integral value codes can effectively compress data while retaining the main features of the data, providing a basis for subsequent data management and optimization.

[0102] Each edge node packages the coded data blocks and corresponding flight data integral values ​​within it into a data packet; the data packet contains metadata, such as edge node ID, timestamp, data block size, etc.; the data packet is uploaded to the central server through a secure communication link (such as VPN, encrypted channel) established with the central server; the central server is provided with a data receiving end, which is used to receive data packets from each edge node, parse the received data packets, extract the coded data blocks, flight data integral values ​​and metadata; the extracted data is stored in the central database and indexed by the edge node ID and timestamp.

[0103] The central server maintains a priority queue and sorts all edge nodes from large to small according to the flight data integral value; regularly takes out several edge nodes with the highest flight data integral values ​​from the priority queue; decodes the encoded data blocks of these edge nodes to reconstruct the original low-altitude economic flight data; analyzes and mines the low-altitude economic flight data to generate decision support information for the planning and management of low-altitude economic flight; the central server dynamically adjusts the priority queue according to the real-time status of the analysis to increase the processing priority of important data; at the same time, according to the load status of the edge nodes, dynamically allocates computing resources to avoid overloading of certain edge nodes.

[0104] The central server can efficiently manage and schedule the coded data blocks distributed in various edge nodes, fully utilizing the advantages of edge computing to achieve real-time processing and decision support of low-altitude economic flight data.

[0105] In this embodiment, by constructing a distributed edge computing network, data processing tasks are dispersed to edge nodes, which greatly improves data processing capabilities, effectively alleviates the pressure on the central server, and realizes efficient and real-time data collection and processing; at the same time, it reduces the demand for network bandwidth, reduces the delay in data transmission, and ensures the real-time requirements of low-altitude economic flight; secondly, the massive low-altitude economic flight data is highly compressed, which greatly reduces the bandwidth and space required for data transmission and storage, and reduces the system operating costs; at the same time, the encoding process retains the main features of the data, laying the foundation for subsequent data analysis and decision support; in addition, the data processing priority is dynamically adjusted according to the importance of the flight data, and computing resources are reasonably allocated to ensure that key data can be processed first, which improves the flexibility and adaptability of the system and meets the urgent real-time requirements of low-altitude economic flight; it provides strong support for the planning, scheduling and management of low-altitude economic flight.

[0106] Example 2

[0107] See also Figure 2 As shown, the part not described in detail in this embodiment is described in Example 1, which provides a low-altitude economic flight data management system based on edge computing, including:

[0108] A network establishment module is used to establish a distributed edge computing network. The distributed edge computing network consists of several edge nodes and a central server. Each edge node is responsible for collecting low-altitude economic flight data from aircraft.

[0109] The data partitioning module is used to divide the low-altitude economic flight data in all edge nodes into n data blocks; n is an integer greater than 1;

[0110] The coding management and scheduling module is used to perform integral coding on each data block in the edge node to obtain a coded data block and calculate the flight data integral value corresponding to the coded data block; the edge node uploads the coded data block and the flight data integral value to the central server, and the central server manages and schedules the coded data block of the edge node according to the flight data integral value; each module is connected by wire and / or wireless means to realize data transmission between modules.

[0111] Example 3

[0112] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the operation mode of the low-altitude economic flight data management method based on edge computing provided above is implemented.

[0113] Since the electronic device introduced in this embodiment is an electronic device used to implement a low-altitude economic flight data management method based on edge computing in the embodiment of the present application, based on the low-altitude economic flight data management method based on edge computing introduced in the embodiment of the present application, the technical personnel of this field can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application is not introduced in detail here. As long as the technical personnel of this field implement the electronic device used in the low-altitude economic flight data management method based on edge computing in the embodiment of the present application, it belongs to the scope of protection of this application.

[0114] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0115] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical users in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A low-altitude economic flight data management method based on edge computing, characterized in that: include: S1. Establish a distributed edge computing network, which consists of several edge nodes and a central server. Each edge node is responsible for collecting low-altitude economic flight data from aircraft. S2, divide the low-altitude economic flight data in all edge nodes into n data blocks; S3, performing integral encoding on each data block in the edge node to obtain an encoded data block, and calculating the flight data integral value corresponding to the encoded data block; the edge node uploads the encoded data block and the flight data integral value to the central server, and the central server manages and schedules the encoded data block of the edge node according to the flight data integral value; The data block division method includes: The low-altitude economic flight data is constructed into a vector representation to obtain a flight data vector; the similarity between any two edge nodes is calculated and represented by the cosine similarity between the corresponding flight data vectors; a similarity threshold is defined, all edge nodes are traversed, two edge nodes whose similarity is greater than the similarity threshold are connected by a virtual undirected edge, and the corresponding similarity is used as the weight of the undirected edge; then a complete weighted undirected graph is constructed as a data network graph; data blocks are divided based on the data network graph to obtain n data blocks; The method of performing integral encoding includes: The flight trajectory data in each data block are fitted into a flight trajectory curve, and the flight trajectory curve is divided into several trajectory segments, each trajectory segment corresponds to a continuous flight process, and the attached attributes of each trajectory segment are defined as low-altitude economic flight data in addition to the flight trajectory data; each point on the trajectory segment corresponds to a set of low-altitude economic flight data in addition to the flight trajectory data; the trajectory segment is projected onto a two-dimensional plane; the two-dimensional plane is divided into regular grids; for each grid, the integral value of the trajectory segment in the corresponding grid is calculated; for each grid, its integral weight is defined Among them, θ is the distance adjustment parameter; τ, ω and is the periodic parameter; T is the current time period; dw is the distance between the corresponding grid and the center of the two-dimensional plane; The integral values ​​of all grids are multiplied by the corresponding integral weights and concatenated in a certain order to obtain the encoding vector of the corresponding trajectory segment; for each grid, the encoding vectors of all trajectory segments within it are concatenated to obtain an encoding data block; The flight data integral value FG=v1×L1+v2×V1+v3×T1; wherein L1 is the total length of all trajectory segments in the encoded data block, V1 is the average speed of all trajectory points in the encoded data block, T1 is the total time of all trajectory segments in the encoded data block, and v1, v2 and v3 are the weight coefficients of the corresponding items.

2. The low-altitude economic flight data management method based on edge computing according to claim 1 is characterized in that: The method of establishing a distributed edge computing network includes: Determine the data collection area, and determine the number of edge nodes to be deployed based on the size of the data collection area and the number of aircraft; determine the geographic boundaries of the data collection area, and divide the entire data collection area into regular grids, with each grid representing the coverage of a potential edge node; collect the flight trajectory data of all aircraft in a fixed time period, and project it onto the grid, and fit the flight trajectory data through the spline interpolation method to obtain the flight trajectory; for each flight trajectory, divide it into a series of trajectory points in chronological order; traverse each trajectory point and determine the grid to which it belongs based on its latitude and longitude coordinates; for each grid, count the number of all trajectory points that appear in the grid as the number of aircraft in the corresponding grid; for each grid, calculate the cumulative flight time of all trajectory points in the corresponding grid as the total flight time; The weight value of the corresponding grid is calculated based on the number of aircraft and the total flight time; the calculation formula of the weight value is: W_i(t)=α×N_i(t)+β×T_i(t)+γ×G_i(t)×H_i(t); where α, β and γ are grid weight coefficients between 0 and 1; W_i(t) is the weight value of the ith grid at time t; N_i(t) is the standardized number of aircraft in the ith grid at time t; T_i(t) is the standardized total flight time of the ith grid at time t; G_i(t) is the geographic location weight coefficient of the ith grid at time t, and H_i(t) is the time weight coefficient of the ith grid at time t; Among them, α1, α2 and α3 are the weights of the corresponding items; d(i,t) is the Euclidean distance or great circle distance from the geographical coordinates of the i-th grid at time t to the city center, d0 is the reference distance; h(i,t) is the average altitude of the location of the i-th grid at time t; h0 is the reference altitude; p(i,t) is the ratio of the population to the area of ​​the i-th grid at time t, that is, the population density; p_max is the maximum population density; σ is the standard deviation; Based on the weight values ​​of the grids, the grids are classified into several grid groups; for each grid group, the weight values ​​of all the grids therein are obtained, and the grid with the largest weight value is used as the representative grid of the corresponding grid group; the geographical location of the representative grid is determined as the preliminary deployment location of the edge node, and all the grids in the corresponding grid group are within the coverage area of ​​the edge node; repeat until all grid groups have determined the preliminary deployment location of the corresponding edge nodes; The initial deployment positions of the edge nodes are comprehensively adjusted to obtain the final deployment positions of several edge nodes; a communication link is established between the edge nodes and the aircraft, and a central server is deployed as the control center of the entire distributed edge computing network. A wired network or a wireless network is used to establish a communication link between the central server and each edge node, thus completing the construction of the distributed edge computing network.

3. The low-altitude economic flight data management method based on edge computing according to claim 2 is characterized in that: The method of classifying the grids to obtain a plurality of grid groups includes: The longitude and latitude coordinates of each grid and the corresponding weight value are combined as the input data point; the desired number of groups k, the average number of dimensions l and the neighborhood radius r are defined; m input data points are randomly selected from all the input data points as the initial seed points; the n1-l data points closest to the corresponding seed point are selected from the circular area constructed with the seed point as the center and the neighborhood radius r as the radius, where n1 is the dimension of the input data point; PCA principal component analysis is performed on these nearest n1-l data points to find l principal component dimensions; the distance from each input data point to the seed point is calculated on these l principal component dimensions, and the input data points whose distance is less than the preset distance threshold are divided into a grid group.

4. The low-altitude economic flight data management method based on edge computing according to claim 3 is characterized in that: The method of comprehensively adjusting the initial deployment position of the edge node includes: The coverage areas of all edge nodes are represented as polygons, and a boundary table and a seed set are created; the polygons are placed in a pre-constructed two-dimensional coordinate system, the edges of all polygons are sorted from small to large according to their vertical coordinates in the two-dimensional coordinate system, and a scanning queue is constructed; Take out the edges with the smallest vertical coordinate from the scan queue in turn and add them to the boundary table. In the two-dimensional coordinate system, move a horizontal scan line from top to bottom or from bottom to top. For each edge in the boundary table, calculate the horizontal coordinate of the intersection with the horizontal scan line, sort them from small to large according to the horizontal coordinate, and add the line segments between the intersections to the seed set. If two adjacent intersections belong to the same polygon, the line segment between them will not be added to the seed set; remove the processed edges from the boundary table and add new edges; repeat until the boundary table is empty; reallocate the edge nodes based on the seed set to obtain the final deployment position of the edge nodes.

5. The low-altitude economic flight data management method based on edge computing according to claim 4 is characterized in that: The method of reallocating edge nodes includes: Discretize the data collection area and divide it into a large number of small rectangular cells, each of which is equivalent to a pixel. For each line segment in the seed set, its two endpoints are used as seed points. Starting from the seed point, the seed filling algorithm is used to fill the surrounding pixels. During the filling process, if the boundary of other polygons is encountered, the filling is stopped. For each filled area, the edge node to which it belongs is determined. For each filled area, if it is covered by several edge nodes, the corresponding area is an overlapping area. For each overlapping area, the optimization index value when the overlapping area is assigned to different edge nodes is calculated. The calculation formula of the optimization index value is: Wherein, w1 is the balance coefficient, F_j is the optimization index value when the overlapping area is assigned to the jth edge node, Pf is the fixed energy consumption of the corresponding edge node, Pv is the variable energy consumption of the corresponding edge node, Ae is the area of ​​the original coverage area of ​​the corresponding edge node; Ao is the area of ​​the corresponding overlapping area, De is the sum of the delays of the original coverage area of ​​the corresponding edge node, Do is the sum of the delays of the overlapping area; A_max is the maximum coverage area of ​​the edge node; The edge node with the smallest optimization index value is selected as the representative node of the corresponding overlapping area, and the edge nodes except the representative node in the corresponding overlapping area are removed from the overlapping area; the edge nodes except the representative node are recorded as non-representative nodes; all pixels in the overlapping area are traversed, and for each pixel, if it is covered by the coverage area of ​​the non-representative node, the pixel is removed from the coverage area of ​​the non-representative node; for each edge node, the polygon of its coverage area is recalculated according to the removed pixels; and its centroid is calculated as the final deployment position of the corresponding edge node.

6. The low-altitude economic flight data management method based on edge computing according to claim 5 is characterized in that: The low-altitude economic flight data includes aircraft information, flight trajectory data, mission information, environmental data and communication data; Aircraft information includes aircraft ID, aircraft type, payload, maximum flight time and battery power; flight trajectory data includes timestamp, latitude and longitude coordinates, altitude and speed; mission information includes mission type, take-off time, landing time, route planning and cargo information; environmental data includes temperature, humidity, wind speed, wind direction and visibility; communication data includes signal strength, latency and packet loss rate.

7. The low-altitude economic flight data management method based on edge computing according to claim 1 is characterized in that: The step of dividing the data blocks based on the data network diagram comprises: Step 1: The edge nodes in the data network graph are recorded as graph nodes; initially, each graph node is initialized as an independent community; Step 2: traverse each graph node I, and for each graph node J connected to the graph node I, calculate the network modularity increment ΔQ of moving the graph node I from the current community to the community where the graph node J is located; Among them, ER_I is the sum of the weights of the undirected edges in the community where graph node I moves to graph node J; KR_I is the sum of the weights of the undirected edges between graph node I and other nodes in the community where graph node J is located; m' is the total number of undirected edges in the data network graph, RE is the sum of the weights of the undirected edges in the entire data network graph; K_I is the sum of the weights of the undirected edges between graph node I and all other graph nodes in the entire data network graph; dn is the average time distance between graph node I and other graph nodes in the community where graph node J is located after graph node I moves to the community where graph node J is located; dm is the maximum time distance between any two graph nodes in the data network graph; ε is a positive real number; If ΔQ is greater than 0, move graph node I from the current community to the community where graph node J is located; repeat until ΔQ is less than or equal to 0; and obtain several mid-segment communities; Step 3: Treat each mid-segment community as a new node; the weight of the undirected edges between the new nodes is the sum of the weights of the undirected edges between the nodes in the original community internal graph; and obtain a new community network; Step 4: Repeat steps 2 and 3 on the new community network until the network modularity increment no longer increases; finally, assign the graph nodes in the data network graph to the corresponding communities, and each community is a data block.

8. A low-altitude economic flight data management system based on edge computing, which is used to implement the low-altitude economic flight data management method based on edge computing according to any one of claims 1 to 7, characterized in that: include: A network establishment module is used to establish a distributed edge computing network. The distributed edge computing network consists of several edge nodes and a central server. Each edge node is responsible for collecting low-altitude economic flight data from aircraft. A data partitioning module is used to divide the low-altitude economic flight data in all edge nodes into n data blocks; The coding management and scheduling module is used to perform integral coding on each data block in the edge node to obtain a coded data block and calculate the flight data integral value corresponding to the coded data block; the edge node uploads the coded data block and the flight data integral value to the central server, and the central server manages and schedules the coded data block of the edge node according to the flight data integral value; each module is connected by wired and / or wireless means.

Images