A method for intelligently generating flight routes for flight management equipment

Abstract

This application discloses a method for intelligently generating routes for flight management equipment, belonging to the field of flight control technology, and comprising the following steps: Step 1: Generate a basic route through the flight management equipment, wherein the conditions for generating the basic route include at least the departure airport, landing airport, flight direction, civil aviation route, and navigation database; Step 2: Optimize the basic route using a deep learning network model: Based on historical route generation experience, associated learning rules, and combined with network-acquired weather information and air traffic control information corresponding to the basic route, optimize and adjust the route to generate a flight route that meets safety requirements; Step 3: Use global high-precision terrain elevation data to complete three-dimensional route obstacle avoidance. The present invention improves the efficiency and accuracy of route input, as well as the efficiency and safety of flight management.

Description

Technical Field

[0001] The present application relates to a method for intelligently generating routes for flight management equipment, and belongs to the field of flight control technology. Background Art

[0002] With the rapid development of air traffic, the efficiency and safety of flight management are receiving increasing attention. Flight routes play a key role in flight management, including route input, new route development, and creation. However, there is a pressing problem: a method for rapidly generating and analyzing routes, as well as a solution capable of evaluating them using big data.

[0003] Current route input methods are cumbersome and prone to errors. Improving the efficiency and accuracy of route input is a pressing issue. To address this challenge, a method is needed that can rapidly generate routes and evaluate them using big data. Summary of the Invention

[0004] According to one aspect of the present application, a method for intelligently generating routes for flight management equipment is provided, which improves the efficiency and accuracy of route input. Compared with the original route input method, this method relies entirely on algorithms and software for automatic generation, thereby improving the efficiency and safety of flight plan management.

[0005] A method for intelligently generating flight routes for flight management equipment, characterized by comprising the following steps:

[0006] Step 1: Generate a basic route through the flight management device, where the basic route is generated based on at least the conditions of the departure airport, landing airport, flight direction, civil aviation route and navigation database;

[0007] Step 2: Optimize the basic route using a deep learning network model. Based on historical route generation experience, associated learning rules, and weather information and air traffic control information obtained from the network, the route is optimized and adjusted to generate a flight route that meets safety requirements.

[0008] Step 3: Use global high-precision terrain elevation data to complete three-dimensional route obstacle avoidance.

[0009] Furthermore, the step 2 includes:

[0010] Dividing the basic route into flight segments to obtain a plurality of first flight segments, and requesting weather information and air traffic control information for each first flight segment;

[0011] Conduct a quantitative analysis of the risk level of meteorological information and air traffic control information for each first flight segment to determine whether it will affect flight safety;

[0012] If there is a safety impact, the flight segment will be replanned and the route will be adjusted to bypass the meteorological danger zone and avoid the waypoints in the danger zone. At the same time, new waypoints will be defined and combined to generate a new flight segment.

[0013] Generate flight paths that meet safety requirements.

[0014] Furthermore, the step three includes:

[0015] The two-dimensional plane data in the flight route obtained in step 2 is converted into three-dimensional data through an obstacle avoidance algorithm, wherein the basis for data conversion includes at least: airport altitude, cruising altitude and global high-precision terrain elevation data.

[0016] Furthermore, the method further comprises:

[0017] Generate preferred flight routes:

[0018] Generate an initial database based on departure and arrival airports,

[0019] According to the pilot route system evaluation table, weighted calculation is performed on the option scores in the pilot route system evaluation table to generate a score for the route segment;

[0020] Each time a route is generated later, the routes retrieved based on the departure and landing airports will be ranked by score, with routes with higher scores being used first.

[0021] Furthermore, the scoring options of the pilot route system evaluation form include at least:

[0022] Route comfort, route safety and route integrity.

[0023] Furthermore, the use of global high-precision terrain elevation data to complete three-dimensional route obstacle avoidance includes:

[0024] The pilot enters the cruising altitude of the flight;

[0025] According to the cruising altitude, data comparison is performed with global high-precision terrain elevation data;

[0026] Dividing the flight route into flight segments to obtain a plurality of second flight segments;

[0027] Each second flight segment is tested. If the cruising altitude of the current second flight segment minus the terrain altitude is less than 5000m, the current second flight segment is replanned.

[0028] When there is a waypoint within the set range of the second flight segment and the obstacle can be successfully avoided, the waypoint is used for detour; when there is no waypoint within the set range, the optimal route of the current second flight segment is calculated, several custom waypoints are generated, and a complete route is generated by combining the flight cruising speed and cruising altitude with the custom waypoints, thereby completing the route planning with three-dimensional data.

[0029] Furthermore, the global high-precision terrain elevation data has a tile data accuracy of 30m, which is used for stereo analysis of three-dimensional routes.

[0030] The beneficial effects of this application include:

[0031] The present application provides a method for intelligently generating routes for flight management equipment. The method has simple steps and does not require complicated operations. It only requires inputting the take-off airport, landing airport, and cruising altitude to generate a basic route. The basic route is optimized through a deep network model to generate a flight route that meets safety requirements. At the same time, the obstacle avoidance algorithm is used to realize the perception and obstacle avoidance functions of dangerous terrain. Thus, optimal flight route generation is achieved. The implementation of the present invention improves the efficiency and accuracy of route input, thereby improving the efficiency and safety of flight management. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is the overall flow chart of the method for intelligently generating routes for flight management equipment provided by the present invention;

[0033] Figure 2 The method for intelligently generating routes for flight management equipment provided by the present invention generates a basic route schematic diagram by drawing lines on a map;

[0034] Figure 3 A basic route generation diagram for the flight management equipment route intelligent generation method provided by the present invention;

[0035] Figure 4 A flowchart of the flight route generation method for intelligently generating flight routes for flight management equipment provided by the present invention;

[0036] Figure 5 A schematic diagram of obstacle avoidance in the intelligent route generation method for flight management equipment provided by the present invention;

[0037] Figure 6 An evaluation table for the intelligent route system in the flight management equipment route intelligent generation method provided by the present invention;

[0038] Figure 7 Schematic diagram of the three-dimensional route obtained by using the visual graph method in the intelligent generation method of the flight management equipment route provided by the present invention Figure 1 ;

[0039] Figure 8Schematic diagram of obtaining a three - dimensional flight route using the visibility graph method in the flight route intelligent generation method provided by the present invention Figure 2 ;

[0040] Figure 9 Effect diagram of obtaining a three - dimensional flight route using the visibility graph method in the flight route intelligent generation method provided by the present invention Detailed implementation manners

[0041] The present application will be described in detail below in conjunction with embodiments, but the present application is not limited to these embodiments

[0042] Refer to Figure 1-9 As shown in Figure 1 A flight route intelligent generation method for a flight management device, characterized by including the following steps

[0043] Step 1: Generate a basic flight route through the flight management device. As shown in Figure 2-3 , the basis conditions for generating the basic flight route at least include the departure airport, the arrival airport, the flight direction, the civil aviation route, and the navigation database

[0044] Specifically, by inputting the departure airport and the arrival airport, roughly draw the flight route and the cruising altitude on the map, and the flight management device generates a basic flight route

[0045] It should be noted that there are不外乎 two possibilities from the departure airport (S) to the arrival airport (E). Method 1 is to directly go from the departure airport to the arrival airport, and Method 2 is to pass through several nodes (P). Step 1 of the present application is applicable to Method 2. According to the flight direction, first divide several nodes (P), obtain the longitude and latitude information of several nodes (P). DIS(S,E) is the total distance between multiple nodes (P). For each node (P), obtain multiple waypoints (N) near the node (P) from the civil aviation navigation database. DIS(P,N) is the total distance between the node (P) and multiple waypoints (N). Calculate DIS(P,N) and calculate the previous DTK (Desired Track Angle) angle value. Select the waypoint (N) that meets the minimum value among multiple waypoints (N) and satisfies that the angle at the two - end DTK minus the P angle is less than 90° to replace the current node (P) in the flight route, and then traverse the subsequent nodes (P) one by one to find the waypoint (N) that meets the conditions. Finally, it should be satisfied that (DIS(S,E) / π)<DIS(S,N0)+DIS(N0,N...)+DIS(N...,E) to obtain a complete basic flight route that meets the departure airport (S), the arrival airport (E), and multiple waypoints (N) that meet the flight direction

[0046] Step 2: Optimize the basic route using a deep learning network model: Based on historical route generation experience, associated learning rules, and combined with weather information and air traffic control information corresponding to the basic route obtained by the network, the route is optimized and adjusted to generate a flight route that meets safety requirements;

[0047] Step 3: Use global high-precision terrain elevation data to complete three-dimensional route obstacle avoidance.

[0048] Specifically, the pilot of the present application only needs to input the take-off airport, landing airport and cruising altitude to generate a basic route, and optimize the basic route through a deep network model to generate a flight route that meets safety requirements. At the same time, the obstacle avoidance algorithm is used to realize the perception and obstacle avoidance functions of dangerous terrain; thereby achieving the optimal flight route generation. The implementation of the present invention improves the efficiency and accuracy of route input, thereby improving the efficiency and safety of flight management.

[0049] The second step includes:

[0050] Dividing the basic route into flight segments to obtain a plurality of first flight segments, and requesting weather information and air traffic control information for each first flight segment;

[0051] Conduct a quantitative analysis of the risk level of meteorological information and air traffic control information for each first flight segment to determine whether it will affect flight safety;

[0052] If there is a safety impact, the flight segment will be replanned and the route will be adjusted to bypass the meteorological danger zone and avoid the waypoints in the danger zone. At the same time, new waypoints will be defined and combined to generate a new flight segment.

[0053] Generate flight paths that meet safety requirements.

[0054] Specifically, if Figure 4 As shown, the entire route is divided into 10-meter segments, and meteorological and air traffic control information is requested for each segment. For example, if a segment contains various severe weather phenomena that endanger aircraft safety, such as cumulonimbus clouds, active thunderstorms, hail, tropical storms, severe squall lines, tornadoes, strong sandstorms, high winds, heavy snow, heavy rain, freezing rain, low clouds, low visibility, cloud cover over mountains, downdrafts, strong wind shear, severe turbulence, and severe icing, the system analyzes each phenomenon based on its quantitative hazard level to determine whether it impacts flight safety. If so, the corresponding flight segment is replanned and route adjustments are made to circumvent the meteorological hazard area. The system automatically finds the nearest waypoint to avoid the hazard area, or generates custom waypoints based on an algorithm and combines these new waypoints to create a new segment.

[0055] The step three includes:

[0056] The two-dimensional plane data in the flight route obtained in step 2 is converted into three-dimensional data through an obstacle avoidance algorithm, wherein the basis for data conversion includes at least: airport altitude, cruising altitude and global high-precision terrain elevation data.

[0057] Specifically, if Figure 5 As shown in the figure, the route obstacle avoidance algorithm based on the safe altitude and high-precision elevation data of the global terrain can only reflect the flight guidance of the horizontal route because the generated route is two-dimensional plane data. At this time, the obstacle avoidance algorithm extends the two-dimensional route into three-dimensional data, which can provide maximum safety guidance for vertical flight. At this time, the pilot needs to input the cruising altitude. The built-in algorithm of the equipment uses the airport altitude, cruising altitude, and global high-precision terrain elevation data to optimize the unsafe terrain in the route through the obstacle avoidance algorithm again, so that the route can accurately avoid dangerous terrain such as mountains, canyons, obstacles, etc., further improving flight safety.

[0058] Specifically, the aircraft, target point and each vertex of the polygonal obstacle are combined and connected through the visual graph method. The connecting straight lines are regarded as arcs. It is required that the lines between the aircraft and each vertex of the obstacle, between the target point and each vertex of the obstacle, and between the vertices of the obstacle cannot pass through the obstacle. In this way, the three-dimensional route data is obtained and the route is available.

[0059] Visibility graph methods assume the route and obstacles are a point mass and an approximate polygon, respectively. Visible line segments are used to connect the point mass, target point, and polygon vertices in an orderly fashion. This means that the connecting lines cannot cross the obstacle polygon. This transforms the route path planning problem into a search for the shortest set of line segments between the starting point and the target point. Improved algorithms are often used to search for the shortest path. Because the searched path consists of line segments connecting obstacle vertices, the aircraft is prone to collision with obstacles. When there are many obstacles and feature information in the environment, the path search takes a long time. Furthermore, visibility graph methods lack flexibility, adaptability, and real-time performance, and cannot guarantee that the path they find is the global optimal one.

[0060] Therefore, the path planning algorithm based on the visibility graph method mainly includes the following two steps:

[0061] ① Construction of visual graph;

[0062] ② Use some optimization method to search for the optimal path on the constructed visibility graph;

[0063] In the visibility graph algorithm, obstacles are described by polygons, and the starting point S, the target point G and the vertices V of the polygon obstacle are o, as the vertices V of the visible graph, connect these vertices to each other, and keep the lines that do not cross obstacles as the edges E of the visible graph. Then, weight these edges according to some criteria, such as using the length of these edges as their weights. Then, use some optimization method to search for the required optimal path on the constructed visible graph. According to the above process, it is easy to know that the final result is a set of vertices including S and G. These vertices are connected in order to form the obtained path, such as Figure 7 shown.

[0064] From the above process, we can see that the key to constructing a visibility graph is to determine whether the line connecting any two vertices passes through an obstacle. We can simply and crudely detect whether the line connecting any two vertices passes through an obstacle, or we can use the following method to simplify the judgment:

[0065] ① The lines connecting adjacent vertices of the same obstacle must not pass through the obstacle (i.e., a certain boundary of the obstacle). Non-adjacent vertices of the same obstacle are generally considered to pass through the obstacle (however, for concave polygon obstacles, this may sometimes lead to misjudgment, because the lines connecting some non-adjacent nodes of concave polygons also do not pass through the obstacle).

[0066] ② The judgment of whether the vertices between different obstacles pass through the obstacle can be transformed into judging whether the line connecting the vertices intersects with the line connecting the two vertices of the edge forming the obstacle. For example, in the figure below, the line between V1 and V7 passes through the line between V2 and V3, so the line between V1 and V7 passes through the obstacle. Figure 8 shown.

[0067] The path found by the visibility graph method is close to the edge of the obstacle. You can use the method of extending the vertices of the obstacle outward by a set radius. The effect is as follows: Figure 9 As shown:

[0068] The method further comprises:

[0069] Generate preferred flight routes:

[0070] Generate an initial database based on departure and arrival airports,

[0071] Among them, the device stores the civil aviation navigation database, which is the initial database, and has built-in civil aviation routes, which can match the routes.

[0072] According to the pilot route system evaluation table, the option scores in the pilot route system evaluation table are weighted and calculated to generate the score of the route segment. The evaluation table is as follows: Figure 6 As shown;

[0073] Each time a route is generated later, the routes retrieved based on the departure and landing airports will be ranked by score, with routes with higher scores being used first.

[0074] Specifically, an initial database is generated using the departure and arrival airports. A weighted calculation is performed based on the scores of each option according to the pilot route system evaluation table. The scoring weights are shown in Table 1. This generates a score for this route. The next time a route is generated, the routes retrieved based on the departure and arrival airports are prioritized for ranking, with routes with higher scores being used first.

[0075] As shown in Table 1:

[0076] project Weight Route comfort 15％ Route safety 30％ Route integrity 25％ Does the route meet demand? 20％ Overall evaluation 10％

[0077] Table 1

[0078] Specifically, AI big data algorithms are used to self-learn and generate optimal flight routes for continuous optimization;

[0079] Its main contents include:

[0080] Support: The support of an itemset is defined as the proportion of records containing the itemset in the dataset to the total records.

[0081] Confidence is defined for a route association rule. The confidence of this rule is defined as "support(route {01,02}) / support({01})." A confidence of 0.75 means that for all records containing "01," the association rule applies to 75% of them.

[0082] When generating optimal flight routes, the goal is to find a set of key waypoints that frequently appear at takeoff and landing airports. The support of a set is used to measure its frequency of occurrence. The support of a set refers to the percentage of routes that contain it.

[0083] 1) First, a list of itemsets of all single waypoints is generated;

[0084] 2) Check which item sets meet the minimum support requirement through past route records, and those that do not meet the minimum support will be removed;

[0085] 3) Combine the remaining sets to generate an itemset containing two elements;

[0086] 4) Next, re-query the previous route records and remove the item sets that do not meet the minimum support. Repeat until all item sets are removed.

[0087] To find association rules, we start with a frequent item set. We want to know whether the elements in the frequent item set can be used to obtain other content, that is, whether a certain element or a certain set can be used to infer another element. If there is a frequent item set {waypoint 1, waypoint 2}, then there may be an association rule "waypoint 1 --> waypoint 2", which means that if waypoint 1 is selected, there is a statistically high probability of connecting to waypoint 2. However, the reverse is not necessarily true.

[0088] Based on the frequent itemset, a list of possible rules is generated. The credibility of each rule is then tested. If the credibility does not meet the minimum requirement, the rule is removed. Similar to the generation of frequent itemsets mentioned above, a frequent itemset can generate many possible association rules. If the number of rules can be reduced before calculating the credibility of the rules, the computational efficiency will be greatly improved.

[0089] One of the rules is: if a rule does not meet the minimum credibility requirement, then all subsets of the rule will not meet the minimum credibility requirement;

[0090] Once the confidence threshold is lowered, more rules can be obtained, thereby improving the accuracy and processing speed of the self-learning algorithm.

[0091] The scoring options of the pilot route system evaluation form include at least:

[0092] Route comfort, route safety and route integrity.

[0093] The use of global high-precision terrain elevation data to complete three-dimensional route obstacle avoidance includes:

[0094] The pilot enters the cruising altitude of the flight;

[0095] According to the cruising altitude, data comparison is performed with global high-precision terrain elevation data;

[0096] Dividing the flight route into flight segments to obtain a plurality of second flight segments;

[0097] Each second flight segment is tested. If the cruising altitude of the current second flight segment minus the terrain altitude is less than 500m, the current second flight segment is replanned.

[0098] The cruising altitude is consistent in each section and does not involve departure, arrival and approach procedures.

[0099] When there is a waypoint within the set range of the second flight segment and the obstacle can be successfully avoided, the waypoint is used for detour; when there is no waypoint within the set range, the optimal route of the current second flight segment is calculated, several custom waypoints are generated, and a complete route is generated by combining the flight cruising speed and cruising altitude with the custom waypoints, thereby completing the route planning with three-dimensional data.

[0100] Specifically, if Figure 5 As shown, the pilot inputs the flight cruising altitude, and the data is compared with the global high-precision terrain elevation data based on the cruising altitude. The terrain is detected every 30 meters. When it is found that the cruising altitude minus the terrain height is less than 500M, the current segment is replanned. When there are waypoints nearby and obstacles can be successfully avoided, the aircraft is detoured through the nearby waypoints, that is, the waypoints within the set range. When there are no waypoints within the set range, the optimal route for the segment is automatically calculated, and several custom points are generated. A complete route can be generated by combining the flight cruising speed and cruising altitude with these custom points, but the route must meet the flight requirements, thereby realizing three-dimensional route planning and completing the planning of the entire flight route.

[0101] The global high-precision terrain elevation data is tile data with an accuracy of 30m, which is used for stereo analysis of three-dimensional routes.

[0102] Specifically, tile data is a cached image set formed by preprocessing vector or image data using an efficient caching mechanism (such as a pyramid), which is organized in a "level, row, and column" manner and can be quickly loaded on a web page. Therefore, tile map loading is based on the map range and level requested by the client, and the tiles of the grid at the corresponding level (i.e., the images pre-cropped by the server) are obtained by calculating the row and column numbers, and a map is formed on the client from these tile sets. Tile data with an accuracy of 30m is used in this application for stereo analysis of three-dimensional routes.

[0103] The above descriptions are merely a few embodiments of the present application and do not constitute any form of limitation to the present application. Although the present application discloses the preferred embodiments as above, they are not intended to limit the present application. Any technical personnel familiar with the present profession, without departing from the scope of the technical solution of the present application, using the technical content disclosed above to make slight changes or modifications are equivalent to equivalent implementation cases and fall within the scope of the technical solution.

Claims

1. A method for intelligently generating flight routes for flight management equipment, characterized in that: The steps include: Step 1: Generate a basic route through the flight management device, where the basic route is generated based on at least the conditions of the departure airport, landing airport, flight direction, civil aviation route and navigation database; Step 2: Optimize the basic route using a deep learning network model: Based on historical route generation experience, associated learning rules, and combined with weather information and air traffic control information corresponding to the basic route obtained by the network, the route is optimized and adjusted to generate a flight route that meets safety requirements; Step 3: Use global high-precision terrain elevation data to complete three-dimensional route obstacle avoidance; The second step includes: Dividing the basic route into flight segments to obtain a plurality of first flight segments, and requesting weather information and air traffic control information for each first flight segment; Conduct a quantitative analysis of the risk level of meteorological information and air traffic control information for each first flight segment to determine whether it will affect flight safety; If there is a safety impact, the flight segment is replanned and the route is adjusted to bypass the meteorological risk area and avoid waypoints in the risk area. At the same time, new waypoints are defined and combined to generate a new flight segment; thus, a flight route that meets safety requirements is generated; The use of global high-precision terrain elevation data to complete three-dimensional route obstacle avoidance includes: The pilot enters the cruising altitude for the flight path; According to the cruising altitude, global high-precision terrain elevation data is compared; Dividing the flight route into flight segments to obtain a plurality of second flight segments; Each second flight segment is tested. If the cruising altitude of the current second flight segment minus the terrain altitude is less than 5000m, the current second flight segment is replanned. If there is a waypoint within the set range of the second flight segment and the obstacle can be successfully avoided, the waypoint is used for detour. If there is no waypoint within the set range, the optimal route for the current second flight segment is calculated, a number of custom waypoints are generated, and a complete route is generated by combining the flight cruising speed and cruising altitude with the custom waypoints, thereby completing the route planning with three-dimensional data. The global high-precision terrain elevation data has a 30m accuracy tile data, which is used for stereo analysis of three-dimensional routes; Utilize AI big data algorithms to self-learn and generate optimal flight routes for continuous optimization. The main contents include: 1) First, a list of itemsets of all single waypoints is generated; 2) Check which item sets meet the minimum support requirement through past route records, and those that do not meet the minimum support will be removed; 3) Combine the remaining sets to generate an itemset containing two elements; 4) Next, re-query the previous route records and remove the itemsets that do not meet the minimum support. Repeat this process until all itemsets are removed. To find association rules, start with a frequent itemset. An element or a set will deduce another element. If there is a frequent itemset {waypoint 1, waypoint 2}, then there is an association rule "waypoint 1-->waypoint 2", which means that if waypoint 1 is selected, then the probability of connecting waypoint 2 is statistically higher. Based on the frequent itemset, a list of possible rules is generated, and then the credibility of each rule is tested. If the credibility does not meet the minimum requirement, the rule is removed. Similar to the frequent itemset generation mentioned above, a frequent itemset can generate many possible association rules. If the number of rules can be reduced before calculating the credibility of the rules, the computational efficiency will be greatly improved. One of the rules is: if a rule does not meet the minimum credibility requirement, then all subsets of the rule will not meet the minimum credibility requirement; once the credibility threshold is lowered, more rules can be obtained, thereby improving the accuracy and processing speed of the self-learning algorithm.

2. A flight management equipment route intelligent generation method according to claim 1, characterized in that: The step three includes: The two-dimensional plane data of the flight path obtained in step 2 is converted into three-dimensional data through the obstacle avoidance algorithm. The basis for data conversion includes at least: airport altitude, cruising altitude and global high-precision terrain elevation data.

3. The method for intelligently generating flight routes for flight management equipment according to claim 1, characterized in that: The method further comprises: Generate preferred flight routes: Generate an initial database based on departure and arrival airports; According to the pilot route system evaluation table, weighted calculation is performed on the option scores in the pilot route system evaluation table to generate a score for the route segment; Each time a route is generated later, the routes retrieved based on the departure and landing airports will be ranked by score, with routes with higher scores being used first.

4. The method for intelligently generating flight routes for flight management equipment according to claim 3, characterized in that: The scoring options of the pilot route system evaluation form include at least: Route comfort, route safety and route integrity.

Images