Aircraft conflict resolution method based on knowledge graph

Abstract

The invention discloses an aircraft conflict resolution method based on a knowledge graph, and belongs to the technical field of air traffic control, and the method comprises the steps: obtaining the air dynamic data and airspace environment data of an aircraft; acquiring a control rule, and analyzing the control rule into structured data; knowledge extraction is carried out on multi-source heterogeneous historical operation data composed of aerial dynamic data and airspace environment data of an aircraft, and entities, attributes and relationships are extracted; establishing a flight conflict resolution knowledge graph by adopting a knowledge fusion method; the air traffic situation is evaluated according to the established flight conflict resolution knowledge graph, whether flight conflicts exist or not is judged, whether a resolution strategy is executed or not is judged, and an air traffic situation evaluation model is established to evaluate the complexity of the air traffic situation. The invention provides a reliable and practical aircraft conflict resolution method, the process of judging and executing aircraft conflict resolution by an operator is simplified, and more efficient aircraft conflict resolution operation is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of air traffic control, and in particular to an aircraft conflict resolution method based on a knowledge graph. Background Art

[0002] During air traffic operations, controllers will estimate conflict-free trajectories for each aircraft in advance and issue control instructions. However, due to the continuous dynamic changes in factors such as bad weather, the operating conditions of air traffic control facilities and equipment, and personnel operations, aircraft conflicts may still occur, especially in airspace with frequent ascent / descent and limited maneuvering space. Aircraft conflict resolution is the last line of defense for control operation safety. When an aircraft conflict is detected, the controller must rely on control experience to quickly determine the conflict resolution method. The current system cannot provide an aircraft conflict resolution solution, and the method of relying solely on controller decisions to implement aircraft conflict control is prone to decision-making errors and is inefficient, and is not suitable for complex and dynamic operating environments.

[0003] Scholars at home and abroad have proposed a large number of methods for aircraft conflict detection and resolution. In the early days of aircraft conflict detection research, deterministic conflict detection methods were mainly used for conflict detection. This method predicts the future trajectory of an aircraft based on its current trajectory and determines whether a conflict has occurred. This method has low conflict detection accuracy and a high false alarm rate. Accordingly, some deterministic algorithms have also been applied to conflict resolution. As the research deepened, researchers realized that uncertain factors such as weather, navigation errors, and flight intentions also have an important impact on the accuracy of trajectory prediction. Based on this, a probabilistic conflict detection method was proposed to quantify the possibility of a conflict. The accuracy of this method is affected by uncertain factors such as flight intentions.

[0004] At present, optimization algorithms such as genetic algorithm, ant colony algorithm, and optimal control theory are used to design aircraft conflict resolution methods. The aircraft conflict resolution methods based on traditional optimization algorithms lack systematic consideration of the controller's knowledge level and practical experience. The proposed aircraft conflict resolution methods are inconsistent with the controller's operating habits and may even interfere with the controller's decision-making.

[0005] Knowledge Graph is a modern theory that achieves the purpose of multidisciplinary integration by combining the theories and methods of applied mathematics, graphics, information visualization technology, information science and other disciplines with metrological citation analysis, co-occurrence analysis and other methods, and using visual graphs to vividly display the core structure, development history, frontier fields and overall knowledge architecture of the discipline. Summary of the invention

[0006] The present invention aims to establish a set of aircraft conflict resolution knowledge graphs taking into account existing control rules and provide a reliable and practical aircraft conflict resolution method.

[0007] To achieve the above object, the technical solution of the present invention is:

[0008] The method for resolving aircraft conflicts based on knowledge graph includes the following steps:

[0009] Step S1, obtaining the aerial dynamic data and airspace environment data of the aircraft;

[0010] Step S2, obtaining control rules and parsing unstructured control rules into structured data;

[0011] Step S3, extracting knowledge from multi-source heterogeneous historical operation data consisting of aircraft air dynamic data, airspace environment data, and structured control rules, and extracting entities, attributes, and relationships;

[0012] Step S4, using knowledge fusion method to solve the problems of redundancy, conflict and non-standardization of multi-heterogeneous historical operation data in the extracted entities, attributes and relationships;

[0013] Step S5, establishing an inference decision tree model based on the established flight conflict resolution knowledge graph, and determining whether there is a flight conflict between each pair of aircraft;

[0014] Step S6, based on the inference decision tree model, determine whether there is a flight conflict between the aircraft pairs, evaluate the difficulty of conflict resolution for each aircraft with flight conflict, sort the aircraft pairs according to the difficulty of conflict resolution, and use the control strategies of the inference decision tree model in order from difficult to easy to resolve the conflict.

[0015] In a preferred embodiment of the present invention, the step S3 further comprises:

[0016] The entities constitute an entity set E (Atc, Fli, Tsd, Rul, Beh), where Atc represents the controller entity, Fli represents the aircraft entity, Tsd represents the air traffic situation entity, Rul represents the control rule entity, and Beh represents the control strategy entity; each entity has multiple attributes.

[0017] In a preferred embodiment of the present invention, the conflict coordination difficulty is evaluated, specifically including, based on the acquired multi-source heterogeneous historical operation data and structured data, an air traffic situation assessment model establishes an air traffic complexity measurement index system based on the number of aircraft, dynamic characteristics, immediate conflicts between aircraft, future conflicts between aircraft, and the difficulty of conflict coordination between aircraft; the dynamic characteristics include: the number of heading changes, speed changes, and altitude changes of the aircraft within the evaluation time.

[0018] In a preferred embodiment of the present invention, the calculation method of the instantaneous conflict between aircraft is:

[0019] Define an aircraft pair consisting of aircraft i and aircraft j, where the positions of aircraft i and aircraft j are P i and P j , the velocity vectors of aircraft i and aircraft j are V i and V j , then the relative distance d of the pair of aircraft ij for:

[0020]

[0021] In the formula, x j (t) is the horizontal coordinate of aircraft j at time t, y j (t) is the ordinate of aircraft j at time t, x i (t) is the horizontal coordinate of aircraft i at time t, y i (t) is the ordinate of aircraft i at time t;

[0022] The relative speed V of the aircraft pair ij for:

[0023]

[0024] Using the exponential function P i,j (t) Define the proximity of aircraft i and aircraft j in the aircraft pair;

[0025]

[0026] In the formula, α is the weighting coefficient, D sep A pre-set safety interval;

[0027] The immediate inter-aircraft conflict Ic(i) of aircraft i is:

[0028]

[0029] Where m is the total number of aircraft.

[0030] In a preferred embodiment of the present invention, an exponential function Fc is used. i,j (t) reflects the future conflict between aircraft, exponential function Fc i,j (t) is:

[0031]

[0032]

[0033] Where D i,j (t) represents the distance between aircraft i and aircraft j at the time t;

[0034] The future conflict Fc(i) of aircraft i is:

[0035]

[0036] In a preferred embodiment of the present invention, the relative distance gradient change is measured by an aircraft. To reflect the difficulty of conflict coordination between aircraft, aircraft's relative distance gradient changes for:

[0037]

[0038]

[0039] In the formula, v i and v j are the speeds of aircraft i and aircraft j, respectively, θ i and θ j are the headings of aircraft i and aircraft j respectively, Δ x =x j -x i , Δ y =y j -y i ;

[0040] The difficulty of conflict coordination between aircraft Sd(i) is:

[0041]

[0042] In a preferred embodiment of the present invention, the aircraft conflict resolution method includes establishing branches of a decision tree from the aircraft's flight status (ascending, level flight and descending), heading relationship (same track, opposite track and intersecting track) and altitude relationship (same altitude and different altitudes); the root nodes of the decision tree are respectively same-direction track conflict, opposite track conflict and intersecting track conflict; the intermediate nodes of the decision tree are the aircraft's flight status and altitude relationship; the leaf nodes of the decision tree are the controller's control intention, and the control intention includes adjusting altitude, adjusting speed and adjusting heading.

[0043] In a preferred embodiment of the present invention, the processing logic of the decision tree for the same-direction track conflict is:

[0044] When the horizontal distance between the two aircraft is less than or equal to the preset horizontal safety interval, and the vertical distance between the two aircraft is less than or equal to the preset vertical safety interval, the corresponding air traffic control instruction is selected according to the real-time status of the two aircraft; the air traffic control instruction includes:

[0045] When both aircraft are ascending, the leading aircraft ascends to a preset altitude, and the trailing aircraft descends to a preset altitude;

[0046] When one aircraft is ascending and the other is descending, the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude.

[0047] When one aircraft is ascending and the other is flying level, the ascending aircraft ascends to a preset altitude and the flying level aircraft descends to a preset altitude;

[0048] When both aircraft are descending, the leading aircraft descends to a preset altitude, while the trailing aircraft ascends to a preset altitude;

[0049] When one aircraft is descending and the other aircraft is flying level, the descending aircraft descends to a preset altitude and the flying aircraft ascends to a preset altitude;

[0050] When both aircraft are in level flight, the leading aircraft ascends to a preset altitude and the trailing aircraft descends to a preset altitude.

[0051] In a preferred embodiment of the present invention, the processing logic of the decision tree for the cross-track conflict is:

[0052] When the horizontal distance between the two aircraft is less than or equal to the preset horizontal safety interval, the corresponding air traffic control instruction is selected according to the real-time status of the two aircraft and the horizontal distance between the two aircraft; the air traffic control instruction includes:

[0053] When the vertical distance between two aircraft is less than or equal to the preset vertical safety interval:

[0054] When both aircraft are ascending, the aircraft with better performance ascends to a preset altitude, while the aircraft with worse performance descends to a preset altitude;

[0055] When one aircraft is ascending and the other aircraft is descending, the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude;

[0056] When one aircraft is ascending and the other is flying level, the ascending aircraft ascends to a preset altitude and the flying level aircraft descends to a preset altitude;

[0057] When both aircraft are descending, the aircraft with relatively better performance descends to the preset altitude, while the aircraft with relatively worse performance ascends to the preset altitude;

[0058] When one aircraft is descending and the other aircraft is flying level, the descending aircraft descends to a preset altitude and the flying aircraft ascends to a preset altitude;

[0059] When both aircraft are in level flight, the aircraft with relatively better performance ascends to a preset altitude, while the aircraft with relatively worse performance descends to a preset altitude;

[0060] When the vertical distance between two aircraft is less than or equal to the preset vertical safety interval:

[0061] When both aircraft are ascending, the aircraft with a relatively far intersection point descends to a preset altitude, and the aircraft with a relatively close intersection point ascends to a preset altitude;

[0062] When one aircraft is ascending and the other aircraft is descending, the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude;

[0063] When one aircraft is ascending and the other is flying level, the ascending aircraft ascends to a preset altitude and the flying level aircraft descends to a preset altitude;

[0064] When both aircraft are descending, the aircraft with a relatively far intersection point ascends to a preset altitude, and the aircraft with a relatively close intersection point descends to a preset altitude;

[0065] When one aircraft is descending and the other aircraft is flying level, the descending aircraft descends to a preset altitude and the flying aircraft ascends to a preset altitude;

[0066] When both aircraft are in level flight, the aircraft with a relatively far intersection point descends to a preset altitude, and the aircraft with a relatively close intersection point ascends to a preset altitude.

[0067] In a preferred embodiment of the present invention, the processing logic of the decision tree for reverse track conflict is: when the horizontal distance between the two aircraft is less than or equal to the preset horizontal safety interval, and the vertical distance between the two aircraft is less than or equal to the preset vertical safety interval, the corresponding air traffic control instruction is selected according to the real-time status of the two aircraft; the air traffic control instruction includes:

[0068] When both aircraft are ascending, the aircraft with better performance ascends to a preset altitude, while the aircraft with worse performance descends to a preset altitude;

[0069] When one aircraft is ascending and the other aircraft is descending, the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude;

[0070] When one aircraft is ascending and the other is flying level, the ascending aircraft ascends to a preset altitude and the flying level aircraft descends to a preset altitude;

[0071] When both aircraft are descending, the aircraft with relatively better performance ascends to a preset altitude, while the aircraft with relatively worse performance descends to a preset altitude;

[0072] When one aircraft is descending and the other aircraft is flying level, the descending aircraft descends to a preset altitude and the flying aircraft ascends to a preset altitude;

[0073] When both aircraft are in level flight, the aircraft with relatively better performance ascends to a preset altitude, while the aircraft with relatively worse performance descends to a preset altitude.

[0074] Beneficial effect: The present invention provides a set of reliable and practical aircraft conflict resolution methods based on the aircraft conflict resolution knowledge graph, simplifies the process of operators judging and executing aircraft conflict resolution, and realizes more efficient aircraft conflict resolution operations.

[0075] In order to make the above features and advantages of the invention more obvious and easy to understand, embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 The present invention is a flowchart of an aircraft conflict resolution method based on a knowledge graph.

[0077] Figure 2 It is a schematic diagram of the inference decision tree model in the present invention.

[0078] Figure 3 Schematic diagram of the decision tree model for resolving conflicts between same-direction tracks.

[0079] Figure 4 Schematic diagram of the decision tree model for cross-track conflict resolution.

[0080] Figure 5 Schematic diagram of the decision tree model for reverse track conflict resolution. DETAILED DESCRIPTION

[0081] In order to make the purpose and technical solution of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment 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.

[0082] Figure 1 The present invention relates to a flow chart of an aircraft conflict resolution method based on a knowledge graph, such as Figure 1 As shown, the aircraft conflict resolution method based on knowledge graph of the present invention includes the following steps.

[0083] Step S1, obtaining the aircraft's aerial dynamic data and airspace environment data.

[0084] The air dynamic data includes historical monitoring data and flight plan data, and the airspace environment data includes meteorological data, basic air traffic control data, and interactive command data. The historical monitoring data is mainly the aircraft position information detected by ground-based and satellite-based monitoring equipment, which is generally updated every 4 seconds or 1 second. The flight plan data is mainly the planned messages such as the aircraft's departure airport, destination airport, route and airway. The meteorological data is mainly the meteorological conditions such as precipitation and visibility at the airport or route. The basic air traffic control data is information such as the location of navigation stations and airways in the air, operating restrictions, etc. The interactive command data is the information issued by the controller to adjust the aircraft's altitude, speed, heading, etc.

[0085] The above-mentioned historical monitoring data and flight plan data belong to structured data, that is, they are all data composed of a fixed format and a clear structure, usually presented in a tabular form. The meteorological data, basic air traffic control data, and interactive instruction data in the airspace environment data belong to perception data. The above-mentioned data are heterogeneous and heterogeneous, but have a certain correlation. For example, meteorological data and basic air traffic control data often affect flight plan data, interactive instruction data, etc. The various types of perception data are themselves time-series related, for example, the meteorological data at adjacent moments are similar, and the historical monitoring data at adjacent moments are similar. Standardizing and uniformly representing the multivariate heterogeneous historical operation data composed of various types of structured data and perception data is conducive to providing data support for subsequent control reasoning and decision-making.

[0086] Step S2, obtaining control rules, and parsing unstructured control rules into structured data.

[0087] Control rules are unstructured data, which usually include the following aspects: descriptions of complex situations described in natural language, such as special control measures under specific meteorological conditions; descriptions of controllers' empirical judgments and verbal instructions; textual descriptions of the principles for handling some rare but critical flight conditions; descriptions of flexible control requirements for different types of aircraft in special airspaces or specific time periods; descriptions of multi-department coordination and communication mechanisms, lacking clear processes and division of responsibilities; general descriptions of response strategies for emergencies or emergencies, lacking specific operational steps and quantitative standards. Unstructured control rules can use named entity recognition (NER) algorithms to identify key entities in the text, such as aircraft types, airspace areas, altitude restrictions, etc.

[0088] Step S3, extracting knowledge from multi-source heterogeneous historical operation data consisting of aircraft air dynamic data, airspace environment data, and structured control rules, and extracting entities, attributes, and relationships.

[0089] The entities constitute an entity set E (Atc, Fli, Tsd, Rul, Beh), where Atc represents the controller entity, Fli represents the aircraft entity, Tsd represents the air traffic situation entity, Rul represents the control rule entity, and Beh represents the control strategy entity.

[0090] Each entity has multiple attributes. For example, the attributes of the air traffic situation entity include: flight number, aircraft type, departure airport, destination airport, route, etc.; the attributes of the aircraft entity include: aircraft type, climb rate, descent rate, maximum takeoff weight, cruising speed, etc. The relationship between each entity is different. For example, the relationship between the controller entity and the aircraft entity is a monitoring relationship, a command relationship, etc.

[0091] Step S4, using knowledge fusion method to solve the redundancy, conflict and irregularity problems in the extracted entities, attributes and relationships, and establish a flight conflict resolution knowledge graph.

[0092] Specifically, it is necessary to streamline and normalize the extracted entities, attributes, and relationships, remove noise, redundancy, and conflicts, and provide a high-quality data foundation for subsequent processing. The deep learning model Transformer architecture can be introduced to optimize and improve knowledge fusion. The Transformer architecture performs well in natural language processing tasks and can capture long-distance dependencies and contextual information, thereby more effectively handling complex knowledge fusion problems and improving the quality and accuracy of knowledge graphs. It has obvious advantages over traditional methods.

[0093] Specifically, the entities, attributes and relationships extracted from multi-dimensional heterogeneous historical operation data all need to be fused. For example, a specific aircraft has different forms in different data. In historical monitoring data, it is called a secondary code, in flight plan data it is called a flight number, call sign, etc., and in interactive command data it includes ascent, descent, ascent to a certain altitude, descent to a certain altitude, etc.

[0094] According to the controller's control and command process and main tasks, the control and command tasks are divided into monitoring, conflict judgment, deployment, and monitoring execution, and a knowledge graph for flight conflict resolution is established. The establishment of a knowledge graph for flight conflict resolution includes constructing and storing the data structures of "controller entity-monitoring relationship-air traffic situation entity", "aircraft-compliance relationship-control rule entity", "aircraft-execution relationship-control strategy", and "aircraft-formation relationship-control situation".

[0095] Step S5, establishing an inference decision tree model based on the established flight conflict resolution knowledge graph, and determining whether there is a flight conflict between each pair of aircraft.

[0096] Step S6, based on the inference decision tree model, determine whether there is a flight conflict between the aircraft pairs, evaluate the difficulty of conflict resolution for each aircraft with flight conflict, sort the aircraft pairs according to the difficulty of conflict resolution, and use the control strategies of the inference decision tree model in order from difficult to easy to resolve the conflict.

[0097] The assessment of the difficulty of the above-mentioned conflict resolution includes the following specific contents.

[0098] Based on the multi-source heterogeneous historical operation data and structured data obtained above, the air traffic situation assessment model establishes an air traffic complexity measurement index system based on the number of aircraft, dynamic characteristics, immediate conflicts between aircraft, future conflicts between aircraft, and difficulty of conflict coordination between aircraft. Among them, the dynamic characteristics specifically include: the number of heading changes (HC), speed changes (SC), and altitude changes (AC) of aircraft during the assessment time.

[0099] The microstructure between aircraft in the airspace also reflects the complexity of the airspace, and the immediate conflicts between aircraft, future conflicts between aircraft, and the difficulty of conflict coordination between aircraft are considered as airspace complexity assessment indicators. The relative position and relative speed of aircraft are the most basic elements for analyzing this impact.

[0100] Define an aircraft pair consisting of aircraft i and aircraft j, where the positions of aircraft i and aircraft j are P i and P j , the velocity vectors of aircraft i and aircraft j are V i and V j , then the relative distance d of the pair of aircraft ij for:

[0101]

[0102] In formula (1), x j (t) is the horizontal coordinate of aircraft j at time t, y j (t) is the ordinate of aircraft j at time t, x i (t) is the horizontal coordinate of aircraft i at time t, y i (t) is the ordinate of aircraft i at time t.

[0103] The relative speed V of the aircraft pair ij for:

[0104]

[0105] The relative distance d between the aircraft pairs ij The larger the relative distance d, the weaker the proximity of the two aircrafts; vice versa. ij Smaller than the preset safety interval Dsep When , it means that the collision risk is extremely prominent, the urgency of the approach between aircraft increases dramatically, and the control difficulty doubles. On the contrary, it means that the urgency between aircraft is not obvious. In order to reflect this nonlinear relationship, the exponential function P is used i,j (t) Define the proximity of aircraft i and aircraft j in the aircraft pair.

[0106]

[0107] In formula (3), α is the weighting coefficient.

[0108] The immediate inter-aircraft conflict Ic(i) of aircraft i is:

[0109]

[0110] In formula (4), m is the total number of aircraft.

[0111] Relative speed V of the aircraft pair ij Reflecting the degree of future conflict between aircraft, the convergence approach to urgency is a nonlinear process, using the exponential function Fc i,j (t) Reflecting this process, future conflicts between aircraft are:

[0112]

[0113]

[0114] In formula (5) and formula (6), D i,j (t) represents the distance between aircraft i and aircraft j at time t.

[0115] The future conflict Fc(i) of aircraft i is:

[0116]

[0117] For the same immediate conflict between aircraft and future conflict between aircraft, the difficulty of controller deployment is still different. |To reflect the difficulty of conflict coordination between aircraft.

[0118]

[0119]

[0120] In formula (8) and formula (9), v i and v j are the speeds of aircraft i and aircraft j, respectively, θ i and θ j are the headings of aircraft i and aircraft j respectively, Δ x =x j -x i , Δ y =y j -y i .

[0121] The difficulty of conflict coordination between aircraft Sd(i) is:

[0122]

[0123] Figure 2 Schematic diagram of the inference decision tree model in the present invention. Figure 2 As shown, the inference decision tree model specifically includes the following steps.

[0124] Step S10, determining an aircraft pair, that is, selecting two aircraft that need to be judged for flight conflict.

[0125] Step S20, determining whether the safety interval between the aircraft pairs is met, if the safety interval is met, proceeding to step S30, if not, proceeding to step S40.

[0126] The interval between aircraft can be divided into longitudinal interval, lateral interval and vertical interval, among which longitudinal interval and lateral interval are collectively referred to as horizontal interval. Aircraft conflict means that the horizontal interval and vertical interval between aircraft do not meet the prescribed safety interval at the same time, for example, the horizontal safety interval is 10km and the vertical safety interval is 300m.

[0127] Step S30, determine the flight type, and execute step S31, step S32, or step S33 according to the flight type.

[0128] Step S31, if the flight type is determined to be approach, approach control is performed.

[0129] Step S32, judging that the line type is departure, then performing departure control.

[0130] Step S33, judging that the row type is overflight, then performing overflight control.

[0131] Step S40, determining the heading relationship, and executing step S41, step S42, or step S43 according to the heading relationship.

[0132] Step S41, determining that there is a same-direction track conflict, and executing the same-track conflict resolution.

[0133] Step S42: determine that there is a cross-track conflict and execute cross-track conflict resolution.

[0134] Step S43: determine that it is a reverse track conflict and execute reverse track conflict resolution.

[0135] After the above steps are completed, return to step S10 to judge the next aircraft pair. A same-direction track conflict refers to an aircraft conflict when the tracks of two aircraft are parallel, or the track difference is less than 45 degrees, or greater than 315 degrees, and the tracks of the two aircraft are less than the safety interval. A reverse track conflict refers to an aircraft conflict when two aircraft fly in opposite directions, the interval is less than the safety interval, and the track difference is between 135 degrees and 225 degrees. A cross-track conflict refers to an aircraft conflict when the track difference between two aircraft is between 45 degrees and 135 degrees, or the track difference is between 225 degrees and 315 degrees, and the interval between the aircraft is less than the safety interval.

[0136] The intermediate nodes of the inference decision tree are the relationship between the flight status and altitude of the aircraft, and the leaf nodes are the control intentions of the controller, which include adjusting altitude, speed and heading.

[0137] Adjusting the altitude is to adjust the aircraft to fly at a new altitude layer according to the altitude layer division rules. In order to meet the needs of providing air traffic services, the controller will issue speed adjustment instructions to the aircraft according to the situation to ensure the spacing, reasonable sorting and overall control. At altitudes of 7600m and above, the horizontal speed is adjusted in multiples of 0.01 Mach number. At altitudes below 7600m, the horizontal speed is adjusted in multiples of the indicated airspeed of 20km / h or 10kt. Heading adjustment is to guide the aircraft to fly to the desired position, avoid conflict with other aircraft, or maintain or increase the lateral or longitudinal spacing of the aircraft to avoid bad weather and special airspace.

[0138] Figure 3 3 is a schematic diagram of a decision tree model for resolving conflicts in the same direction. The reasoning model for resolving conflicts in the same direction is shown in FIG3 . The processing logic of the reasoning decision tree for the conflicts in the same direction includes the following steps.

[0139] Step S411, determining that the two aircraft are on the same-direction track conflict.

[0140] Step S412: determine whether the horizontal distance D between the two aircraft is less than or equal to the preset horizontal safety interval ST hor , and when determining whether the vertical distance H between the two aircraft is less than or equal to the preset vertical safety interval ST ver If both of the above judgment results are yes, execute step S413, otherwise execute step S414.

[0141] Step S413, select the corresponding air traffic control instruction according to the real-time status of the two aircraft. That is, when the horizontal distance D between the two aircraft is less than or equal to the preset horizontal safety interval ST hor, and the vertical distance H between the two aircraft is less than or equal to the preset vertical safety interval ST ver , select the corresponding air traffic control instructions according to the real-time status of the two aircraft.

[0142] In step S413, the selectable air traffic control instructions include the following modes according to the real-time status of the two aircraft.

[0143] When both aircraft are ascending (i.e. ascending-ascending): the leading aircraft ascends to a preset altitude, and the trailing aircraft descends to a preset altitude. The formula is: h f1 =h f0 +sep,h r1 =h r0 -sep, where h f0 represents the original altitude of the aircraft ahead, h f1 Indicates the adjusted altitude of the aircraft ahead, h r0 represents the original altitude of the rear aircraft r, h r1 It indicates the adjusted altitude of the rear aircraft r, and sep indicates the altitude layer interval.

[0144] When one aircraft is ascending and the other is descending (i.e. ascending-descending): the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude. The formula is: h c1 =h c0 +sep,h d1 =h d0 -sep, where h c0 represents the original altitude of the ascending aircraft c, h c1 represents the adjusted altitude of the ascending aircraft c, h d0 represents the original altitude of the descending aircraft d, h d1 d represents the adjusted altitude of the descending aircraft, and sep represents the altitude layer interval.

[0145] When one aircraft is ascending and the other is level (i.e. ascend-level): the ascending aircraft ascends to a preset altitude, and the level-flying aircraft descends to a preset altitude. The formula is: h c1 =h c0 +sep,h m1 =h m0 -sep, where h c0 represents the original altitude of the ascending aircraft c, h c1 represents the adjusted altitude of the ascending aircraft c, h m0 represents the original altitude of the aircraft in level flight m, h m1 m represents the adjusted altitude of a level flight aircraft, and sep represents the altitude layer interval.

[0146] When both aircraft are descending (i.e., descending-descending): the leading aircraft descends to a preset altitude, and the trailing aircraft ascends to a preset altitude. The formula is: h f1 =h f0 -sep,h r1 =h r0 +sep, where h f0 represents the original altitude of the aircraft ahead, h f1 Indicates the adjusted altitude of the aircraft ahead, h r0 represents the original altitude of the rear aircraft r, h r1 It indicates the adjusted altitude of the rear aircraft r, and sep indicates the altitude layer interval.

[0147] When one aircraft is descending and the other is flying level (i.e., descending-leveling): the descending aircraft descends to a preset altitude, and the level-flying aircraft ascends to a preset altitude. The formula is: h d1 =h d0 -sep,h m1 =h m0 +sep, where h d0 represents the original altitude of the descending aircraft d, h d1 represents the adjusted altitude of the descending aircraft d, h m0 represents the original altitude of the aircraft in level flight m, h m1 m represents the adjusted altitude of a level flight aircraft, and sep represents the altitude layer interval.

[0148] When both aircraft are flying level (i.e., level-level): the leading aircraft ascends to a preset altitude, and the trailing aircraft descends to a preset altitude. The formula is: h f1 =h f0 -sep,h r1 =h r0 +sep, where h f0 represents the original altitude of the aircraft ahead, h f1 Indicates the adjusted altitude of the aircraft ahead, h r0 represents the original altitude of the rear aircraft r, h r1 It indicates the adjusted altitude of the rear aircraft r, and sep indicates the altitude layer interval.

[0149] Step S414, for aircraft without conflict, adjust the altitude according to the gradient, that is, gradually change the flight altitude with a slope or gradient determined by considering the aircraft performance, control rules, flight phase, and aircraft position relationship, to avoid flight conflict caused by the following aircraft crossing the altitude layer of the preceding aircraft.

[0150] Figure 4FIG4 is a schematic diagram of a decision tree model for cross-track conflict resolution. The cross-track conflict resolution intention reasoning model is shown in FIG4. The processing logic of the cross-track conflict reasoning decision tree includes the following steps.

[0151] Step S421, determining that two aircraft have a cross-track conflict.

[0152] Step S422: determine whether the horizontal distance D between the two aircraft is less than or equal to the preset horizontal safety interval ST hor If the result is yes, execute step S423, if the result is no, execute step S425.

[0153] Step S423, determine whether the vertical distance H between the two aircraft is less than or equal to the preset vertical safety interval ST ver If the result is yes, execute step S424, if the result is no, execute step S427.

[0154] Step S424, selecting corresponding air traffic control instructions according to the real-time status of the two aircrafts. Selecting corresponding air traffic control instructions according to the real-time status of the two aircrafts, the selectable air traffic control instructions include the following modes.

[0155] When both aircraft are at an ascending altitude (i.e., ascend-ascend): the aircraft with relatively better performance ascends to a preset altitude, and the aircraft with relatively worse performance descends to a preset altitude.

[0156] When one aircraft is ascending and the other is descending (i.e. ascending-descending): the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude. The formula is: h c1 =h c0 +sep,h d1 =h d0 -sep, where h c0 represents the original altitude of the ascending aircraft c, h c1 represents the adjusted altitude of the ascending aircraft c, h d0 represents the original altitude of the descending aircraft d, h d1 d represents the adjusted altitude of the descending aircraft, and sep represents the altitude layer interval.

[0157] When one aircraft is ascending and the other is level (i.e. ascend-level): the ascending aircraft ascends to a preset altitude, and the level-flying aircraft descends to a preset altitude. The formula is: h c1 =h c0 +sep,h m1 =h m0 -sep, where h c0 represents the original altitude of the ascending aircraft c, h c1 represents the adjusted altitude of the ascending aircraft c, hm0 represents the original altitude of the aircraft in level flight m, h m1 m represents the adjusted altitude of a level flight aircraft, and sep represents the altitude layer interval.

[0158] When both aircraft are descending (ie, descend-descend): the aircraft with relatively better performance descends to a preset altitude, and the aircraft with relatively worse performance ascends to a preset altitude.

[0159] When one aircraft is descending and the other is flying level (i.e., descending-leveling): the descending aircraft descends to a preset altitude, and the level-flying aircraft ascends to a preset altitude. The formula is: h d1 =h d0 -sep,h m1 =h m0 +sep, where h d0 represents the original altitude of the descending aircraft d, h d1 represents the adjusted altitude of the descending aircraft d, h m0 represents the original altitude of the aircraft in level flight m, h m1 m represents the adjusted altitude of a level flight aircraft, and sep represents the altitude layer interval.

[0160] When both aircraft are in level flight (i.e., level flight-level flight): the aircraft with relatively better performance ascends to a preset altitude, and the aircraft with relatively worse performance descends to a preset altitude.

[0161] The above-mentioned aircraft performance compares the performance of the two aircraft by considering the aircraft speed, ceiling, climb rate, stability and maneuverability.

[0162] Step S425, determining whether the vertical distance H between the two aircraft is less than or equal to the preset vertical safety interval ST ver If the result is yes, execute step S426, if the result is no, execute step S427.

[0163] Step S426, select the corresponding air traffic control instruction according to the real-time status of the two aircraft. The corresponding air traffic control instruction is selected according to the real-time status of the two aircraft. The selectable air traffic control instructions include the following modes. When both aircraft are at an ascending altitude (i.e., ascending-ascending): the aircraft with a relatively far intersection descends to a preset altitude, and the aircraft with a relatively close intersection ascends to a preset altitude. The formula is: h l1 =h l0 -sep,h s1 =h s0 +sep, where h l0 Indicates the original altitude of the aircraft l when the intersection point is relatively far away, h l1 Indicates that the intersection point is relatively far away. The adjusted height of the aircraft l, h s0Indicates the original altitude of the aircraft s relative to the intersection point, h s1 It indicates the adjusted altitude of aircraft s that is relatively close to the intersection point, and sep indicates the altitude layer interval.

[0164] When one aircraft is ascending and the other is descending (i.e. ascending-descending): the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude. The formula is: h c1 =h c0 +sep,h d1 =h d0 -sep, where h c0 represents the original altitude of the ascending aircraft c, h c1 represents the adjusted altitude of the ascending aircraft c, h d0 represents the original altitude of the descending aircraft d, h d1 d represents the adjusted altitude of the descending aircraft, and sep represents the altitude layer interval.

[0165] When one aircraft is ascending and the other is level (i.e. ascend-level): the ascending aircraft ascends to a preset altitude, and the level-flying aircraft descends to a preset altitude. The formula is: h c1 =h c0 +sep,h m1 =h m0 -sep, where h c0 represents the original altitude of the ascending aircraft c, h c1 represents the adjusted altitude of the ascending aircraft c, h m0 represents the original altitude of the aircraft in level flight m, h m1 m represents the adjusted altitude of a level flight aircraft, and sep represents the altitude layer interval.

[0166] When both aircraft are descending (i.e., descending-descending): the aircraft with a relatively far intersection point ascends to a preset altitude, and the aircraft with a relatively close intersection point descends to a preset altitude. The formula is: h l1 =h l0 +sep,h s1 =h s0 -sep, where h l0 Indicates the original altitude of the aircraft l when the intersection point is relatively far away, h l1 Indicates that the intersection point is relatively far away. The adjusted height of the aircraft l, h s0 Indicates the original altitude of the aircraft s relative to the intersection point, h s1 It indicates the adjusted altitude of aircraft s that is relatively close to the intersection point, and sep indicates the altitude layer interval.

[0167] When one aircraft is descending and the other is flying level (i.e., descending-leveling): the descending aircraft descends to a preset altitude, and the level-flying aircraft ascends to a preset altitude. The formula is: h d1 =h d0 -sep,h m1 =h m0 +sep, where h d0 represents the original altitude of the descending aircraft d, h d1 represents the adjusted altitude of the descending aircraft d, h m0 represents the original altitude of the aircraft in level flight m, h m1 m represents the adjusted altitude of a level flight aircraft, and sep represents the altitude layer interval.

[0168] When both aircraft are flying level (i.e., level-level): the aircraft with a relatively far intersection point descends to a preset altitude, and the aircraft with a relatively close intersection point ascends to a preset altitude. The formula is: h l1 =h l0 -sep,h s1 =h s0 +sep, where h l0 Indicates the original altitude of the aircraft l when the intersection point is relatively far away, h l1 Indicates that the intersection point is relatively far away. The adjusted height of the aircraft l, h s0 Indicates the original altitude of the aircraft s relative to the intersection point, h s1 It indicates the adjusted altitude of aircraft s that is relatively close to the intersection point, and sep indicates the altitude layer interval.

[0169] Step S427, for aircraft without conflict, adjust the altitude according to the gradient, that is, gradually change the flight altitude with a slope or gradient determined by considering the aircraft performance, control rules, flight phase, and aircraft position relationship, to avoid flight conflict caused by the following aircraft crossing the altitude layer of the preceding aircraft.

[0170] Figure 5 Schematic diagram of the decision tree model for reversing track conflict resolution. The inference model for the intention of reversing track conflict resolution is shown in Figure 5. The processing logic of the inference decision tree for reversing track conflict includes the following steps.

[0171] Step S431, determining that the two aircraft are in conflict with each other in the opposite direction.

[0172] Step S432: determine whether the horizontal distance D between the two aircraft is less than or equal to the preset horizontal safety interval ST hor , and when determining whether the vertical distance H between the two aircraft is less than or equal to the preset vertical safety interval ST ver If both of the above two judgment results are yes, execute step S433, otherwise execute step S434.

[0173] Step S433, select the corresponding air traffic control instruction according to the real-time status of the two aircraft. That is, when the horizontal distance D between the two aircraft is less than or equal to the preset horizontal safety interval ST hor , and the vertical distance H between the two aircraft is less than or equal to the preset vertical safety interval ST ver When the corresponding air traffic control instruction is selected according to the real-time status of the two aircrafts, the air traffic control instructions that can be selected in step S433 include the following modes.

[0174] When both aircraft are at an ascending altitude (i.e., ascend-ascend): the aircraft with relatively better performance ascends to a preset altitude, and the aircraft with relatively worse performance descends to a preset altitude.

[0175] When one aircraft is ascending and the other is descending (i.e. ascending-descending): the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude. The formula is: h c1 =h c0 +sep,h d1 =h d0 -sep, where h c0 represents the original altitude of the ascending aircraft c, h c1 represents the adjusted altitude of the ascending aircraft c, h d0 represents the original altitude of the descending aircraft d, h d1 d represents the adjusted altitude of the descending aircraft, and sep represents the altitude layer interval.

[0176] When one aircraft is ascending and the other is level (i.e. ascend-level): the ascending aircraft ascends to a preset altitude, and the level-flying aircraft descends to a preset altitude. The formula is: h c1 =h c0 +sep,h m1 =h m0 -sep, where h c0 represents the original altitude of the ascending aircraft c, h c1 represents the adjusted altitude of the ascending aircraft c, h m0 represents the original altitude of the aircraft in level flight m, h m1 m represents the adjusted altitude of a level flight aircraft, and sep represents the altitude layer interval.

[0177] When both aircraft are descending (ie, descend-descend): the aircraft with relatively better performance ascends to a preset altitude, and the aircraft with relatively worse performance descends to a preset altitude.

[0178] When one aircraft is descending and the other is flying level (i.e., descending-leveling): the descending aircraft descends to a preset altitude, and the level-flying aircraft ascends to a preset altitude. The formula is: h d1 =hd0 -sep,h m1 =h m0 +sep, where h d0 represents the original altitude of the descending aircraft d, h d1 represents the adjusted altitude of the descending aircraft d, h m0 represents the original altitude of the aircraft in level flight m, h m1 m represents the adjusted altitude of a level flight aircraft, and sep represents the altitude layer interval.

[0179] When both aircraft are in level flight (i.e., level flight-level flight): the aircraft with relatively better performance ascends to a preset altitude, and the aircraft with relatively worse performance descends to a preset altitude.

[0180] Step S434, for aircraft without conflict, adjust the altitude according to the gradient, that is, gradually change the flight altitude with a slope or gradient determined by considering the aircraft performance, control rules, flight phase, and aircraft position relationship, to avoid flight conflicts caused by the following aircraft crossing the altitude layer of the preceding aircraft.

[0181] In summary, the present invention provides a set of reliable and practical aircraft conflict resolution methods based on the aircraft conflict resolution knowledge graph, simplifies the process of operators judging and executing aircraft conflict resolution, and realizes more efficient aircraft conflict resolution operations.

[0182] Although the present invention has been disclosed as above by way of embodiments, it is not intended to limit the present invention. Any person having ordinary knowledge in the technical field may make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be determined by the scope of the attached patent application.

Claims

1. The method for resolving aircraft conflicts based on knowledge graph is characterized by: The following steps are included: Step S1, obtaining the aerial dynamic data and airspace environment data of the aircraft; Step S2, obtaining control rules and parsing unstructured control rules into structured data; Step S3, extracting knowledge from multi-source heterogeneous historical operation data consisting of aircraft air dynamic data, airspace environment data, and structured control rules, and extracting entities, attributes, and relationships; Step S4, using knowledge fusion method to solve the problems of redundant, conflicting and irregular operation data in the extracted entities, attributes and relationships, and establishing a knowledge graph for flight conflict resolution; Step S5, establishing an inference decision tree model based on the established flight conflict resolution knowledge graph, and determining whether there is a flight conflict between each pair of aircraft; Step S6, based on the inference decision tree model, determine whether there is a flight conflict between the aircraft pairs, evaluate the difficulty of conflict resolution for each aircraft with flight conflict, sort the aircraft pairs according to the difficulty of conflict resolution, and use the control strategies of the inference decision tree model in order from difficult to easy to resolve the conflict.

2. The method for resolving aircraft conflicts based on a knowledge graph as claimed in claim 1, characterized in that: The step S3 further comprises: The entities constitute an entity set E (Atc, Fli, Tsd, Rul, Beh), where Atc represents the controller entity, Fli represents the aircraft entity, Tsd represents the air traffic situation entity, Rul represents the control rule entity, and Beh represents the control strategy entity; each entity has multiple attributes.

3. The method for resolving aircraft conflicts based on knowledge graph according to claim 1, characterized in that: The difficulty of conflict resolution is assessed, including: Based on the acquired multi-source heterogeneous historical operation data and structured data, the air traffic situation assessment model establishes an air traffic complexity measurement index system based on the number of aircraft, dynamic characteristics, immediate conflicts between aircraft, future conflicts between aircraft, and difficulty of conflict coordination between aircraft; The dynamic characteristics include: the number of heading changes, the number of speed changes, and the number of altitude changes of the aircraft within the evaluation time.

4. The method for resolving aircraft conflicts based on a knowledge graph as claimed in claim 3, characterized in that: The calculation method of the immediate conflict between aircraft is as follows: Define an aircraft pair consisting of aircraft i and aircraft j, where the positions of aircraft i and aircraft j are P i and P j , the velocity vectors of aircraft i and aircraft j are V i and V j , then the relative distance d of the pair of aircraft ij for: In the formula, x j (t) is the horizontal coordinate of aircraft j at time t, y j (t) is the ordinate of aircraft j at time t, x i (t) is the horizontal coordinate of aircraft i at time t, y i (t) is the ordinate of aircraft i at time t; The relative speed V of the aircraft pair ij for: Using the exponential function P i,j (t) Define the proximity of aircraft i and aircraft j in the aircraft pair; In the formula, α is the weighting coefficient, D sep A pre-set safety interval; The immediate inter-aircraft conflict Ic(i) of aircraft i is: Where m is the total number of aircraft.

5. The method for resolving aircraft conflicts based on knowledge graph according to claim 4, characterized in that: Use the exponential function Fc i,j (t) reflects the future conflict between aircraft, exponential function Fc i,j (t) is: Where D i,j (t) represents the distance between aircraft i and aircraft j at the time t; The future conflict Fc(i) of aircraft i is:

6. The method for resolving aircraft conflicts based on knowledge graph according to claim 5, characterized in that: Using aircraft to measure relative distance gradient To reflect the difficulty of conflict coordination between aircraft, aircraft's relative distance gradient changes for: In the formula, v i and v j are the speeds of aircraft i and aircraft j, respectively, θ i and θ j are the headings of aircraft i and aircraft j respectively, Δ x =x j -x i , Δ y =y j -y i ; The difficulty of conflict coordination between aircraft Sd(i) is:

7. The method for resolving aircraft conflicts based on knowledge graph according to claim 1, characterized in that: The aircraft conflict resolution method comprises the following steps: establishing branches of a decision tree from the flight status (ascending, level flight and descending), heading relationship (same track, opposite track and crossing track) and altitude relationship (same altitude and different altitude) of the aircraft; the root nodes of the decision tree are respectively the same track conflict, the opposite track conflict and the crossing track conflict; the intermediate nodes of the decision tree are the flight status and altitude relationship of the aircraft; the leaf nodes of the decision tree are the control intention of the controller, and the control intention includes adjusting the altitude, adjusting the speed and adjusting the heading.

8. The method for resolving aircraft conflicts based on knowledge graph as claimed in claim 7, characterized in that: The processing logic of the decision tree for the same-direction track conflict is: When the horizontal distance between the two aircraft is less than or equal to the preset horizontal safety interval, and the vertical distance between the two aircraft is less than or equal to the preset vertical safety interval, the corresponding air traffic control instruction is selected according to the real-time status of the two aircraft; the air traffic control instruction includes: When both aircraft are ascending, the leading aircraft ascends to a preset altitude, and the trailing aircraft descends to a preset altitude; When one aircraft is ascending and the other is descending, the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude. When one aircraft is ascending and the other is flying level, the ascending aircraft ascends to a preset altitude and the flying level aircraft descends to a preset altitude; When both aircraft are descending, the leading aircraft descends to a preset altitude, while the trailing aircraft ascends to a preset altitude; When one aircraft is descending and the other aircraft is flying level, the descending aircraft descends to a preset altitude and the flying aircraft ascends to a preset altitude; When both aircraft are in level flight, the leading aircraft ascends to a preset altitude and the trailing aircraft descends to a preset altitude.

9. The method for resolving aircraft conflicts based on knowledge graph as claimed in claim 7, characterized in that: The processing logic of the decision tree for the cross-track conflict is: When the horizontal distance between the two aircraft is less than or equal to the preset horizontal safety interval, the corresponding air traffic control instruction is selected according to the real-time status of the two aircraft and the horizontal distance between the two aircraft; the air traffic control instruction includes: When the vertical distance between two aircraft is less than or equal to the preset vertical safety interval: When both aircraft are ascending, the aircraft with better performance ascends to a preset altitude, while the aircraft with worse performance descends to a preset altitude; When one aircraft is ascending and the other aircraft is descending, the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude; When one aircraft is ascending and the other is flying level, the ascending aircraft ascends to a preset altitude and the flying level aircraft descends to a preset altitude; When both aircraft are descending, the aircraft with relatively better performance descends to the preset altitude, while the aircraft with relatively worse performance ascends to the preset altitude; When one aircraft is descending and the other aircraft is flying level, the descending aircraft descends to a preset altitude and the flying aircraft ascends to a preset altitude; When both aircraft are in level flight, the aircraft with relatively better performance ascends to a preset altitude, while the aircraft with relatively worse performance descends to a preset altitude; When the vertical distance between two aircraft is less than or equal to the preset vertical safety interval: When both aircraft are ascending, the aircraft with a relatively far intersection point descends to a preset altitude, and the aircraft with a relatively close intersection point ascends to a preset altitude; When one aircraft is ascending and the other aircraft is descending, the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude; When one aircraft is ascending and the other is flying level, the ascending aircraft ascends to a preset altitude and the flying level aircraft descends to a preset altitude; When both aircraft are descending, the aircraft with a relatively far intersection point ascends to a preset altitude, and the aircraft with a relatively close intersection point descends to a preset altitude; When one aircraft is descending and the other aircraft is flying level, the descending aircraft descends to a preset altitude and the flying aircraft ascends to a preset altitude; When both aircraft are in level flight, the aircraft with a relatively far intersection point descends to a preset altitude, and the aircraft with a relatively close intersection point ascends to a preset altitude.

10. The method for resolving aircraft conflicts based on knowledge graph according to claim 7, characterized in that: The processing logic of the decision tree for reverse track conflict is: when the horizontal distance between the two aircraft is less than or equal to the preset horizontal safety interval, and the vertical distance between the two aircraft is less than or equal to the preset vertical safety interval, the corresponding air traffic control instruction is selected according to the real-time status of the two aircraft; the air traffic control instruction includes: When both aircraft are ascending, the aircraft with better performance ascends to a preset altitude, while the aircraft with worse performance descends to a preset altitude; When one aircraft is ascending and the other aircraft is descending, the ascending aircraft ascends to a preset altitude and the descending aircraft descends to a preset altitude; When one aircraft is ascending and the other is flying level, the ascending aircraft ascends to a preset altitude and the flying level aircraft descends to a preset altitude; When both aircraft are descending, the aircraft with relatively better performance ascends to a preset altitude, while the aircraft with relatively worse performance descends to a preset altitude; When one aircraft is descending and the other aircraft is flying level, the descending aircraft descends to a preset altitude and the flying aircraft ascends to a preset altitude; When both aircraft are in level flight, the aircraft with relatively better performance ascends to a preset altitude, while the aircraft with relatively worse performance descends to a preset altitude.

Images