An airport flight area crossing traffic command system based on A-SMGCS system

Abstract

The application provides an airport flight area crossing traffic command system based on an A-SMGCS system, and belongs to the technical field of intelligent traffic, and comprises: a sensor layer, which is used for collecting aircraft dynamic data, to-be-passed authorization object data, environmental meteorological data and dynamic obstacle data; a data processing and decision-making layer, which is used for standardizing processing and space-time mapping processing of the collected data, judging existing potential conflicts in combination with a dynamic threshold adjustment mechanism, and generating and issuing instructions based on multi-objective optimization decision-making; and an execution layer, which is used for controlling crossing traffic control equipment according to the issued instructions, and feeding back execution results to the data processing and decision-making layer in real time to form a closed-loop control. The application effectively reduces the traffic conflict risk of the crossing of the airport flight area, improves the crossing efficiency, reduces the vehicle waiting time and the aircraft taxi delay, adapts to the fine scheduling demand of the airport scene movement, and provides reliable support for the safe and efficient operation of the airport flight area traffic.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation technology, and in particular to an airport flight area intersection traffic control system based on the A-SMGCS system. Background Technology

[0002] Airport runway crossings are crucial nodes connecting the flight area (aircraft takeoff and landing, taxiing core area) and the non-flying area. The safety and efficiency of traffic control at these crossings directly impact the overall operation of the airport. With the increase in airport throughput, the number of vehicles passing through these crossings, such as catering vehicles, maintenance vehicles, and emergency vehicles, has surged. Traditional crossing control relies on manual verification of permissions and manual judgment of conflicts between aircraft and other vehicles, resulting in problems such as delayed response and large judgment errors.

[0003] While there have been attempts to apply the A-SMGCS system to level crossing control in existing technologies, the following shortcomings still exist:

[0004] Lack of spatiotemporal mapping at bifurcation points: The correspondence between bifurcation points at flight area crossings (such as bifurcations in the direction of the apron or maintenance area) and associated taxiways is not clearly defined. Only a general judgment of overall crossroad conflicts is made, which cannot accurately locate the risks of specific bifurcation points and may easily lead to false prohibition of passage or missed judgment of conflicts.

[0005] The conflict classification and handling logic is unclear: it does not distinguish whether the aircraft taxiing path can be changed, and only generates prohibition / permission instructions. It cannot flexibly deal with scenarios such as aircraft path lock but vehicle can be detoured, and vehicle path fixed but aircraft speed can be adjusted, resulting in waste of crossing resources.

[0006] Instruction execution mismatch with branching points: Existing instructions only apply to the level crossing as a whole and do not specify corresponding branching points, which can easily lead to vehicles mistakenly entering conflicting branching points and increase the risk of ground accidents.

[0007] Therefore, this invention proposes an airport flight area crossing traffic control system based on the A-SMGCS system. Summary of the Invention

[0008] This invention provides an airport flight area crossing traffic control system based on the A-SMGCS system to solve the aforementioned technical problems.

[0009] This invention provides an airport flight area crossing traffic control system based on the A-SMGCS system, comprising:

[0010] The sensor layer is used to collect dynamic data of aircraft, data of objects awaiting authorization, environmental and meteorological data, and dynamic obstacle data.

[0011] The data processing and decision-making layer is used to standardize the collected data and perform spatiotemporal mapping processing according to the bifurcation of the target crossing in the flight area. It also uses a dynamic threshold adjustment mechanism to judge potential conflicts and generate and issue instructions based on multi-objective optimization decision-making. The bifurcation corresponds to a unique associated taxiway and is a lane bifurcation within the crossing that connects different target areas.

[0012] The execution layer is used to control the traffic control equipment at the intersection according to the issued instructions, and to feed back the execution results to the data processing and decision-making layer in real time to form a closed-loop control.

[0013] Preferably, the aircraft dynamic data includes: real-time coordinates, taxiing speed, locked taxiing trajectory, and predicted taxiing path for the next time period T, including the coordinate range of the associated taxiway;

[0014] The data of the object to be authorized to pass includes: identity identifier, target area, vehicle type and rated speed, and estimated time of entry into the crossing.

[0015] The environmental meteorological data includes: visibility;

[0016] The dynamic obstacle data includes: temporary construction areas and the location of disabled vehicles.

[0017] Preferably, the data processing and decision-making layer includes:

[0018] The standardization processing unit is used to standardize the collected data.

[0019] The time interval determination unit is used to determine the first time interval for the object to pass through the bifurcation and the second time interval for the aircraft to pass through the associated taxiway, based on aircraft dynamic data, data of the object to be authorized to pass and dynamic obstacle data.

[0020] The probability determination unit is used to determine the spatiotemporal intersection probability of each branch and the corresponding associated taxiway under the target crossing in the flight area based on the first time interval and the second time interval. At the same time, it dynamically sets the conflict threshold according to the conflict risk level, environmental meteorological data, traffic density and aircraft type.

[0021] The judgment unit is used to determine that there is no conflict at the corresponding bifurcation if the spatiotemporal intersection probability is less than the conflict threshold, and to generate a permission instruction.

[0022] Otherwise, determine the existence of potential conflicts;

[0023] Topology building units are used to extract the immutable and modifiable conflicts of the potential conflicts and construct a conflict topology graph.

[0024] An associated expansion unit is used to determine the maximum waiting time based on the immutable conflict, lock the immutable conflict at the first position in the conflict topology graph, and expand the associated node.

[0025] The decision generation unit is used to generate multi-objective optimization decisions based on rolling time-domain optimization, with the goal of maximizing the throughput of the A-SMGCS system in the future time.

[0026] Anchor point determination unit, used to determine conflict anchor points based on the maximum waiting time and the expanded nodes based on the multi-objective optimization decision;

[0027] The instruction matching unit is used to match the issued instruction that matches the conflicting anchor point from the anchor point-instruction type lookup table, wherein the issued instruction is a delayed passage instruction or a path adjustment instruction.

[0028] Preferably, when a passage permission instruction is issued, the corresponding green light at the fork in the road is turned on, the barrier is raised, a voice announcement is made, and a terminal instruction is pushed; when a passage delay instruction or a route adjustment instruction is issued, the yellow light is turned on, the delay duration is announced, or an alternative fork in the road is selected.

[0029] Preferably, the data processing and decision-making layer is also used to automatically trigger the emergency command mechanism when it receives a sudden deviation of an aircraft from its taxiway, a malfunction in the crossing area, or an emergency rescue request, generate a no-passing instruction, and push the emergency information to the airport operations command center. After the emergency is handled, traffic control at the crossing is restored.

[0030] When a no-passage order is issued, all green lights will be turned off, barriers will be lowered, red lights will be turned on, and an emergency warning will be broadcast.

[0031] Preferably, the topology building unit includes:

[0032] Category-based sub-units are used to divide the set of conflict factors for each potential conflict into subsets according to factor categories, and to determine the conflict gain of each subset;

[0033] The gain judgment subunit is used to treat several subsets with conflict gain greater than the preset gain as the first set, combine any factor in each first set, and use the maximum number of factors in all first sets as the benchmark, and input the resulting combinations into the factor analysis model in sequence to obtain the first conflict coordination label for each combination.

[0034] If there is no subset with conflict gain greater than the preset gain, the main factors of each subset are extracted to construct a combination of conflict vectors, which are then input into the factor analysis model to obtain the second conflict coordination label.

[0035] The feature extraction subunit is used to extract conflict features for each subset and construct a conflict-response vector by performing event response on the conflict features;

[0036] The vector parsing subunit is used to construct auxiliary vectors based on the conflict-response vector of each subset and all the labels involved in each subset. These vectors are then input into the vector event parsing model to obtain the conflict events that exist in the corresponding subset and the conflict change probability based on each conflict event.

[0037] The relationship determination subunit is used to determine immutable and mutable conflicts based on all conflict events and conflict change probabilities, and to obtain conflict relationships based on the event attributes of the conflict events involved in each immutable and mutable conflict.

[0038] The graph construction sub-unit is used to construct a conflict topology graph with the conflict relationship as the edge and the bifurcation point and the corresponding associated slide as the node.

[0039] Preferably, the feature extraction subunit includes:

[0040] The conflict feature acquisition block is used to collect multi-dimensional conflict features of M historical conflicts at airport flight area crossings.

[0041] The initial vector generation block is used to take the deviation parameters between the multi-dimensional conflict features of each conflict and the baseline conflict features as vector primitives, and generate the initial conflict vector after sorting the vector primitives corresponding to each conflict.

[0042] The statistics block is used to determine the frequency of each feature in all initial conflict vectors and to count the first feature corresponding to the maximum frequency.

[0043] The associated combination block is used to take each first feature as the parent feature and analyze the associated features of the initial conflict vector with the parent feature in turn to construct a comprehensive feature and place the blank vector.

[0044] The partitioning block is used to partition the initial conflict vector based on a preset response time window to obtain a preceding response sub-vector and a following response sub-vector. The dimension of the preceding response sub-vector is the left dimension after partitioning the composite feature with a blank vector according to the preset response time window, and the dimension of the following response sub-vector is the right dimension after partitioning the composite feature with a blank vector according to the preset response time window. If the feature in the corresponding dimension of the preceding or following response sub-vector is empty, the feature in the corresponding dimension that is closest to the corresponding initial conflict vector in the next historical time is locked and filled.

[0045] The compensation determination block is used to adjust the preset response time window based on the response time variation coefficient after M conflict processing and to further divide each divided combination vector to obtain intermediate sub-vectors and end sub-vectors. The feature expressions of the preceding response sub-vector, the following response sub-vector, the intermediate sub-vector, and the end sub-vector are solved respectively. The response compensation parameters of the conflict are calculated based on the feature weights of the conflict corresponding to each feature expression.

[0046] The correction block is used to perform similarity analysis on the conflict features of each subset with each initial conflict vector to filter the maximum similar vector, and obtain the required combination vector that is consistent with the maximum similar vector and the difference factor based on each element in the required combination vector. The corresponding difference factor in the required combination vector is then corrected according to the response compensation parameter to obtain the conflict-response vector.

[0047] Preferably, the associated combination block includes:

[0048] The feature group construction sub-block is used to select the first feature as the parent feature, filter out the initial conflict vectors containing the parent feature to form a set of parent feature vectors, and sequentially extract other features that are related to the parent feature in each parent feature vector to construct multiple related feature groups;

[0049] The quantity determination sub-block is used to determine the number of dimensions of the feature placement blank vector based on the distribution range of the parent feature in the initial conflict vector space. According to the occurrence time and correlation of features in each associated feature group, feature placeholders are set on the corresponding dimensions of the feature placement blank vector to generate the feature placement blank vector for subsequent conflict-response vector completion optimization.

[0050] Compared with the prior art, the beneficial effects of this application are as follows:

[0051] By deploying multiple types of devices at the sensor layer to comprehensively collect data on aircraft, authorized vehicles, the environment, and obstacles, the system ensures complete and accurate data dimensions. The data processing and decision-making layer eliminates data discrepancies through standardization, clarifies the spatiotemporal relationship between vehicles and aircraft through spatiotemporal mapping at bifurcation points, and employs a dynamic threshold adjustment mechanism to adapt to different scenarios. Multi-objective optimization decisions balance safety and efficiency, generating precise instructions. The execution layer coordinates the execution of instructions through multiple devices and provides real-time feedback, with closed-loop control ensuring timely correction of anomalies. This effectively reduces the risk of traffic conflicts at airport runway intersections, improves intersection efficiency, reduces vehicle waiting times and aircraft taxiing delays, and adapts to the refined scheduling needs of surface movement, providing reliable support for the safe and efficient operation of airport runway traffic.

[0052] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0053] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0054] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0055] Figure 1 This is a structural diagram of an airport flight area crossing traffic control system based on the A-SMGCS system, according to an embodiment of the present invention. Detailed Implementation

[0056] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0057] This invention provides an airport flight area intersection traffic control system based on the A-SMGCS system, such as... Figure 1 As shown, it includes:

[0058] The sensor layer is used to collect dynamic data of aircraft, data of objects awaiting authorization, environmental and meteorological data, and dynamic obstacle data.

[0059] The data processing and decision-making layer is used to standardize the collected data and perform spatiotemporal mapping processing according to the bifurcation of the target crossing in the flight area. It also uses a dynamic threshold adjustment mechanism to judge potential conflicts and generate and issue instructions based on multi-objective optimization decision-making. The bifurcation corresponds to a unique associated taxiway and is a lane bifurcation within the crossing that connects different target areas.

[0060] The execution layer is used to control the traffic control equipment at the intersection according to the issued instructions, and to feed back the execution results to the data processing and decision-making layer in real time to form a closed-loop control.

[0061] Preferably, the aircraft dynamic data includes: real-time coordinates, taxiing speed, locked taxiing trajectory, and predicted taxiing path for the next time period T, including the coordinate range of the associated taxiway;

[0062] The data of the object to be authorized to pass includes: identity identifier, target area, vehicle type and rated speed, and estimated time of entry into the crossing.

[0063] The environmental meteorological data includes: visibility;

[0064] The dynamic obstacle data includes: temporary construction areas and the location of disabled vehicles.

[0065] In this embodiment, A-SMGCS (Advanced Surface Movement Guidance and Control System) is an airport surface management system that integrates real-time monitoring, conflict warning, path planning, and automated guidance.

[0066] In this embodiment, the sensor layer is a hardware combination used to centrally collect various key data required for traffic control at airport flight area crossings. By deploying suitable sensing devices around and in related areas of the flight area crossings, real-time acquisition of data on aircraft, authorized ground vehicles, environment, and obstacles is achieved, providing a basic data source for subsequent data processing.

[0067] In this embodiment, aircraft dynamic data refers to a set of data reflecting the real-time status, predetermined route, and future taxiing trend of an aircraft during its taxiing on the ground in the flight area. Real-time coordinates are the real-time geographical coordinates of the aircraft during taxiing on the ground in the flight area, typically based on the airport's geographic information system. Taxiing speed is obtained as the ratio of the aircraft's displacement difference to the time interval during ground taxiing. The locked taxiway is the aircraft taxiing route pre-planned by the airport's A-SMGCS system and confirmed by air traffic controllers, clearly defining the taxiways and turning points the aircraft must traverse from its starting position to its target position. For example, a Boeing 737 taxiing from parking stand 3 to runway 36L has a locked taxiway as follows: Parking stand 3 → Taxiway B2 → Taxiway A3 → Taxiway A5 → Runway 36L, with the trajectory including the coordinates of each turning point.

[0068] The predicted taxiing path for the next time period (T) is based on the aircraft's current taxiing speed and locked taxi trajectory. It predicts the path the aircraft will take within a specific timeframe, including the taxiway segments involved and the estimated arrival time at each key node. Half of this specific timeframe is 5 to 10 minutes. Specifically, the sensor layer transmits the collected real-time speed and locked trajectory data to the local edge computing module. Using a path prediction algorithm based on uniform motion, the future path is calculated according to the set value for time period T. The prediction results are updated every 20 seconds to ensure they match the aircraft's actual taxiing status.

[0069] The coordinate range of the associated taxiway is the geographic coordinate boundary of the taxiways involved in the predicted taxiway path in the future time period T, which clarifies the specific spatial range associated with the taxiway and the taxiway branching point. Specifically, the coordinate boundary data of all taxiways are pre-stored through the airport geographic information system. After generating the predicted taxiway path, the edge computing module of the sensor layer calls the taxiway coordinate interface of the GIS system to match the taxiways involved in the predicted path, extract and output the coordinate range of the taxiway.

[0070] In this embodiment, the data for vehicles awaiting authorization refers to data related to ground vehicles that have obtained airport access permits and plan to pass through the airfield gates. It only includes data on legally verified vehicles to ensure compliance. Identification is a unique identifier used to distinguish vehicles awaiting authorization, ensuring vehicle traceability. This is achieved by installing RFID readers on pillars at the gate entrance and affixing dedicated RFID tags to authorized vehicles. The target area is the specific functional area within the airfield that the vehicle plans to visit, clearly defining its final destination. Vehicle type refers to the specific category to which the vehicle belongs. Different vehicle types have different driving characteristics, requiring targeted control strategies. Rated speed is the upper limit of the designed safe driving speed for the vehicle awaiting authorization, determined by the vehicle's owner based on factory parameters and airport safety regulations. For example, the rated speed for refueling trucks must be within the range of 8-12 km / h.

[0071] The estimated time of entry into the crossing is the estimated time when the vehicle waiting to be authorized to pass will arrive at the crossing entrance based on its current location and driving speed.

[0072] Dynamic obstacle data refers to data on objects or areas that affect traffic flow at airport flight area crossings and related areas, which needs to be updated in real time to avoid traffic conflicts.

[0073] Temporary construction areas are areas within the flight zone that are temporarily designated as prohibited or restricted areas for airport pavement maintenance, equipment installation, or other work. These areas have clearly defined construction periods and boundaries.

[0074] The location of a disabled vehicle is the real-time geographical location of an authorized vehicle that is unable to move normally due to mechanical failure, traffic accident, or other reasons at the entrance to the flight area and on related roads. It needs to be identified in a timely manner to avoid collisions with other vehicles or aircraft.

[0075] Standardization processing refers to converting data collected by the sensor layer from different sources, in different formats, and with different coordinate systems into a unified format and coordinate system that conforms to airport data specifications. Specifically, this involves deploying a data standardization program on the server at the data processing and decision-making layer. This program pre-configures the airport data specifications to use the WGS84 geographic coordinate system and the JSON format. It processes the real-time data transmitted from the sensor layer by calling the coordinate conversion interface and data format conversion functions of the airport GIS system.

[0076] In this embodiment, the target access point in the flight area refers to a specific airport flight area access point that currently requires traffic control. Each access point has a unique number, corresponds to a specific geographical location and associated taxiway, and is the specific object controlled by the system. It is based on the pre-stored basic information of all flight area access points, and the basic information includes the access point number, geographical coordinates, a list of associated taxiways, and the target area range served. For example, access point 3 in the airport flight area is located at the intersection of taxiway A5 and the maintenance area road, and mainly serves vehicles going to parking positions 2-5 and aircraft passing through taxiway A5.

[0077] In this embodiment, a fork in the road refers to a lane branching point within the target access point of the flight area that connects different target areas. Each fork corresponds to a unique associated taxiway, used to divert vehicles to different destinations. For example, there are two forks in access point 3. Fork 1 connects to parking positions 2 and 3 in the maintenance area, corresponding to associated taxiway A5. In this case, the route taken by a vehicle after passing through fork 1 is adjacent to the taxiing route of the aircraft on taxiway A5, and there is a possibility of spatiotemporal overlap. Fork 2 connects to parking positions 4 and 5 in the maintenance area, corresponding to associated taxiway A6. In this case, the route taken by a vehicle after passing through fork 2 is associated with the taxiing route of the aircraft on taxiway A6.

[0078] In this embodiment, it is determined whether the coordinate range of the bifurcation point overlaps with the coordinate range of the associated taxiway and whether the two time intervals overlap. A spatiotemporal mapping relationship table is constructed to realize spatiotemporal mapping processing. For example, a vehicle waiting to pass plans to pass through bifurcation point 1 of intersection 3, with an estimated passing time interval of 9:18-9:20. The coordinate range of bifurcation point 1 is 116.524°-116.525°E and 39.931°-39.932°N. An aircraft is expected to pass through the associated taxiway A5 corresponding to bifurcation point 1, with a passing time interval of 9:17-9:19. The coordinate range of associated taxiway A5 is 116.522°-116.528°E and 39.930°-39.935°N. After spatiotemporal mapping, it was found that the time intervals of the vehicle and the aircraft overlapped: 9:18-9:19. The spatial ranges also intersected: the coordinates of the bifurcation point 1 were within the coordinate range of the associated taxiway A5, providing an intuitive spatiotemporal correlation basis for subsequent conflict judgment.

[0079] In this embodiment, the conflict risk level and dynamic threshold adjustment mechanism are preset, and some of the content is shown in Table 1:

[0080] Table 1. Partial Contents of the Conflict Risk Level - Dynamic Threshold Adjustment Mechanism

[0081]

[0082] In this embodiment, the probability of spatiotemporal intersection = (duration of spatiotemporal intersection / duration of the vehicle's first time interval + area of ​​spatial intersection / area of ​​the bifurcation) / 2.

[0083] In this embodiment, multi-objective optimization decision refers to formulating the optimal traffic control strategy at the crossing by comprehensively considering the control objectives and combining them with the current traffic conditions at the airport when a potential conflict is determined to exist.

[0084] In this embodiment, the issued instructions are obtained based on the optimal control strategy determined by multi-objective optimization decision-making. For example, the optimal decision is to delay the refueling truck with the identification number JW-20240508 through intersection 1 of level 3 by 2 minutes. The generated instruction is intersection 1-vehicle JW-20240508-delay passage-2 minutes. The instruction format adopts the standardized code: F1-CJW20240508-YC-2, where F1 represents intersection 1. The instruction types include: allow passage, delay passage, route adjustment, prohibit passage, etc.

[0085] In this embodiment, the traffic control equipment at the level crossings is hardware deployed at level crossings and junctions within the flight zone to implement traffic guidance and passage control. It includes traffic lights, barriers, voice broadcasting devices, and vehicle-mounted terminal command push devices, serving as the execution layer for implementing control actions. Specifically: traffic lights are installed at the entrance of each junction; barriers are installed at the entrance of each junction; and voice broadcasting devices are installed on pillars on both sides of the level crossing. The terminal command push device refers to the device that pushes command content to the vehicle-mounted terminals of vehicles waiting to pass via wireless communication.

[0086] In this embodiment, the execution result refers to the status result of the traffic control equipment at the intersection actually performing control actions after receiving the issued instructions. This includes information such as whether the equipment acts according to the instructions, whether the actions are in place, and whether there are any malfunctions. It serves as the basis for data processing and decision-making to judge the effectiveness of instruction execution.

[0087] In this embodiment, closed-loop control refers to the complete control cycle of instruction issuance, execution, feedback, decision optimization, and new instruction issuance.

[0088] The beneficial effects of the above technical solution are as follows: By deploying multiple types of equipment at the sensor layer to comprehensively collect data on aircraft, authorized vehicles, the environment, and obstacles, data dimensions are ensured to be complete and accurate. At the data processing and decision-making layer, standardized processing eliminates data discrepancies, and spatiotemporal mapping at bifurcation points clarifies the spatiotemporal relationship between vehicles and aircraft. A dynamic threshold adjustment mechanism adapts to different scenarios, and multi-objective optimization decisions balance safety and efficiency, generating precise instructions. At the execution layer, multiple devices collaboratively execute instructions and provide real-time feedback, with closed-loop control ensuring timely correction of anomalies. This effectively reduces the risk of traffic conflicts at airport flight area intersections, improves intersection efficiency, reduces vehicle waiting times and aircraft taxiing delays, adapts to the refined scheduling needs of surface movement, and provides reliable support for the safe and efficient operation of airport flight area traffic.

[0089] This invention provides an airport flight area intersection traffic control system based on the A-SMGCS system, wherein the data processing and decision-making layer includes:

[0090] The standardization processing unit is used to standardize the collected data.

[0091] The time interval determination unit is used to determine the first time interval for the object to pass through the bifurcation and the second time interval for the aircraft to pass through the associated taxiway, based on aircraft dynamic data, data of the object to be authorized to pass and dynamic obstacle data.

[0092] The probability determination unit is used to determine the spatiotemporal intersection probability of each branch and the corresponding associated taxiway under the target crossing in the flight area based on the first time interval and the second time interval. At the same time, it dynamically sets the conflict threshold according to the conflict risk level, environmental meteorological data, traffic density and aircraft type.

[0093] The judgment unit is used to determine that there is no conflict at the corresponding bifurcation if the spatiotemporal intersection probability is less than the conflict threshold, and to generate a permission instruction.

[0094] Otherwise, determine the existence of potential conflicts;

[0095] Topology building units are used to extract the immutable and modifiable conflicts of the potential conflicts and construct a conflict topology graph.

[0096] An associated expansion unit is used to determine the maximum waiting time based on the immutable conflict, lock the immutable conflict at the first position in the conflict topology graph, and expand the associated node.

[0097] The decision generation unit is used to generate multi-objective optimization decisions based on rolling time-domain optimization, with the goal of maximizing the throughput of the A-SMGCS system in the future time.

[0098] Anchor point determination unit, used to determine conflict anchor points based on the maximum waiting time and the expanded nodes based on the multi-objective optimization decision;

[0099] The instruction matching unit is used to match the issued instruction that matches the conflicting anchor point from the anchor point-instruction type lookup table, wherein the issued instruction is a delayed passage instruction or a path adjustment instruction.

[0100] Preferably, when a passage permission instruction is issued, the corresponding green light at the fork in the road is turned on, the barrier is raised, a voice announcement is made, and a terminal instruction is pushed; when a passage delay instruction or a route adjustment instruction is issued, the yellow light is turned on, the delay duration is announced, or an alternative fork in the road is selected.

[0101] In this embodiment, the first time interval is the time range from the expected entry of the vehicle onto the taxiway to its passage through the target branch. For example, if a maintenance vehicle is expected to enter the taxiway at 9:05:00, and the distance from the taxiway to branch 1 is 250 meters, the passage time is approximately 250 meters ÷ 3.33 meters / second = 75 seconds, then the first time interval is 9:05:00 to 9:06:15. However, due to a temporary construction area, the vehicle's speed is reduced to 2.22 meters / second, extending the passage time to 250 meters ÷ 2.22 meters / second = 113 seconds. Therefore, the first time interval is adjusted to 9:05:00 to 9:06:53, including a 30-second error redundancy. The second time interval is the time range from the expected entry of the aircraft onto the associated taxiway to its complete passage through the associated taxiway. For example, the associated taxiway A5 of the Airbus A320 is 300 meters long, the taxiing speed is 5 meters / second, and the passage time is 300 meters ÷ 5 meters / second = 60 seconds. It is expected to enter the associated taxiway at 9:04:30. The second time interval is from 9:04:30 to 9:06:30, which includes a 60-second error redundancy.

[0102] In this embodiment, traffic density refers to the number of vehicles waiting to pass within a unit of time (unit: vehicles / 10 minutes), reflecting the degree of traffic congestion at the intersection. Aircraft types are classified according to wingspan and fuselage length into small, medium, and large aircraft.

[0103] The permission instruction includes the instruction type (permission allowed), the target bifurcation number (e.g., bifurcation 1), and the instruction effective time (e.g., immediate effective). For example, after determining that there is no conflict at bifurcation 2, a permission instruction is generated: Instruction type: Permission allowed; Target bifurcation: 2; Effective time: 9:07:00.

[0104] In this embodiment, the unchangeable conflict is a type of potential conflict that cannot be adjusted due to objective limitations. It is usually an aircraft-related conflict, such as an aircraft that has locked its taxiing trajectory, and adjustments would require replanning and affect the overall scheduling, or a large aircraft with large taxiing inertia that cannot temporarily change its path.

[0105] In this embodiment, a modifiable conflict is a type of potential conflict that can be eliminated by adjusting the parameters of the object to be passed, such as delaying the passage time or switching the driving path to other forks. It is usually a conflict related to ground vehicles, such as a conflict between an aircraft maintenance vehicle and an aircraft. By delaying the vehicle's passage by 5 minutes or switching to fork 2, the spatiotemporal overlap with the aircraft is eliminated, and this conflict is determined to be a modifiable conflict.

[0106] In this embodiment, the conflict topology graph is a visual graphical model constructed with bifurcations and associated slideways as nodes and conflict relationships as edges, which is used to intuitively display the conflict relationships between nodes.

[0107] In this embodiment, the maximum waiting time is the longest that an object waiting to pass can wait to avoid global scheduling delays caused by unchangeable conflicts. If this time is exceeded, the conflict must be resolved through other means, such as route adjustment. The maximum waiting time is set according to airport operation specifications and the urgency of the vehicle's task. For example, the maximum waiting time is 5 minutes for ordinary vehicles and 2 minutes for emergency vehicles. For instance, if an aircraft with an unchangeable conflict is scheduled to leave the associated taxiway at 9:10 AM, and the current time is 9:05 AM, the maximum waiting time for an ordinary maintenance vehicle is set to 5 minutes. This means that the vehicle must pass through the junction no later than 9:10 AM, or if it still cannot pass through after 9:10 AM, the route will be adjusted.

[0108] In this embodiment, the first position refers to marking the node and edge corresponding to the immutable conflict in the conflict topology graph as the core position. For example, the immutable conflict node between bifurcation 1 and A5 is placed in the center of the topology graph, the node size is 1.5 times that of other nodes, the edge is bold red, and it is marked as the first position.

[0109] In this embodiment, the expansion of associated nodes involves adding other nodes associated with the unchangeable conflict nodes in the conflict topology graph. For example, taking bifurcation 1-A5 as the core of the unchangeable conflict, the associated nodes bifurcation 2, bifurcation 3, and associated taxiway A6 are expanded. These nodes have path associations with the core node and serve as alternatives for vehicle path adjustment.

[0110] In this embodiment, the rolling time-domain optimization uses a fixed future time window as the optimization period, updates the traffic data within the window every minute, and re-optimizes the decision to ensure that the decision is adapted to the real-time traffic conditions.

[0111] In this embodiment, maximizing the throughput of the A-SMGCS system within the future timeframe refers to the core objective of multi-objective optimization. The throughput of the A-SMGCS system is the number of aircraft and objects waiting to pass through the airport surface per unit time. Maximizing throughput requires reducing waiting and delays and improving traffic efficiency while ensuring safety. For example, within the window of 9:05-9:10, the optimization objective is: ≥3 vehicles passing through the junction and ≥2 aircraft passing through the associated taxiway, ensuring maximum throughput.

[0112] In this embodiment, multi-objective optimization decision-making refers to generating the optimal control scheme while considering sub-objectives such as vehicle waiting time ≤ maximum waiting time, no aircraft taxiing delays, and minimum route adjustment costs, in addition to the core objective of maximizing throughput. Specifically, a rolling time-domain optimization program is deployed on the server, with a 5-minute optimization time window. Real-time data within the window is read every minute. The NSGA-II algorithm is used, with maximizing throughput as the core objective and waiting time ≤ maximum waiting time and no aircraft delays as constraints. The objective function value of different control schemes is calculated as: throughput score × 0.6 + waiting time score × 0.3 + adjustment cost score × 0.1. The scheme with the highest objective function value is selected as the multi-objective optimization decision.

[0113] In this embodiment, a conflict anchor point refers to a decision-identifier mapping point formed by binding core parameters of multi-objective optimization decisions, such as delay duration and alternative nodes, to fixed identifiers. Each anchor point corresponds to a unique control strategy. For example, if the optimization decision is: delay 2 minutes + bifurcation 1, combined with the maximum waiting time of 5 minutes and the expansion node bifurcation 2, the conflict anchor point is determined to be anchor point A2, and the parameters corresponding to this decision are: delay 2 minutes, alternative bifurcation 2. If the decision is to delay passage, the anchor point format is: anchor point + delay duration + target bifurcation, for example, anchor point A2 - 2 minutes - bifurcation 1; if the decision is to adjust the path, the anchor point format is: anchor point + alternative bifurcation + delay duration (if any), for example, anchor point B3 - bifurcation 2 - 0 minutes.

[0114] In this embodiment, the anchor-instruction type lookup table refers to a pre-defined database table that stores the correspondence between conflicting anchors and issued instructions. The table includes the anchor number, instruction type (delayed passage / path adjustment), and instruction parameters (delay duration, alternative branch points), ensuring that anchors can quickly match unique instructions. For example, the lookup table entries might be: Anchor A2: Instruction type - delayed passage, parameter - delay 2 minutes, target branch point 1; Anchor B3: Instruction type - path adjustment, parameter - alternative branch point 2, delay 0 minutes. The instruction format involved is: Instruction type - target branch point - parameter - effective time.

[0115] The beneficial effects of the above technical solution are as follows: by standardizing data benchmarks, the spatiotemporal intervals and conflict probabilities of objects to be passed and aircraft are accurately calculated, and dynamic thresholds are used to achieve accurate conflict determination; furthermore, by classifying conflicts, constructing topologies and expanding associations, optimization constraints are clarified, and decisions that take into account throughput and waiting time are generated based on rolling time-domain optimization. Finally, instructions are matched and linked with execution layer equipment to achieve visualized and personalized control, effectively reducing the conflict misjudgment rate and safety risks at airport flight area intersections, shortening vehicle waiting time, ensuring the efficiency of priority passage for aircraft, and providing reliable technical support for the safe, orderly and efficient operation of airport flight area traffic.

[0116] This invention provides an airport flight area crossing traffic control system based on the A-SMGCS system. The data processing and decision-making layer is also used to automatically trigger the emergency command mechanism when it receives a sudden deviation of an aircraft from its taxiway, a malfunction in the crossing area, or an emergency rescue request, generate a prohibition instruction, and push the emergency information to the airport operations command center. The crossing traffic control is restored after the emergency is handled.

[0117] When a no-passage order is issued, all green lights will be turned off, barriers will be lowered, red lights will be turned on, and an emergency warning will be broadcast.

[0118] In this embodiment, the triggering conditions of the emergency command mechanism are preset, such as: sudden deviation of the aircraft from the locked trajectory: deviation from the locked trajectory by ≥3 meters or deviation angle by ≥8°, lasting for more than 2 seconds; level crossing failure: the barrier control module feeds back a fault code, such as E01: motor failure, or the road surface sensor detects a foreign object, such as one weighing more than 50kg.

[0119] The beneficial effect of the above technical solution is that by timely triggering of emergency-related events, a prohibition order can be effectively generated, thus ensuring traffic safety.

[0120] This invention provides an airport flight area crossing traffic control system based on the A-SMGCS system, wherein the topology building unit includes:

[0121] Category-based sub-units are used to divide the set of conflict factors for each potential conflict into subsets according to factor categories, and to determine the conflict gain of each subset;

[0122] The gain judgment subunit is used to treat several subsets with conflict gain greater than the preset gain as the first set, combine any factor in each first set, and use the maximum number of factors in all first sets as the benchmark, and input the resulting combinations into the factor analysis model in sequence to obtain the first conflict coordination label for each combination.

[0123] If there is no subset with conflict gain greater than the preset gain, the main factors of each subset are extracted to construct a combination of conflict vectors, which are then input into the factor analysis model to obtain the second conflict coordination label.

[0124] The feature extraction subunit is used to extract conflict features for each subset and construct a conflict-response vector by performing event response on the conflict features;

[0125] The vector parsing subunit is used to construct auxiliary vectors based on the conflict-response vector of each subset and all the labels involved in each subset. These vectors are then input into the vector event parsing model to obtain the conflict events that exist in the corresponding subset and the conflict change probability based on each conflict event.

[0126] The relationship determination subunit is used to determine immutable and mutable conflicts based on all conflict events and conflict change probabilities, and to obtain conflict relationships based on the event attributes of the conflict events involved in each immutable and mutable conflict.

[0127] The graph construction sub-unit is used to construct a conflict topology graph with the conflict relationship as the edge and the bifurcation point and the corresponding associated slide as the node.

[0128] In this embodiment, the set of conflict factors for potential conflicts refers to the set of all factors that may lead to or affect the handling of conflicts, covering multiple dimensions such as aircraft, objects to be passed, environment, and traffic. For example, the set of conflict factors with a 77.3% spatiotemporal intersection probability between bifurcation 1 and associated taxiway A5 is as follows: {Aircraft type: medium-sized Airbus A320, aircraft taxiing speed: 18km / h, vehicle type: aircraft maintenance vehicle, vehicle rated speed: 12km / h, visibility: 300 meters, wind speed: 5m / s, number of vehicles to be passed: 5, temporary construction area 200 meters east of the crossing}.

[0129] In this embodiment, the factor category refers to the classification of conflict factors based on their attributes. In the context of airport crossings, it is usually divided into four categories: factors affecting aircraft taxiing, factors affecting vehicle passage, meteorological factors affecting passage safety, and temporary factors affecting the path.

[0130] In this embodiment, a subset refers to a set of factors under a single category obtained after splitting the conflict factor set according to factor categories. Each subset contains only factors of the same category. For example, the subsets after splitting the conflict factor set are: the subset affecting aircraft taxiing {Airbus A320, 18km / h}, the subset affecting vehicle passage {aircraft maintenance vehicle, 12km / h}, the subset affecting passage safety {300 meters, 5m / s}, and the subset affecting the path {5 vehicles, temporary construction}.

[0131] In this embodiment, the conflict gain refers to the weight of a single subset on the severity of a potential conflict. It is obtained by weighted summation of the risk coefficients of each factor within the subset, with a value ranging from 0 to 1. The risk coefficients are determined based on historical conflict data of the airport. For example, the risk coefficient for visibility of 300 meters is 0.8, and the risk coefficient for a medium-sized aircraft is 0.6. For example, the risk coefficients of factors affecting traffic safety are: visibility of 300 meters (0.8) and wind speed of 5 m / s (0.3), both with a weight of 0.5, and the conflict gain is: (0.8 × 0.5 + 0.3 × 0.5) = 0.55; the risk coefficients of factors affecting the path are: 5 vehicles waiting to pass (0.9) and temporary construction (0.7), both with a weight of 0.5, and the conflict gain is: (0.9 × 0.5 + 0.7 × 0.5) = 0.8.

[0132] In this embodiment, the preset gain refers to the critical value that a pre-set subset of judgments has a significant impact on the conflict. It is determined according to the airport safety operation standards and is usually set to 0.6. That is, a subset with a conflict gain > 0.6 is considered to have a significant contribution to the conflict. The preset value is fine-tuned by the administrator based on the quarterly conflict statistics results. For example, if the quarterly conflict rate is high, it is lowered to 0.5 to improve sensitivity.

[0133] In this embodiment, the first set refers to the set of all subsets whose conflict gain is greater than the preset gain. For example, if the conflict gain of only the subset affecting the path (0.8) is greater than 0.6, then the first set = {the subset affecting the path {5 vehicles waiting to pass, temporary construction}}; if there is also a subset affecting aircraft taxiing (gain 0.7), then the first set = {the subset affecting the path, the subset affecting aircraft taxiing}. For example, the first set contains the subset affecting the path, which contains 2 factors: number of vehicles, temporary construction, and the subset affecting aircraft taxiing, which contains 2 factors: aircraft type, taxiing speed. In this case, the maximum number of factors is 2, and the combination method is: extract 2 factors from the subset affecting the path (number of vehicles + temporary construction), and extract 2 factors from the subset affecting aircraft taxiing (aircraft type + taxiing speed), forming two types of combinations: number of vehicles + temporary construction, and aircraft type + taxiing speed; if the maximum number of factors is 3, then all combinations must contain 3 factors. If a subset only has 2 factors, 1 factor can be extracted repeatedly to complete it.

[0134] In this embodiment, the factor analysis model refers to an algorithm model used to determine the direction of conflict coordination based on the combination of factors. It is trained based on historical conflict handling data of the airport. The input is a combination of factors, such as 5 vehicles waiting to pass + temporary construction. The output is the object or parameter label that needs to be adjusted corresponding to the combination. A random forest model is used. The training data contains 1000 sets of factor combinations - historical coordination schemes, such as 5 vehicles + construction → adjust vehicle path. The model learns from historical data and outputs the coordination direction corresponding to the combination.

[0135] In this embodiment, the first conflict coordination label refers to the conflict coordination suggestion label output by the factor analysis model after analyzing the factor combination of the first set. The label content clearly defines the actionable adjustment target. For example, if the combination of 5 vehicles waiting to pass + temporary construction is input into the factor analysis model, the model outputs the first conflict coordination label as: some vehicles need to be guided to switch to other intersections to avoid the construction area; if the combination of Airbus A320 + 18km / h is input, the output label is: the aircraft taxiing speed is stable and no adjustment is needed.

[0136] In this embodiment, the main factors refer to the 1-2 factors with the highest risk coefficients selected from each subset when there is no subset with a conflict gain greater than the preset gain, i.e., all subsets have a low contribution to the conflict. For example, if the conflict gain of all subsets is less than 0.6, such as the subset affecting traffic safety having a gain of 0.55 and the subset affecting the path having a gain of 0.58, then the main factors selected from the subset affecting traffic safety are: visibility 300 meters, with a risk coefficient of 0.8 > wind speed 0.3; and the main factors selected from the subset affecting the path are: 5 vehicles waiting to pass, with a risk coefficient of 0.9 > temporary construction 0.7.

[0137] In this embodiment, the conflict vector combination refers to arranging the main factors of each subset in the order of factor categories, such as aircraft-related → objects to be passed → environmental weather → traffic obstacles, to form a quantitative vector. Non-numerical factors need to be converted into numerical codes, such as aircraft type: small aircraft = 1, medium aircraft = 2, large aircraft = 3; temporary construction: present = 1, absent = 0. For example, if the main factors of each subset are: aircraft type: Airbus A320 (2), vehicle rated speed: 12km / h, visibility: 300 meters, number of vehicles to be passed: 5, then the conflict vector combination is [2, 12, 300, 5].

[0138] In this embodiment, the second conflict coordination label refers to the comprehensive coordination suggestion label output by the factor analysis model after analyzing the combination of conflict vectors, targeting the main factors of all subsets. The difference between the second label and the first label is that the first label only targets the core subset, while the second label covers the main factors of all subsets, emphasizing overall coordination. For example, inputting the conflict vector combination [2,12,300,5] into the model will output the second conflict coordination label as: The vehicle speed needs to be reduced to 10km / h, and the vehicle waiting time should be shortened to within 3 minutes.

[0139] In this embodiment, conflict features refer to the specific indicators extracted from each subset that reflect the impact of that subset on conflict. The feature dimensions need to be determined in conjunction with the subset category: the features of the subset affecting aircraft taxiing are: aircraft type code and taxiing speed (km / h); the features of the subset affecting vehicle passage are vehicle type code and rated speed (km / h); the features of the subset affecting traffic safety are visibility (meters) and wind speed (m / s); and the features of the subset affecting the path are the number of vehicles waiting to pass (vehicles / 10 minutes) and the distance to the construction area (meters, set to 0 if there is no construction). For example, the conflict features of the subset affecting aircraft taxiing are [2 (medium-sized aircraft), 18], and the conflict features affecting traffic safety are [300, 5].

[0140] In this embodiment, event response refers to the control actions and effect data taken by the airport in the past when dealing with conflicts with similar characteristics to the current subset of conflicts. This includes response action types, such as delayed passage and route adjustment; response parameters, such as delay duration and alternative branching point numbers; and conflict resolution rate after response, such as a conflict resolution rate of 90% after a 5-minute delay. These are used to construct the correlation between features and responses.

[0141] In this embodiment, the conflict-response vector refers to a quantized vector formed by combining the conflict features of each subset with the corresponding event response parameters in a fixed order. The vector dimension is fixed, and non-numerical parameters need to be encoded, such as response action types: delayed passage = 1, path adjustment = 2. For example, if the conflict features of the subset affecting the path are [5 (number of vehicles), 200 (construction distance)], and the corresponding event responses are: path adjustment (2), alternative fork 2 (2), conflict elimination rate 85%, then the conflict-response vector of this subset is [5, 200, 2, 2, 85].

[0142] In this embodiment, each subset involves all tags, referring to the conflict coordination tags corresponding to that subset in the gain judgment subunit. If the subset belongs to the first set, it is the first conflict coordination tag; if it does not belong, it is the second conflict coordination tag, as well as the adjustment priority implied in the tag. For example, if the tag includes priority adjustment, the priority is 1; if it includes ordinary adjustment, the priority is 2. The tag needs to be converted into a numerical code. For example, the code for the tag that adjusts the vehicle path is 3, and the priority is 1. For instance, if the subset affecting the path belongs to the first set, and the corresponding first conflict coordination tag is guiding some vehicles to switch to the branch point, its code is 3, and the priority is 1, then all tags involved in this subset are coded as [3,1].

[0143] In this embodiment, the auxiliary vector refers to the extended vector formed by concatenating the conflict-response vector of each subset with all the label codes involved in that subset in the order of conflict-response vector first and label code last. It is used to supplement the coordination direction information of the vector and improve the accuracy of model parsing. For example, if the conflict-response vector of the subset affecting the path is [5,200,2,2,85] and the label code is [3,1], then the auxiliary vector is [5,200,2,2,85,3,1]; if the conflict-response vector of the subset affecting the aircraft taxiing is [2,18,0,0,0] and the label code is [4 (no adjustment required), 3 (low priority)], then the auxiliary vector is [2,18,0,0,0,4,3].

[0144] In this embodiment, the vector event parsing model refers to an algorithmic model used to identify conflict event types and quantify the difficulty of conflict modification based on auxiliary vectors. It is trained using a deep learning framework, with fixed-dimensional auxiliary vectors as input and output including conflict event names, such as the probability of a vehicle conflicting with an aircraft due to construction and the probability of such a conflict modification. The model training data consists of auxiliary vectors with manually labeled event types and modification probabilities, totaling over 1500 samples. For example, by learning vector features such as a construction distance of 200 meters and 5 vehicles, the model can output the probability of a construction-impact vehicle-aircraft conflict and its modification.

[0145] In this embodiment, the conflict event pointer event analysis model outputs a precise description of the underlying cause of the current potential conflict. This description must include the objects involved in the conflict, such as vehicles and aircraft, and the core triggers, such as construction and low visibility, to facilitate subsequent judgment on whether the conflict can be altered. For example, after analyzing the auxiliary vectors of a subset of the influencing paths, the output conflict event could be: "Due to construction on the east side of the crossing, the maintenance vehicle is queuing up, causing a conflict with an Airbus A320 on taxiway A5."

[0146] In this embodiment, the conflict change probability refers to the quantitative value of the possibility of eliminating the conflict event by adjusting the control strategy, output by the quantitative event analysis model. The value ranges from 0 to 1: when 0.6 ≤ probability ≤ 1, the conflict is easily eliminated by adjusting the vehicle / route (can be changed); when 0 ≤ probability < 0.4, the conflict is difficult to adjust due to aircraft / core environmental factors (cannot be changed); when 0.4 ≤ probability < 0.6, it needs to be judged in conjunction with other events. For example, the change probability of the above-mentioned construction-affected vehicle-aircraft conflict is 0.8, which means that the construction can be avoided by adjusting the vehicle route, and it is easy to eliminate; the change probability of the extended spatiotemporal intersection caused by low visibility is 0.3, which means that visibility cannot be manually adjusted, and it is difficult to eliminate.

[0147] In this embodiment, all conflict events and conflict change probability pointers are output by the corresponding subset of potential conflicts and conflict events and change probabilities, and are summarized to form an event-probability list. For example, the summary list of a potential conflict is: {1. Construction-affected vehicle-aircraft conflict (0.8), 2. Low visibility leads to extended spatiotemporal intersection (0.3), 3. Medium-sized aircraft taxiing trajectory stability (0.2), 4. Vehicle rated speed adapts to the crossing (0.5)}.

[0148] In this embodiment, an unchangeable conflict refers to a type of conflict that cannot be eliminated by manually adjusting the control strategy, based on the change probability of all conflict events. The determination rule is: there is at least one conflict event with a change probability < 0.4, and this event is directly related to the aircraft, such as aircraft type, locked trajectory, taxiing speed, etc., or related to uncontrollable environmental factors, such as low visibility, strong winds, etc. This type of conflict requires priority to protect the aircraft or wait for the environment to improve. For example, in the summary list above, low visibility leads to an extension of the spatiotemporal intersection (0.3) and the taxiing trajectory of a medium-sized aircraft is stable (0.2), both of which are < 0.4, and involve the aircraft and the uncontrollable environment. Therefore, this potential conflict is determined to be an unchangeable conflict.

[0149] In this embodiment, a modifiable conflict refers to a conflict type where the probability of modification for all conflict events is ≥0.6, and the events are only related to vehicles or controllable obstacles, such as the number of vehicles or construction areas. Such conflicts can be eliminated by adjusting vehicle routes, delay times, or other strategies. For example, a potential conflict might have the following events: five vehicles queuing leading to a conflict with an aircraft (0.9) and a temporary construction area occupying a lane (0.8). Both events have a probability of ≥0.6 and only involve vehicles and construction, thus qualifying as modifiable conflicts.

[0150] In this embodiment, event attributes refer to the inherent attributes of the conflict event itself, used to describe the scope of the conflict's impact, such as affecting only a single bifurcation, affecting multiple bifurcations, and the degree of urgency, such as ordinary conflict and emergency conflict. The attribute values ​​are extracted from the conflict event description. For example, the scope of the construction impact type conflict = single bifurcation, and the degree of urgency = ordinary; the scope of the aircraft deviation from the trajectory conflict = multiple bifurcations, and the degree of urgency = emergency.

[0151] In this embodiment, the conflict relationship refers to the type of conflict between a bifurcation and its corresponding associated taxiway, determined based on event attributes. It includes three dimensions: conflict type (unchangeable / changeable), scope of impact, and urgency, used to clarify the association between nodes in the topology graph. For example, the conflict relationship corresponding to an unchangeable conflict is: Bifurcation 1 - Associated Taxiway A5: Unchangeable conflict, scope of impact = single bifurcation, urgency = normal; if it is a changeable conflict, the relationship is: Bifurcation 2 - Associated Taxiway A6: Changeable conflict, scope of impact = single bifurcation, urgency = normal.

[0152] In this embodiment, conflict relationships are defined as edges in the topology graph, where the conflict relationships output by the relationship-determining sub-unit are used as edges connecting nodes. The attributes of the edges correspond to the conflict relationships: edge color distinguishes conflict types (red for unchangeable conflicts, yellow for changeable conflicts, and green for no conflicts); edge thickness distinguishes urgency (thin lines for ordinary conflicts and thick lines for emergency conflicts); and edge labels indicate the scope of influence, such as single-port influence. For example, if the conflict relationship between bifurcation 1 and associated taxiway A5 is unchangeable, of ordinary urgency, and single-port influence, then the corresponding edge is a thin red line with the label "single-port influence."

[0153] In this embodiment, the bifurcation and its corresponding associated taxiway are considered nodes in the topology diagram. The bifurcations of the target access points in the flight area, such as bifurcation 1 and bifurcation 2, and their corresponding associated taxiways, such as A5 and A6, are treated as independent nodes. The shape and color of the nodes distinguish their types: bifurcations are represented by circles in blue, and associated taxiways by rectangles in gray. Nodes are labeled with their names, such as bifurcation 1 and taxiway A5, ensuring that the node types are clearly identifiable. For example, the nodes for access point 3 include: circular blue nodes: bifurcation 1 and bifurcation 2; and rectangular gray nodes: taxiways A5 and taxiways A6.

[0154] In this embodiment, the conflict topology map refers to a visual graphic model constructed with bifurcations and associated taxiways as nodes and conflict relationships as edges. It is used to intuitively display the conflict relationships between nodes within the target crossing. For example, in the conflict topology map of crossing No. 3, the blue circular bifurcation 1 is connected to the gray rectangular taxiway A5 by a red thin line (edge), and the edge label indicates single-crossing impact; the blue circular bifurcation 2 is connected to the gray rectangular taxiway A6 by a yellow thin line, and the edge label indicates single-crossing impact; there is no direct edge between bifurcation 1 and bifurcation 2 (no conflict). The overall graphic is displayed with the airport GIS map background overlaid, and the node positions are consistent with the actual crossing layout.

[0155] The beneficial effects of the above technical solution are: by breaking down conflict factors, locking down core influencing factors and outputting coordination labels, constructing conflict-response correlation vectors, identifying conflict events and difficulty of modification, determining conflict types and clarifying correlation attributes, and generating a visual topology map, it can achieve a refined breakdown and visual presentation of potential conflicts at airport flight area crossings, accurately distinguish between unchangeable and changeable conflicts, effectively reduce the conflict misjudgment rate, and improve the scientificity and safety of traffic control at crossings.

[0156] This invention provides an airport flight area crossing traffic control system based on the A-SMGCS system, wherein the feature extraction subunit includes:

[0157] The conflict feature acquisition block is used to collect multi-dimensional conflict features of M historical conflicts at airport flight area crossings.

[0158] The initial vector generation block is used to take the deviation parameters between the multi-dimensional conflict features of each conflict and the baseline conflict features as vector primitives, and generate the initial conflict vector after sorting the vector primitives corresponding to each conflict.

[0159] The statistics block is used to determine the frequency of each feature in all initial conflict vectors and to count the first feature corresponding to the maximum frequency.

[0160] The associated combination block is used to take each first feature as the parent feature and analyze the associated features of the initial conflict vector with the parent feature in turn to construct a comprehensive feature and place the blank vector.

[0161] The partitioning block is used to partition the initial conflict vector based on a preset response time window to obtain a preceding response sub-vector and a following response sub-vector. The dimension of the preceding response sub-vector is the left dimension after partitioning the composite feature with a blank vector according to the preset response time window, and the dimension of the following response sub-vector is the right dimension after partitioning the composite feature with a blank vector according to the preset response time window. If the feature in the corresponding dimension of the preceding or following response sub-vector is empty, the feature in the corresponding dimension that is closest to the corresponding initial conflict vector in the next historical time is locked and filled.

[0162] The compensation determination block is used to adjust the preset response time window based on the response time variation coefficient after M conflict processing and to further divide each divided combination vector to obtain intermediate sub-vectors and end sub-vectors. The feature expressions of the preceding response sub-vector, the following response sub-vector, the intermediate sub-vector, and the end sub-vector are solved respectively. The response compensation parameters of the conflict are calculated based on the feature weights of the conflict corresponding to each feature expression.

[0163] The correction block is used to perform similarity analysis on the conflict features of each subset with each initial conflict vector to filter the maximum similar vector, and obtain the required combination vector that is consistent with the maximum similar vector and the difference factor based on each element in the required combination vector. The corresponding difference factor in the required combination vector is then corrected according to the response compensation parameter to obtain the conflict-response vector.

[0164] In this embodiment, M historical conflicts refer to the total number of conflict events that have been resolved and are similar to the current potential conflict scenario, selected from the historical operation records of airport flight area gates. Typically, M ≥ 300 times to ensure coverage of different weather, traffic, and aircraft type scenarios. In addition, 500 conflicts that occurred at the No. 3 gate of the airport in the past 2 years are selected as M historical conflicts, covering different scenarios such as rainy days, foggy days, and morning and evening rush hours.

[0165] In this embodiment, the multi-dimensional conflict features refer to the features extracted from each historical conflict that cover four core dimensions: aircraft, objects to be passed, environment, and obstacles. Each dimension includes 2-3 key indicators: the aircraft dimension includes aircraft type (small / medium / large) and taxiing speed (km / h); the objects to be passed dimension includes vehicle type (maintenance vehicle / refueling vehicle / shuttle vehicle) and number of vehicles to be passed; the environment dimension includes visibility (meters) and wind speed (m / s); and the obstacle dimension includes whether there is temporary construction (yes / no) and the distance of the disabled vehicle from the crossing (meters).

[0166] In this embodiment, the baseline conflict feature refers to the feature value of a typical conflict scenario calculated based on the multi-dimensional features of M historical conflicts. It serves as a reference benchmark for measuring the deviation of the features of each historical conflict. The calculation rules are as follows: the classification feature takes the value with the highest frequency, such as the aircraft type taking a medium-sized aircraft; the quantitative feature takes the mean, such as taxiing speed taking 16 km / h and visibility taking 400 meters. For example, the baseline conflict feature is: aircraft dimension (medium-sized aircraft, 16 km / h), object to be passed dimension (maintenance vehicle, 3 vehicles), environmental dimension (400 meters, 4 m / s), and obstacle dimension (yes, 200 meters).

[0167] In this embodiment, the deviation parameter refers to the difference between the multi-dimensional features of each historical conflict and the baseline conflict features. The classification features are converted into numerical values ​​through encoding to calculate the deviation. For example, the aircraft type encoding is: small aircraft = 1, medium aircraft = 2, large aircraft = 3. If the conflicting aircraft type is a medium aircraft (2), the deviation from the baseline (2) is 0; if it is a large aircraft (3), the deviation is 1. The quantitative features directly calculate the difference. For example, if the taxiing speed of a conflict is 18km / h, the deviation from the baseline 16km / h is +2km / h, etc.

[0168] In this embodiment, a vector primitive refers to the deviation parameter corresponding to a single dimension feature. It is the basic unit that makes up the initial conflict vector. Each vector primitive corresponds one-to-one with a dimension and feature. For example, the aircraft type deviation of 0, the taxiing speed deviation of +2, and the visibility deviation of -100 are all independent vector primitives.

[0169] In this embodiment, the occurrence count refers to the frequency of occurrence of different deviation parameter values ​​in each dimension (a total of 8 dimensions, corresponding to 8 vector primitive positions) after traversing the initial conflict vector of M historical conflicts. For example, if the visibility deviation dimension is at the 5th primitive position, a deviation parameter of -100 corresponds to a visibility of 300 meters occurring 120 times, a deviation parameter of -50 corresponds to a visibility of 350 meters occurring 90 times, and so on.

[0170] The first feature corresponding to the maximum frequency refers to the original conflict feature corresponding to the deviation parameter that appears most frequently in each dimension. That is, it is the actual feature corresponding to the deviation parameter derived in reverse, and it is the most representative core feature in that dimension. For example, in the visibility deviation dimension, -100 appears most frequently, that is, 120 times, so the first feature of this dimension is: visibility 300 meters; in the vehicle quantity deviation dimension, +1 corresponds to 4 vehicles in the 4th primitive, which appears most frequently, 110 times, so the first feature is: 4 vehicles waiting to pass.

[0171] In this embodiment, the parent feature refers to the first feature corresponding to the maximum occurrence frequency of each dimension determined by the statistical block. An initial conflict vector with a parent feature refers to an initial conflict vector that contains at least one vector with a deviation parameter corresponding to the parent feature in a certain dimension, that is, the deviation parameter of the vector in a certain dimension is consistent with the deviation parameter of the parent feature.

[0172] In this embodiment, the "comprehensive feature placement blank vector" refers to an empty vector frame with the parent feature as the core, reserving the corresponding dimension positions of the associated features. The number of vector dimensions is equal to the number of parent feature dimensions plus the number of associated feature dimensions. Each dimension position is labeled with the corresponding feature type, such as parent feature - visibility, associated feature - number of vehicles, for later filling in specific feature values. For example, with visibility of 300 meters (1 parent feature dimension) as the core, and associated features being the number of vehicles (4) and temporary construction (2 associated dimensions), the dimension frame of the comprehensive feature placement blank vector is [parent feature - visibility, associated feature - number of vehicles, associated feature - construction status], with each dimension temporarily blank (marked with NULL).

[0173] In this embodiment, the preset response time window refers to the time range from the occurrence of a conflict to the first control response, which is used to correspond to the time attributes of the conflict characteristics, such as real-time characteristics within the window and feedback characteristics after the window. The window duration is set based on the average response time of M historical conflicts, which is usually 20-40 seconds. For example, if the historical average response time is 25 seconds, the preset window is 30 seconds.

[0174] In this embodiment, the eight dimensions of the initial conflict vector are divided into two types of sub-vectors based on whether they correspond to the pre-defined response time window: the preceding response sub-vector corresponds to the real-time handling features within the window, such as aircraft taxiing speed, current visibility, and number of vehicles, totaling five dimensions; the following response sub-vector corresponds to the feedback optimization features after the window, such as conflict resolution time, equipment execution status, and secondary conflict risk, totaling three dimensions. The sum of the dimensions of the sub-vectors after division is equal to the number of dimensions of the initial vector. For example, the initial conflict vector is [0,+2,0,+1,-100,+1,0,-50], the preceding response sub-vector consists of the first five dimensions [0,+2,0,+1,-100], which correspond to the aircraft, vehicles, and real-time visibility features within the window; the following response sub-vector consists of the last three dimensions [+1,0,-50], which correspond to the wind speed feedback, construction status, and faulty vehicle distance optimization after the window.

[0175] In this embodiment, the dimensions of the composite feature placement blank vector are divided into left and right parts according to the response time window. For example, if the composite feature placement blank vector is [visibility (within the window), number of vehicles (within the window), construction status (after the window)] (2 dimensions on the left and 1 dimension on the right), then the preceding response sub-vector must have 2 dimensions and the following response sub-vector must have 1 dimension. If the preceding sub-vector of the initial vector has 5 dimensions, the first 2 dimensions need to be truncated to match the left dimension of the blank vector.

[0176] In this embodiment, in some historical conflict pre- and post-conflict sub-vectors, a certain dimension may be empty due to data collection delays or missing data, such as the secondary conflict risk dimension of the post-conflict sub-vector having no record. In this case, the feature value of that dimension is extracted from the next historical conflict that immediately follows the previous conflict in time and used to fill the missing value. The most recent history filling method is used to handle missing values, that is, the corresponding dimension feature of the next conflict is queried by sorting by time using SQL, and the filled data is stored in the partitioned sub-vector table.

[0177] In this embodiment, the coefficient of variation of response time after M conflict resolution refers to the dispersion of response time from the occurrence of the conflict to its complete resolution in M ​​historical conflicts. The calculation formula is coefficient of variation = standard deviation of response time / mean of response time. It is used to judge the stability of response time: the smaller the coefficient of variation (usually <0.3), the more stable the time; the larger the coefficient of variation (>0.3), the greater the fluctuation of the time, and the more the response time window needs to be adjusted. For example, the mean response time of M=500 conflicts is 28 seconds, the standard deviation is 9.5 seconds, and the coefficient of variation = 9.5 / 28 = 0.34 (>0.3), indicating that the response time fluctuates greatly.

[0178] In this embodiment, adjusting the preset response time window refers to adjusting the preset window duration according to the magnitude of the coefficient of variation. The adjustment rule is as follows: if the coefficient of variation is greater than 0.3, the window duration is shortened by 10%-15%; if the coefficient of variation is less than 0.2, the window duration is extended by 5%-10%. For example, if the original preset window is 30 seconds and the coefficient of variation is 0.34, the adjusted window duration is 30×(1-10%)=27 seconds.

[0179] In this embodiment, the preceding response subvector and the following response subvector are further divided into intermediate subvectors and final subvectors. For example, the adjusted preceding subvector has 4 dimensions, corresponding to the features within a 27-second window, and is further divided into intermediate subvectors with the first 3 dimensions, corresponding to the real-time features within the first 20 seconds of the window, and final subvectors with the last 1 dimension, corresponding to the adjusted features within the last 7 seconds of the window; the following subvector has 3 dimensions, and is further divided into intermediate subvectors with the first 2 dimensions, corresponding to the feedback features within the first 15 seconds after the window, and final subvectors with the last 1 dimension, corresponding to the optimized features after 15 seconds after the window.

[0180] In this embodiment, the feature expression refers to the mathematical expression corresponding to each sub-vector, which is used to quantify the overall feature contribution of the sub-vector. The expression form is a linear combination, that is, based on the deviation parameters and feature weights of each dimension of the feature. For example, the feature expression of the preceding intermediate sub-vector is: F1=w1×x1+w2×x2+w3×x3, where x1, x2, and x3 are the normalized results of the deviation parameters of the three dimensions of the sub-vector, and w1, w2, and w3 are the feature weights of each dimension.

[0181] In this embodiment, the feature weight refers to the influence coefficient of each dimension feature on the conflict handling effect. The sum of their weights is 1, which is determined by the analytic hierarchy process. For example, the taxiing speed deviation weight of the aircraft dimension is 0.4, the visibility deviation weight of the environment dimension is 0.3, and the quantity deviation weight of the vehicle dimension is 0.3.

[0182] In this embodiment, the response compensation parameter refers to the compensation value calculated based on the feature expressions and feature weights of each sub-vector, used to correct the conflict feature deviation, where the response compensation parameter = .

[0183] In this embodiment, the conflict features of each subset refer to the actual multi-dimensional features of a subset of potential conflicts to be analyzed.

[0184] In this embodiment, similarity analysis refers to calculating the feature similarity between the conflict features of each subset and all initial conflict vectors, and is implemented using the cosine similarity algorithm.

[0185] In this embodiment, the required combined vector refers to a vector frame that is completely consistent with the preorder / postorder / middle / end sub-vector structure of the maximum similar vector. That is, each sub-vector of the maximum similar vector is extracted and combined to form a base vector for subsequent correction. For example, the combined vector of the maximum similar vector is [preorder middle sub-vector, preorder end sub-vector, postorder middle sub-vector, postorder end sub-vector]=[[0,+2,-100],[+1],[0,-50],[0]].

[0186] In this embodiment, the difference factor refers to the difference between the conflict features of the current subset and the corresponding dimension features of the required combination vector, which is used to quantify the degree of difference between the two. For example, if the wind speed deviation in the required combination vector is +1 (corresponding to a wind speed of 5m / s), and the wind speed deviation of the current subset is +2 (corresponding to a wind speed of 6m / s), the difference factor = +2 - (+1) = +1; if the required combination vector for the vehicle number deviation is +1 (4 vehicles), and the current subset is +2 (5 vehicles), the difference factor = +1.

[0187] In this embodiment, the conflict-response vector refers to the response compensation parameter calculated using the compensation determination block, which adjusts the value of the difference factor. The correction rule is: corrected difference factor = original difference factor × (1 + response compensation parameter). The corrected difference factor is integrated with other parameters of the required combined vector to form the final conflict-response vector, which contains complete information on conflict feature deviation, response compensation, and corrected difference. For example, if the original difference factor is +1 and the response compensation parameter of the corresponding sub-vector is -0.05, the corrected difference factor = +1 × (1 - 0.05) = +0.95. Substituting all the corrected difference factors into the required combined vector, the conflict-response vector is obtained as [[0, +2, -100], [+0.95], [0, -50], [0]].

[0188] The beneficial effects of the above technical solution are: through historical data collection, feature quantification, core screening, correlation analysis, structural optimization, dynamic compensation, and precise correction, it effectively improves the completeness, accuracy, and timeliness of airport flight area intersection conflict characteristics, providing high-quality data support.

[0189] This invention provides an airport flight area crossing traffic control system based on the A-SMGCS system, wherein the associated combination block includes:

[0190] The feature group construction sub-block is used to select the first feature as the parent feature, filter out the initial conflict vectors containing the parent feature to form a set of parent feature vectors, and sequentially extract other features that are related to the parent feature in each parent feature vector to construct multiple related feature groups;

[0191] The quantity determination sub-block is used to determine the number of dimensions of the feature placement blank vector based on the distribution range of the parent feature in the initial conflict vector space. According to the occurrence time and correlation of features in each associated feature group, feature placeholders are set on the corresponding dimensions of the feature placement blank vector to generate the feature placement blank vector for subsequent conflict-response vector completion optimization.

[0192] In this embodiment, "other features" refers to non-parent feature dimensions that frequently co-occur with the parent feature in the parent feature vector set. The core criterion for determining the association relationship is the co-occurrence frequency, which is the proportion of times this feature also appears when the parent feature appears, out of the total number of times the parent feature appears. This usually needs to be ≥70%. For example, in the parent feature vector set (120 vectors) for visibility of 300 meters, statistics show that there are 95 vectors with the 4th dimension (vehicle quantity deviation) being +1 (corresponding to 4 vehicles waiting to pass), with a co-occurrence frequency of 95 / 120=79.2%. There are 88 vectors with the 7th dimension (whether there is construction deviation) being 0 (corresponding to temporary construction), with a co-occurrence frequency of 88 / 120=73.3%. Both meet the co-occurrence requirement of ≥70%. Therefore, 4 vehicles waiting to pass and temporary construction are considered other features that are associated with the parent feature of visibility of 300 meters.

[0193] In this embodiment, multiple associated feature groups are multiple feature sets constructed by combining the parent feature with other features of the above-mentioned association relationship according to the combination logic of 1 parent feature + 1-2 associated features. The features in each set are strongly associated with the parent feature, and the associated features of different sets are not repeated.

[0194] In this embodiment, the distribution range is the coverage area of ​​all initial conflict vectors corresponding to the parent feature in a multidimensional space composed of all dimensions of the initial conflict vectors. The initial conflict vector space is an 8-dimensional space, which corresponds to 8 deviation parameter dimensions. Each vector is a point in the space, and the distribution range is the set of value intervals of these points in each dimension, reflecting the spatial representativeness of the parent feature in historical conflicts.

[0195] In this embodiment, the number of dimensions of the blank vector for feature placement is based on the spatial distribution range of the parent feature. It is the total number of dimensions of the empty vector used for subsequent placement of the parent feature and related features. The core is to ensure that the blank vector can completely cover the dimensions corresponding to the parent feature and all strongly related features, and does not include irrelevant dimensions. For example, if the spatial distribution range of the parent feature visibility of 300 meters covers 3 core dimensions: dimension 5: visibility, dimension 4: number of vehicles, and dimension 7: whether construction is underway, and all 3 dimensions have strongly related features, then the number of dimensions of the blank vector for feature placement is set to 3; if the spatial distribution range of the parent feature covers 4 core dimensions, then the number of dimensions is set to 4 to ensure that the blank vector can accommodate all key features.

[0196] In this embodiment, the occurrence sequence refers to the chronological order of the parent feature and related features within the associated feature group during the occurrence and development of historical conflicts. This sequence reflects the natural occurrence logic of features in the conflict scenario and serves as the basis for subsequently determining the order of blank vector dimensions.

[0197] In this embodiment, the degree of association is the frequency of co-occurrence between the parent feature and each associated feature within the associated feature group. For example, in the aforementioned associated feature group, the co-occurrence frequency of the parent feature "visibility 300 meters" and the associated feature "4 vehicles waiting to pass" is 95 times (accounting for 79.2% of the total occurrences of the parent feature), while the co-occurrence frequency with "temporary construction" is 88 times (73.3%). Therefore, the degree of association between the parent feature "4 vehicles waiting to pass" and the associated feature "temporary construction" is higher than that between the parent feature and the associated feature "temporary construction".

[0198] In this embodiment, the feature placeholder is the feature type, association priority and corresponding dimension source information marked on each dimension of the feature placement blank vector. The feature placement blank vector refers to the final generated empty vector with feature placeholder. If a feature of a certain dimension of the sub-vector is empty, the feature type and source to be filled can be accurately located based on the placeholder of the blank vector.

[0199] The beneficial effects of the above technical solution are as follows: by using the first feature as the parent feature, filtering the set of parent feature vectors and extracting related features to construct related feature groups, the intrinsic relationship between core features and other features in historical conflicts can be accurately explored, avoiding interference from irrelevant features in subsequent analysis; the number of blank vector dimensions is determined based on the spatial distribution range of the parent features, and placeholders are set in combination with the feature time sequence and the degree of correlation, generating structured feature placement blank vectors, providing clear guidance for the completion of missing features in subsequent conflict-response vectors.

[0200] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An airport flight zone crossing traffic control system based on an A-SMGCS system, characterized in that, The application relates to an A-SMGCS system, which comprises the following parts: a sensor layer for collecting aircraft dynamic data, data of objects to be passed, environmental meteorological data and dynamic obstacle data; a data processing and decision-making layer for standardizing the collected data and processing the data according to the branch of the target intersection of the flight area, judging the potential conflict existing in combination with a dynamic threshold adjustment mechanism, and generating and issuing an instruction based on multi-objective optimization decision-making, wherein the branch corresponds to a unique associated taxiway, and the branch is a lane branch in the intersection connecting different target areas; an execution layer for controlling intersection traffic control equipment according to the issued instruction and feeding back the execution result to the data processing and decision-making layer in real time to form a closed-loop control; wherein the data processing and decision-making layer comprises the following parts: a standardization processing unit for standardizing the collected data; a time interval determination unit for determining a first time interval of the object to be passed through the branch and a second time interval of the aircraft through the associated taxiway based on the aircraft dynamic data, the data of the object to be passed and the dynamic obstacle data; a probability determination unit for determining the space-time intersection probability of each branch and the corresponding associated taxiway of the target intersection of the flight area based on the first time interval and the second time interval, and dynamically setting a conflict threshold according to the conflict risk level, the environmental meteorological data, the traffic density and the aircraft type; a judgment unit for judging that the corresponding branch has no conflict if the space-time intersection probability is less than the conflict threshold, and generating a passing permission instruction; otherwise, judging that a potential conflict exists; a topology construction unit for extracting the unchangeable conflict and the changeable conflict of the potential conflict and constructing a conflict topology graph; an association expansion unit for determining the maximum waiting time length based on the unchangeable conflict, locking the first position of the unchangeable conflict in the conflict topology graph and expanding the associated nodes; a decision-making generation unit for generating a multi-objective optimization decision based on the rolling horizon optimization and taking the maximum A-SMGCS system throughput in the future time as the target; an anchor point determination unit for determining the conflict anchor point based on the maximum waiting time length and the expanded nodes based on the multi-objective optimization decision; and an instruction matching unit for matching the issued instruction matched with the conflict anchor point from an anchor point-instruction type table, wherein the issued instruction is a delay passing instruction or a path adjustment instruction. The aircraft dynamic data comprises real-time coordinates, taxiway speed, locked taxiway trajectory and predicted taxiway path in a future T period, and the coordinates of the associated taxiway are included; the data of the object to be passed comprises an identity, a target area, a vehicle type, a rated speed and a predicted entering time of the intersection; the environmental meteorological data comprises visibility; and the dynamic obstacle data comprises temporary construction areas and fault vehicle positions. When the passing permission instruction is issued, the corresponding branch is opened, the green light is turned on, the barrier is lifted, the voice is broadcast and the terminal instruction is pushed; when the delay passing instruction or the path adjustment instruction is issued, the yellow light is lit, the delay time length or the alternative branch is broadcast. ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ 2. The A-SMGCS system based airport movement area intersection traffic command system according to claim 1, characterized in that, ​ ​ ​ ​ 3. The A-SMGCS system based airport movement area intersection traffic command system according to claim 1, characterized in that, ​ 4. The A-SMGCS system based airport movement area intersection traffic command system according to claim 1, characterized in that, The data processing and decision layer is further configured to automatically trigger an emergency command mechanism, generate a no-passing instruction, and push emergency information to an airport operation command center when receiving an aircraft sudden deviation from a taxiway trajectory, a failure in a crossing area, or receiving an emergency rescue request, and restore the crossing traffic command after the emergency event is disposed. When the no-passing instruction is issued, all green lights are turned off, the barriers are lowered, the red lights are turned on, and the emergency prompt is broadcasted.

5. The A-SMGCS system based airport movement area intersection traffic command system according to claim 1, characterized in that, The topology construction unit comprises: The category division subunit is configured to divide each set of conflict factors of each potential conflict according to the factor categories to obtain subsets, and determine the conflict gain of each subset; The gain judgment subunit is configured to, if there are several subsets with a conflict gain greater than a preset gain, regard the several subsets as first sets, combine any factor in each first set, take the maximum number of factors in all first sets as a benchmark, and input the obtained combinations into the factor analysis model in sequence to obtain a first conflict coordination label of each combination; If there are no subsets with a conflict gain greater than a preset gain, extract the main factors of each subset to construct a conflict vector combination, and input the conflict vector combination into the factor analysis model to obtain a second conflict coordination label; The feature extraction subunit is configured to extract conflict features of each subset, and construct a conflict-response vector based on event responses of the conflict features; The vector analysis subunit is configured to construct an auxiliary vector based on the conflict-response vector of each subset and all labels involved in each subset, input the auxiliary vector into a vector event analysis model, and obtain a conflict event existing in the corresponding subset and a conflict change probability based on each conflict event; The relationship determination subunit is configured to determine unchangeable conflicts and changeable conflicts based on all conflict events and conflict change probabilities, and obtain conflict relationships according to event attributes of conflict events involved in each unchangeable conflict and changeable conflict; The graph construction subunit is configured to construct a conflict topology graph by taking the conflict relationships as edges and taking branch points and corresponding associated taxiways as nodes.

6. The A-SMGCS system based airport movement area intersection traffic command system according to claim 5, characterized in that, The feature extraction subunit comprises: A conflict feature collection block is configured to collect multi-dimensional conflict features of M historical conflicts of a crossing of an airport flight zone; An initial vector generation block is configured to take a deviation parameter of multi-dimensional conflict features of each conflict from a benchmark conflict feature as a vector element, sort the vector element corresponding to each conflict, and generate an initial conflict vector; A statistics block is configured to determine the number of occurrences of each dimensional feature in all initial conflict vectors, and count a first feature corresponding to a maximum number of occurrences; An association combination block is configured to take each first feature as a parent feature, analyze associated features of the initial conflict vector in which the parent feature exists in sequence, and construct a comprehensive feature to place a blank vector. ​ The division block is configured to divide the initial conflict vector based on a preset response time window to obtain a preceding response sub-vector and a subsequent response sub-vector, wherein a dimension number of the preceding response sub-vector is a left dimension number after the comprehensive feature placement blank vector is divided according to the preset response time window, a dimension number of the subsequent response sub-vector is a right dimension number after the comprehensive feature placement blank vector is divided according to the preset response time window, and if a feature in a corresponding dimension of the preceding response sub-vector or the subsequent response sub-vector is empty, a feature in the corresponding dimension existing at a nearest historical time from the initial conflict vector is locked and filled; The compensation determination block is configured to adjust the preset response time window based on a response time variation coefficient after M times of conflict processing, and divide each divided combined vector again to obtain an intermediate sub-vector and a tail sub-vector, solve feature expressions of the preceding response sub-vector, the subsequent response sub-vector, the intermediate sub-vector and the tail sub-vector respectively, calculate a response compensation parameter of the conflict according to a feature weight of the conflict corresponding to each feature expression, and obtain a conflict-response vector by modifying a corresponding difference factor in the required combined vector according to the response compensation parameter. The correction block is configured to perform similarity analysis on each conflict feature of each subset and each initial conflict vector to select a maximum similar vector, obtain a required combined vector consistent with the maximum similar vector, and obtain a difference factor based on each element in the required combined vector, and modify a corresponding difference factor in the required combined vector according to the response compensation parameter to obtain a conflict-response vector.

7. An A-SMGCS system based airport movement area intersection traffic command system according to claim 6, characterized in that, The association combination block comprises: The feature group construction sub-block is configured to take the first feature as a parent feature, filter initial conflict vectors containing the parent feature to form a parent feature vector set, sequentially extract other features having an association relationship with the parent feature in each parent feature vector, and construct a plurality of associated feature groups. The number determination sub-block is configured to determine a dimension number of a feature placement blank vector based on a distribution range of the parent feature in an initial conflict vector space, set a feature placeholder in a corresponding dimension of the feature placement blank vector according to a time sequence of the features in each associated feature group and an association closeness, and generate the feature placement blank vector for subsequent conflict-response vector completion optimization.

Images