A digital parallel simulation airport aircraft taxiing simulation scheduling system and method

Abstract

The present invention discloses a digital parallel simulation airport aircraft taxiing simulation scheduling system and method. The system comprises an airport aircraft taxiing physical model, an airport aircraft taxiing parallel simulation model, a data-driven airport situation awareness model, and an autonomous learning-based aircraft taxiing scheduling strategy generation model. According to the accurately predicted aircraft entry and departure times and the scene operation situation, the aircraft push-out time and taxiing path plan are continuously and dynamically adjusted to realize the digital parallel simulation technology combining virtual and real, thereby assisting the safe and efficient level of the airport scene operation intelligence.

Description

Technical Field

[0001] The present invention relates to the field of intelligence, and in particular to a digital parallel simulation airport aircraft taxiing simulation scheduling system and method. Background Art

[0002] As the number of flights at airports continues to increase, the pressure on green, safe and efficient surface operations is increasing. Faced with increasingly busy airports, how to improve the efficiency of airport surface resource operations and reduce carbon emissions has become a major issue in civil aviation ground transportation research. There are many factors that affect aircraft taxiing. On the one hand, the calculation scale is large, and it is difficult to calculate a feasible solution in a short time by manual methods; on the other hand, considering multiple objectives and constraints, the calculated solution is not optimal. Furthermore, the various solutions under different objectives and constraints vary greatly, and a multi-dimensional perspective is needed to visualize and evaluate their advantages and disadvantages. With the help of advanced technologies such as the Internet of Things, big data and digital twins, it is urgent to study the intelligent daily operation management of airport aircraft taxiing scheduling, actively perceive environmental changes to identify its operating bottlenecks, and quickly generate a green scheduling plan suitable for environmental changes, reduce aircraft taxiing time, avoid conflicts, and reduce aircraft ground congestion and carbon emissions by improving the efficiency of airport aircraft taxiing operations.

[0003] Through studying and combing a large number of domestic and foreign literature, the existing theoretical research and practice of intelligent taxiing scheduling have two main deficiencies: on the one hand, some studies have established mathematical models of airport taxiing paths in static environments and constructed solution algorithms from the perspective of operations optimization, which is also difficult to apply to taxiing scheduling problems in dynamically changing environments; on the other hand, some studies have simulated the behaviors of captains and controllers from the perspective of simulation optimization to generate taxiing paths, but the deficiency is that the simulation environment is relatively rough and cannot approach the real world. Therefore, a digital parallel simulation of airport aircraft taxiing virtual-real combined scheduling can solve the airport aircraft taxiing scheduling problem, and existing research is relatively rare.

[0004] Therefore, the present invention develops a digital parallel simulation airport aircraft taxiing simulation scheduling system and method, an airport aircraft taxiing physical model, an airport aircraft taxiing parallel simulation model, a data-driven airport situation awareness model, and an aircraft taxiing scheduling strategy generation model based on autonomous learning. Based on intelligent network technology, the physical world and the virtual world are combined, and the aircraft taxiing process is reproduced using real data, scenes, plans and knowledge, and the impact of various traffic events on the scheduling plan is simulated and evaluated, thereby assisting the safe and efficient level of airport field operation intelligence. Summary of the invention

[0005] The purpose of the present invention is to provide a digital parallel simulation airport aircraft taxiing simulation scheduling system and method to solve the problems raised in the background technology.

[0006] The present application provides a digital parallel simulation airport aircraft taxiing simulation scheduling method, the method comprising the following steps:

[0007] Step 1: The airport aircraft taxiing physical model tracks the status of surface operation micro-macro indicators in real time based on intelligent network connection technology, integrates the status of surface operation micro-macro indicators, monitors the operation status of the taxiing network, warns of abnormal information and diagnoses the cause, wherein the surface operation micro-macro indicator status includes the operation track of the aircraft, the operation status of the runway and the meteorological environment data, etc.

[0008] Step 2: The airport aircraft taxiing parallel simulation model builds a virtual reality environment of the airport aircraft taxiing physical model based on high-precision GIS, including parking positions, taxiway networks, and runway gate configurations, etc., accesses real-time field operation multi-source heterogeneous data based on intelligent network technology, and visualizes the historical and current aircraft taxiing process and the actual operation status of the runway;

[0009] Step 3: Based on the data-driven airport situation awareness model, the operation data of the airport aircraft taxiing parallel simulation model is accessed, and the large model technology is used to dynamically evaluate the current scene operation situation and predict the evolution trend of the future scene operation situation, thereby detecting the real traffic events or upcoming events of the airport aircraft taxiing physical model;

[0010] Step 4: The aircraft taxiing scheduling strategy generation model based on autonomous learning evaluates the deficiencies of different traffic events on the current aircraft pushback time and taxiing path plan based on the real-time and future traffic event detection results of the data-driven airport situation awareness model, and dynamically formulates a feasible adjustment plan.

[0011] Furthermore, the airport aircraft taxiing physical model tracks the state of micro- and macro-indicators of the surface operation in real time based on the intelligent network connection technology, integrates the state of the micro- and macro-indicators of the surface operation, monitors the operation state of the taxiing network, warns of abnormal information and diagnoses the causes, including the following steps:

[0012] Step 1.1: Monitor the status of micro-indicators of field operation, including:

[0013] Monitor the flight trajectory of aircraft based on video technology, GPS, and ADS-B, including the aircraft's location and time, speed, energy consumption, and plan execution deviation;

[0014] Monitor the status of whether the aircraft has left the parking position based on RFID;

[0015] Monitor the operation status of the runway based on video technology and RFID, including runway occupancy and idle time;

[0016] Monitoring of abnormal events based on video and radar technology, including the time, location, type, degree and duration of abnormal events;

[0017] Monitoring of meteorological environment data based on meteorological radar sensing technology, involving the time, location, type, degree and duration of extreme weather events;

[0018] Step 1.2: Monitor the status of the macro indicators of the field operation, including:

[0019] Monitor the operation status of each taxiway network node and section based on video technology, including queue length, density, waiting time, etc.

[0020] Monitor the operation status of the taxiway network based on video technology, including the number of taxiing aircraft, number of stops, etc.;

[0021] Step 1.3: Integrate the status of surface operation micro- and macro-indicators to monitor the operation status of the taxiing network. When an indicator is abnormal, warn of abnormal information and diagnose the cause based on the causal correlation between them, such as the impact of weather and intrusion events on aircraft taxiing.

[0022] Furthermore, the airport aircraft taxiing parallel simulation model constructs a virtual reality environment of the airport aircraft taxiing physical model based on high-precision GIS, including parking positions, taxiway networks and runway gate configurations, etc., accesses real-time scene operation multi-source heterogeneous data based on intelligent network technology, and visualizes the historical and current aircraft taxiing process and the real operation status of the runway. The steps include:

[0023] Building a digital virtual reality environment based on high-precision GIS, including a GIS for the physical model of the sliding network;

[0024] In the taxiing network physical model GIS, add different types of equipment layers GIS, including video, RFID, etc., and interact with intelligent network technology to access various micro-macro-indicator status data of the field operation;

[0025] Use multi-source data fusion technology to accurately calibrate aircraft trajectory, weather information, runway status, abnormal events and other data;

[0026] Automatically record the multi-input and multi-output causal relationship label time series data of various micro-macro-indicator states of the scene operation;

[0027] The aircraft's trajectory is associated with weather information, runway status, and abnormal events to automatically form a multi-input and multi-output causal relationship label data;

[0028] Using the taxiing network physical model GIS as the visualization base, the current taxiing process of the aircraft, the current operating status of the runway, and the weather conditions are dynamically tracked at a 1:1 scale;

[0029] Using the taxiing network physical model GIS as the visualization base, the historical taxiing process of aircraft, the current and historical operating status of the runway, and the weather status are reproduced in a 1:1 scale multi-dimensional visualization.

[0030] Furthermore, the data-driven airport situation awareness model accesses the operation data of the airport aircraft taxiing parallel simulation model, uses the large model technology to dynamically evaluate the current scene operation situation, and predicts the evolution trend of the future scene operation situation. The steps of detecting the real traffic events or upcoming events of the airport aircraft taxiing physical model include:

[0031] Step 3.1: According to the situational awareness targets and objects of different airports, a library of multiple data-driven deep learning algorithms is constructed based on time series or causal analysis, and each deep learning algorithm is associated with the target, object and time range one by one;

[0032] Step 3.2: Based on the multi-input-multi-output causal relationship label historical data of various micro-macro indicator states of the field operation, different output labels are selected according to different airport situation awareness goals, and the training samples and test data sets are determined using different airport situation awareness objects and time ranges;

[0033] Step 3.3: Select the best deep learning algorithm from the deep learning algorithm library according to the situation awareness objectives of different airports, run various training samples and test data sets on the field, and obtain a series of association rules, that is, the one-to-one correspondence between each input indicator range and different operation situations / trends / events;

[0034] Step 3.4: According to the current scene operation status, retrieve the possible association rule set that matches the input conditions and evaluate its probability of occurrence, and output the results in descending order of probability.

[0035] Furthermore, the autonomous learning-based aircraft taxiing scheduling strategy generation model evaluates the inadequacies of different traffic events on the current aircraft pushback time and taxiing path plan based on the real-time and future traffic event detection results of the data-driven airport situation awareness model, and the steps of dynamically developing a feasible adjustment plan include:

[0036] Step 4.1: Data preparation, including: (1) aircraft positions, runway gates and paths; (2) real-time and future traffic event detection results based on the data-driven airport situational awareness model, such as planned pushback time, runway incursion and runway occupation time; (3) taxi network data; (4) operation rules, etc.;

[0037] Step 4.2: Construct autonomous learning-based aircraft taxiing scheduling strategy generation model scenarios for different user scenarios, including:

[0038] Step 4.2.1: User scenarios include, for example, separation of incoming and outgoing flights; mixed incoming and outgoing flights, etc.

[0039] Step 4.2.2: Scheduling objectives, such as carbon emissions, taxiing distance, and waiting time;

[0040] Step 4.2.3: Autonomous learning objects, such as taxi time, push-off time, runway incursion;

[0041] Step 4.2.4: Constraints, such as taxi path conflicts, maximum taxi waiting time, maximum push-out delay time, path network flow constraints, etc.

[0042] Step 4.2.5: Decision variables, such as taxi path, aircraft pushback time, and aircraft arrival and departure times through taxi nodes;

[0043] Step 4.3: According to different scenarios, use PyEPO to build an aircraft taxiing scheduling strategy generation model engine based on autonomous learning, integrating data, knowledge, models and algorithms to generate aircraft pushback time and taxiing path plans;

[0044] Step 4.3.1: Use the deep learning module of PyEPO to autonomously predict the evolution of taxi time, pushback time, and runway incursion;

[0045] Step 4.3.2: Use the operations research optimization solver module of PyEPO to establish a distributed robust strategy generation model for aircraft taxiing scheduling, and bring the prediction results of step 4.3.1 into the solver for solution;

[0046] Step 4.3.3: If the deviation of the prediction optimization result is less than the threshold, output the result; otherwise, repeat steps 4.3.1 and 4.3.2 continuously;

[0047] Step 4.4: Import the generated aircraft pushback time and taxiing path plan into the airport aircraft taxiing parallel simulation model based on virtual-real integration constructed in step 2, evaluate the impact of different traffic events on it in multiple dimensions, and generate constraints for illegal operation rules. If the constraints exist, return to step 4.2 and update the constraints of step 4.2.4; otherwise, return to step 4.5;

[0048] Step 4.5: Compare the historical and generated aircraft pushback times and taxi path plans, and dynamically formulate a feasible adjustment plan accordingly.

[0049] Another object of the present invention is to provide a digital parallel simulation airport aircraft taxiing simulation scheduling system, the system includes an airport aircraft taxiing physical model, an airport aircraft taxiing parallel simulation model, an execution object (aircraft, etc.), an intelligent network sensor, an intelligent computing platform and a server; wherein the airport aircraft taxiing physical model is an entity, the airport aircraft taxiing parallel simulation model is installed on the server, and has a 1:1 real correspondence with the airport aircraft taxiing physical model; the intelligent network sensor is installed on the execution object and the airport aircraft taxiing physical model, and tracks the state of the micro-macro indicators of the field operation in real time, and stores the data on the server; the intelligent computing platform is installed on the server, and is mainly executed by a data-driven airport situation awareness model and an autonomous learning-based aircraft taxiing scheduling strategy generation model;

[0050] Airport aircraft taxiing physical model, which tracks the status of surface operation micro-macro indicators in real time based on intelligent network connection technology, integrates the status of surface operation micro-macro indicators, monitors the operation status of the taxiing network, warns of abnormal information and diagnoses the cause, wherein the surface operation micro-macro indicators include the operation trajectory of the aircraft, the operation status of the runway and the meteorological environment data;

[0051] The airport aircraft taxiing parallel simulation model builds a virtual reality environment of the airport aircraft taxiing physical model based on high-precision GIS, including parking positions, taxiway networks, and runway gate configurations, etc. It accesses real-time field operation multi-source heterogeneous data based on intelligent network technology, and visualizes the historical and current aircraft taxiing process and the real operation status of the runway;

[0052] The data-driven airport situation awareness model is connected to the operation data of the airport aircraft taxiing parallel simulation model. It uses large model technology to dynamically evaluate the current scene operation situation and predict the evolution trend of the future scene operation situation. Based on this, it detects real traffic events or upcoming events in the airport aircraft taxiing physical model.

[0053] An aircraft taxiing scheduling strategy generation model based on autonomous learning evaluates the deficiencies of different traffic events on the current aircraft pushback time and taxiing path plan in multiple dimensions according to the real-time and future traffic event detection results of the data-driven airport situation awareness model, and dynamically proposes a feasible adjustment plan based on this.

[0054] Beneficial effects achieved by the present invention:

[0055] The present invention is based on the digital parallel simulation technology combining virtuality and reality, and uses high-precision GIS to construct a virtual reality environment of the airport aircraft taxiing physical model. Based on the intelligent network technology, the aircraft's operation trajectory, runway operation status, meteorological environment data and other field operation micro-macro indicators are tracked in real time, and the current and historical taxiing process of the aircraft and the current and historical operation status of the runway are reproduced 1:1. Based on this, the aircraft taxiing scheduling strategy generation model construction process based on autonomous learning is carried out. According to the detected real-time and future traffic events, the shortcomings of the current aircraft push-out time and taxiing path plan are evaluated, and a feasible adjustment plan is dynamically proposed, thereby assisting the safety and efficiency level of the airport field operation intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The present invention is a flowchart of a digital parallel simulation airport aircraft taxiing simulation scheduling method.

[0057] Figure 2 The present invention is a schematic diagram of the structure of a digital parallel simulation airport aircraft taxiing simulation scheduling system.

[0058] Figure 3 This is a schematic diagram of the physical model of airport aircraft taxiing based on intelligent networking technology proposed in the present invention.

[0059] Figure 4 This is a schematic diagram of the airport aircraft taxiing parallel simulation model based on the combination of virtual and real proposed by the present invention.

[0060] Figure 5 This is a schematic diagram of the data-driven airport situation awareness model proposed in the present invention.

[0061] Figure 6 This is a schematic diagram of the aircraft taxiing scheduling strategy generation model based on autonomous learning proposed in the present invention. DETAILED DESCRIPTION

[0062] The technical solution of the present invention is described in detail below in conjunction with the specific drawings.

[0063] like Figure 1 As shown, the present embodiment provides a digital parallel simulation airport aircraft taxiing simulation scheduling method, comprising the following steps:

[0064] Step 1: The airport aircraft taxiing physical model tracks the status of surface operation micro-macro indicators in real time based on intelligent network connection technology (integrating video surveillance, RFID and various sensor technologies), integrates the status of surface operation micro-macro indicators, monitors the operation status of the taxiing network, warns of abnormal information and diagnoses the causes, wherein the surface operation micro-macro indicators include the operation trajectory of the aircraft, the operation status of the runway and the meteorological environment data, etc.

[0065] Step 2: The airport aircraft taxiing parallel simulation model builds a virtual reality environment of the airport aircraft taxiing physical model based on high-precision GIS, including parking positions, taxiway networks, and runway gate configurations, etc., accesses real-time field operation multi-source heterogeneous data based on intelligent network technology, and visualizes the historical and current aircraft taxiing process and the actual operation status of the runway;

[0066] Step 3: Based on the data-driven airport situation awareness model, the operation data of the airport aircraft taxiing parallel simulation model is accessed, and the large model technology is used to dynamically evaluate the current scene operation situation and predict the evolution trend of the future scene operation situation, thereby detecting the real traffic events or upcoming events of the airport aircraft taxiing physical model;

[0067] Step 4: The aircraft taxiing scheduling strategy generation model based on autonomous learning evaluates the deficiencies of different traffic events on the current aircraft pushback time and taxiing path plan based on the real-time and future traffic event detection results of the data-driven airport situation awareness model, and dynamically formulates a feasible adjustment plan.

[0068] like Figure 3 As shown, the airport aircraft taxiing physical model tracks the status of micro- and macro-indicators of the surface operation in real time based on intelligent network connection technology (integrating video surveillance, RFID and various sensor technologies), integrates the status of micro- and macro-indicators of the surface operation, monitors the operation status of the taxiing network, warns of abnormal information and diagnoses the causes, including the following steps:

[0069] Step 1.1: Obtain the aircraft taxiway network layout, including aircraft stands, runway gates, intermediate taxiway nodes, and taxiway sections.

[0070] Step 1.2: Install different intelligent network sensors at nodes on the taxiway network, including aircraft stands, runway entrances, intermediate taxiway nodes, etc., and associate the device IP address with the actual longitude and latitude of the seat.

[0071] Step 1.3: Monitor the status of micro-indicators of the field operation, including:

[0072] Monitor the flight trajectory of aircraft based on video technology, GPS, and ADS-B, including the aircraft's location and time, speed, energy consumption, and plan execution deviation;

[0073] Monitor the status of whether the aircraft has left the parking position based on RFID;

[0074] Monitor the operation status of the runway based on video technology and RFID, including runway occupancy and idle time;

[0075] Monitoring of abnormal events based on video and radar technology, including the time, location, type, degree and duration of abnormal events;

[0076] Monitoring of meteorological environment data based on meteorological radar sensing technology, involving the time, location, type, degree and duration of extreme weather events;

[0077] Step 1.4: Monitor the status of the macro indicators of the field operation, including:

[0078] Monitor the operation status of each taxiway network node and section based on video technology, including queue length, density, waiting time, etc.

[0079] Monitor the operation status of the taxiway network based on video technology, including the number of taxiing aircraft, number of stops, etc.;

[0080] Step 1.5: Integrate the status of surface operation micro- and macro-indicators to monitor the operation status of the taxiing network. When an indicator is abnormal, warn of abnormal information and diagnose the cause based on the causal correlation between them, such as the impact of weather and intrusion events on aircraft taxiing.

[0081] like Figure 4 As shown, the airport aircraft taxiing parallel simulation model constructs a virtual reality environment of the airport aircraft taxiing physical model based on high-precision GIS, including parking positions, taxiway networks and runway gate configurations, etc., accesses real-time scene operation multi-source heterogeneous data based on intelligent network technology, and visualizes the historical and current aircraft taxiing process and the actual operation status of the runway. The steps include:

[0082] Building a digital virtual reality environment based on high-precision GIS, including a GIS for the physical model of the sliding network;

[0083] In the taxiing network physical model GIS, add different types of equipment layers GIS, including video, RFID, etc., and interact with intelligent network technology to access various micro-macro-indicator status data of the field operation;

[0084] Use multi-source data fusion technology to accurately calibrate aircraft trajectory, weather information, runway status, abnormal events and other data;

[0085] Automatically record the multi-input and multi-output causal relationship label time series data of various micro-macro-indicator states of the scene operation;

[0086] The aircraft's trajectory is associated with weather information, runway status, and abnormal events to automatically form a multi-input and multi-output causal relationship label data;

[0087] Using the taxiing network physical model GIS as the visualization base, the current taxiing process of the aircraft, the current operating status of the runway, and the weather conditions are dynamically tracked at a 1:1 scale;

[0088] Using the taxiing network physical model GIS as the visualization base, the historical taxiing process of aircraft, the current and historical operating status of the runway, and the weather status are reproduced in a 1:1 scale multi-dimensional visualization.

[0089] like Figure 5 As shown, the data-driven airport situation awareness model accesses the operation data of the airport aircraft taxiing parallel simulation model, uses the large model technology to dynamically evaluate the current scene operation situation, and predicts the evolution trend of the future scene operation situation. The steps of detecting the real traffic events or upcoming events of the airport aircraft taxiing physical model include:

[0090] Step 3.1: Based on the situation awareness objectives (including operational situation assessment, trend prediction, event detection), objects (including pushback time, taxi time, taxiway network status, runway incursion, etc.) and time range (including long time, short time, etc.) of different airports, a library of multiple data-driven deep learning algorithms is constructed based on time series or causal analysis, and each deep learning algorithm is associated with the objectives, objects and time range one by one;

[0091] Step 3.2: Based on the multi-input-multi-output causal relationship label historical data of various micro-macro indicator states of the field operation, different output labels are selected according to different airport situation awareness goals, and the training samples and test data sets are determined using different airport situation awareness objects and time ranges;

[0092] Step 3.3: Select the best deep learning algorithm from the deep learning algorithm library according to the situation awareness objectives of different airports, run various training samples and test data sets on the field, and obtain a series of association rules, that is, the one-to-one correspondence between each input indicator range and different operation situations / trends / events;

[0093] Step 3.4: According to the current scene operation status, retrieve the possible association rule set that matches the input conditions and evaluate its probability of occurrence, and output the results in descending order of probability.

[0094] like Figure 6 As shown, the autonomous learning-based aircraft taxiing scheduling strategy generation model evaluates the inadequacies of different traffic events on the current aircraft push-out time and taxiing path plan based on the real-time and future traffic event detection results of the data-driven airport situation awareness model, and the steps of dynamically developing a feasible adjustment plan include:

[0095] Step 4.1: Data preparation, including: (1) aircraft positions, runway gates and paths; (2) real-time and future traffic event detection results based on the data-driven airport situational awareness model, such as planned pushback time, runway incursion and runway occupation time; (3) taxi network data; (4) operation rules, etc.;

[0096] Step 4.2: Construct autonomous learning-based aircraft taxiing scheduling strategy generation model scenarios for different user scenarios, including:

[0097] Step 4.2.1: User scenarios include, for example, separation of incoming and outgoing flights; mixed incoming and outgoing flights, etc.

[0098] Step 4.2.2: Scheduling objectives, such as carbon emissions, taxiing distance, and waiting time;

[0099] Step 4.2.3: Autonomous learning objects, such as taxi time, push-off time, runway incursion;

[0100] Step 4.2.4: Constraints, such as taxi path conflicts, maximum taxi waiting time, maximum push-out delay time, path network flow constraints, etc.

[0101] Step 4.2.5: Decision variables, such as taxi path, aircraft pushback time, and aircraft arrival and departure times through taxi nodes;

[0102] Step 4.3: According to different scenarios, use PyEPO to build an aircraft taxiing scheduling strategy generation model engine based on autonomous learning, integrating data, knowledge, models and algorithms to generate aircraft pushback time and taxiing path plans;

[0103] Step 4.3.1: Use the deep learning module of PyEPO to autonomously predict the evolution of taxi time, pushback time, and runway incursion;

[0104] Step 4.3.2: Use the operations research optimization solver module of PyEPO to establish a distributed robust strategy generation model for aircraft taxiing scheduling, and bring the prediction results of step 4.3.1 into the solver for solution;

[0105] Step 4.3.3: If the deviation of the prediction optimization result is less than the threshold, output the result; otherwise, repeat steps 4.3.1 and 4.3.2 continuously;

[0106] Step 4.4: Import the generated aircraft pushback time and taxiing path plan into the airport aircraft taxiing parallel simulation model based on virtual-real integration constructed in step 2, evaluate the impact of different traffic events on it in multiple dimensions, and generate constraints for illegal operation rules. If the constraints exist, return to step 4.2 and update the constraints of step 4.2.4; otherwise, return to step 4.5;

[0107] Step 4.5: Compare the historical and generated aircraft pushback times and taxi path plans, and dynamically formulate a feasible adjustment plan accordingly.

[0108] like Figure 2As shown, an embodiment of the present invention further provides a digital parallel simulation airport aircraft taxiing simulation scheduling system, the system includes an airport aircraft taxiing physical model, an airport aircraft taxiing parallel simulation model, an execution object (aircraft, etc.), an intelligent network sensor, an intelligent computing platform and a server; wherein the airport aircraft taxiing physical model is an entity, the airport aircraft taxiing parallel simulation model is installed on the server, and has a 1:1 real correspondence with the airport aircraft taxiing physical model; the intelligent network sensor is installed on the execution object and the airport aircraft taxiing physical model, and tracks the state of the micro-macro indicators of the field operation in real time, and stores the data on the server; the intelligent computing platform is installed on the server, and is mainly executed by a data-driven airport situation awareness model and an autonomous learning-based aircraft taxiing scheduling strategy generation model;

[0109] Airport aircraft taxiing physical model, which tracks the status of surface operation micro-macro indicators in real time based on intelligent network connection technology, integrates the status of surface operation micro-macro indicators, monitors the operation status of the taxiing network, warns of abnormal information and diagnoses the cause, wherein the surface operation micro-macro indicators include the operation trajectory of the aircraft, the operation status of the runway and the meteorological environment data;

[0110] The airport aircraft taxiing parallel simulation model builds a virtual reality environment of the airport aircraft taxiing physical model based on high-precision GIS, including parking positions, taxiway networks, and runway gate configurations, etc. It accesses real-time field operation multi-source heterogeneous data based on intelligent network technology, and visualizes the historical and current aircraft taxiing process and the real operation status of the runway;

[0111] The data-driven airport situation awareness model is connected to the operation data of the airport aircraft taxiing parallel simulation model. It uses large model technology to dynamically evaluate the current scene operation situation and predict the evolution trend of the future scene operation situation. Based on this, it detects real traffic events or upcoming events in the airport aircraft taxiing physical model.

[0112] An aircraft taxiing scheduling strategy generation model based on autonomous learning evaluates the deficiencies of different traffic events on the current aircraft pushback time and taxiing path plan in multiple dimensions according to the real-time and future traffic event detection results of the data-driven airport situation awareness model, and dynamically proposes a feasible adjustment plan based on this.

[0113] In this embodiment, the airport aircraft taxiing physical model includes:

[0114] The micro-indicator monitoring unit is used to monitor the status of micro-indicators of field operation, including:

[0115] Monitor the flight trajectory of aircraft based on video technology, GPS, and ADS-B, including the aircraft's location and time, speed, energy consumption, and plan execution deviation;

[0116] Monitor the status of whether the aircraft has left the parking position based on RFID;

[0117] Monitor the operation status of the runway based on video technology and RFID, including runway occupancy and idle time;

[0118] Monitoring of abnormal events based on video and radar technology, including the time, location, type, degree and duration of abnormal events;

[0119] Monitoring of meteorological environment data based on meteorological radar sensing technology, involving the time, location, type, degree and duration of extreme weather events;

[0120] The macro-indicator monitoring unit is used to monitor the status of the macro-indicators of the field operation, including:

[0121] Monitor the operation status of each taxiway network node and section based on video technology, including queue length, density, waiting time, etc.

[0122] Monitor the operation status of the taxiway network based on video technology, including the number of taxiing aircraft, number of stops, etc.;

[0123] The abnormal information warning unit is used to integrate the status of micro- and macro-indicators of surface operations and monitor the operating status of the taxiing network. When an indicator is abnormal, it warns of abnormal information and diagnoses the cause based on the causal correlation between their operations, such as the impact of weather and intrusion events on aircraft taxiing.

[0124] In this embodiment, the data-driven airport situation awareness model includes:

[0125] The information association unit is used to build a data-driven library of multiple deep learning algorithms based on time series or causal analysis according to different airport situation awareness goals (including operational situation assessment, trend prediction, event detection), objects (including pushback time, taxi time, taxi network status, runway intrusion, etc.) and time range (including long time, short time, etc.), and associate each deep learning algorithm with the goal, object and time range one by one;

[0126] An output label selection unit is used to select different output labels according to different airport situation awareness targets based on the multi-input-multi-output causal relationship label history data of various micro-macro indicator states of the field operation, and to determine the training samples and test data sets using different airport situation awareness objects and time ranges;

[0127] The association rule acquisition unit is used to select the best deep learning algorithm in the deep learning algorithm library according to different airport situation awareness targets, run various training samples and test data sets on the field, and obtain a series of association rules, that is, the one-to-one correspondence between each input indicator range and different operation situations / trends / events;

[0128] The result output unit is used to retrieve the possible association rule set matching the input conditions and evaluate its occurrence probability according to the current scene operation status, and output the results in descending order of probability.

[0129] In this embodiment, the airport aircraft taxiing parallel simulation model includes:

[0130] Environment construction unit, used to build digital virtual reality environment based on high-precision GIS, including GIS of sliding network physical model;

[0131] Data access unit, used to add different types of equipment layers GIS, including video, RFID, etc., to the taxiing network physical model GIS, and to interact with the intelligent network technology to access various micro-macro-indicator status data of the field operation;

[0132] Data calibration unit, used to accurately calibrate aircraft trajectory, weather information, runway status, abnormal events and other data using multi-source data fusion technology;

[0133] Data recording unit, used to automatically record multi-input and multi-output causal relationship label time series data of various micro-macro-indicator states of scene operation;

[0134] A data generation unit is used to associate the aircraft's trajectory with meteorological information, runway status, and abnormal events to automatically form a multi-input and multi-output causal relationship label data;

[0135] The current taxi tracking unit is used to dynamically track the current taxiing process of the aircraft, the current operating status of the runway, and the weather conditions at a 1:1 ratio, using the taxiing network physical model GIS as a visualization base;

[0136] The historical taxiing reproduction unit is used to reproduce the historical taxiing process of the aircraft, the current and historical operating status of the runway, and the weather status in a 1:1 scale multi-dimensional visualization based on the taxiing network physical model GIS.

[0137] It should be noted that, in this article, the term "comprises" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of more restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0138] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A digital parallel simulation airport aircraft taxiing simulation scheduling method, characterized in that: The method comprises the following steps: Step 1: The airport aircraft taxiing physical model tracks the status of surface operation micro-macro indicators in real time based on intelligent network connection technology, integrates the status of surface operation micro-macro indicators, monitors the operation status of the taxiing network, warns of abnormal information and diagnoses the cause, wherein the surface operation micro-macro indicator status includes the operation track of the aircraft, the operation status of the runway and the meteorological environment data; Step 2: The airport aircraft taxiing parallel simulation model builds a virtual reality environment of the airport aircraft taxiing physical model based on high-precision GIS, including parking positions, taxiway networks and runway gate configurations, accesses real-time field operation multi-source heterogeneous data based on intelligent network technology, and visualizes the historical and current aircraft taxiing process and the actual operation status of the runway; Step 3: Based on the data-driven airport situation awareness model, the operation data of the airport aircraft taxiing parallel simulation model is accessed, and the large model technology is used to dynamically evaluate the current scene operation situation and predict the evolution trend of the future scene operation situation, thereby detecting the real traffic events or upcoming events of the airport aircraft taxiing physical model; Step 4: The aircraft taxiing scheduling strategy generation model based on autonomous learning evaluates the deficiencies of different traffic events on the current aircraft pushback time and taxiing path plan based on the real-time and future traffic event detection results of the data-driven airport situation awareness model, and dynamically formulates a feasible adjustment plan.

2. The digital parallel simulation airport aircraft taxiing scheduling method according to claim 1 is characterized in that: The airport aircraft taxiing physical model tracks the state of micro- and macro-indicators of the surface operation in real time based on the intelligent network connection technology, integrates the state of the micro- and macro-indicators of the surface operation, monitors the operation state of the taxiing network, warns of abnormal information and diagnoses the causes, including the following steps: Step 1.1: Monitor the status of micro-indicators of field operation: Monitor the flight trajectory of aircraft based on video technology, GPS, and ADS-B, including the aircraft's location and time, speed, energy consumption, and plan execution deviation; Monitor the status of whether the aircraft has left the parking position based on RFID; Monitor the operation status of the runway based on video technology and RFID, including runway occupancy and idle time; Monitoring of abnormal events based on video and radar technology, including the time, location, type, degree and duration of abnormal events; Monitoring of meteorological environment data based on meteorological radar sensing technology, involving the time, location, type, degree and duration of extreme weather events; Step 1.2: Monitor the status of macro indicators of scene operation: Monitor the operation status of each taxiway network node and section based on video technology, including queue length, density, and waiting time; Monitor the operation status of the taxiway network based on video technology, including the number of taxiing aircraft and the number of stops; Step 1.3: Integrate the status of the micro- and macro-indicators of the field operation to monitor the operation status of the taxi network. When an indicator is abnormal, warn of the abnormal information and diagnose the cause based on the causal correlation between them.

3. The digital parallel simulation airport aircraft taxiing scheduling method according to claim 1 is characterized in that: The airport aircraft taxiing parallel simulation model constructs a virtual reality environment of the airport aircraft taxiing physical model based on high-precision GIS, including parking positions, taxiway networks and runway gate configurations, accesses real-time field operation multi-source heterogeneous data based on intelligent network connection technology, and visualizes the historical and current aircraft taxiing process and the actual operation status of the runway. The steps include: Building a digital virtual reality environment based on high-precision GIS, including a GIS for the physical model of the sliding network; In the taxiing network physical model GIS, add different types of equipment layer GIS, and interact with the intelligent network technology to access various micro-macro-indicator status data of the field operation; Use multi-source data fusion technology to accurately calibrate aircraft trajectory, weather information, runway status, and abnormal event data; Automatically record the multi-input and multi-output causal relationship label time series data of various micro-macro-indicator states of the scene operation; The aircraft's trajectory is associated with weather information, runway status, and abnormal events to automatically form a multi-input and multi-output causal relationship label data; Using the taxiing network physical model GIS as the visualization base, the current taxiing process of the aircraft, the current operating status of the runway, and the weather conditions are dynamically tracked at a 1:1 scale; Using the taxiing network physical model GIS as the visualization base, the historical taxiing process of aircraft, the current and historical operating status of the runway, and the weather status are reproduced in a 1:1 multi-dimensional visualization.

4. The digital parallel simulation airport aircraft taxiing scheduling method according to claim 1 is characterized in that: The data-driven airport situation awareness model accesses the operation data of the airport aircraft taxiing parallel simulation model, uses the large model technology to dynamically evaluate the current scene operation situation, predicts the evolution trend of the future scene operation situation, and detects the real traffic events or upcoming events of the airport aircraft taxiing physical model based on this, including the following steps: Step 3.1: According to the situation awareness goals, objects and time ranges of different airports, a library of multiple data-driven deep learning algorithms is constructed based on time series or causal analysis, and each deep learning algorithm is associated with the goal, object and time range one by one; Step 3.2: Based on the multi-input-multi-output causal relationship label historical data of the surface operation micro-macro indicator status, different output labels are selected according to different airport situation awareness goals, and the training samples and test data sets are determined using different airport situation awareness objects and time ranges; Step 3.3: Select the best deep learning algorithm from the deep learning algorithm library according to the situation awareness objectives of different airports, run various training samples and test data sets on the field, and obtain a series of association rules, that is, the one-to-one correspondence between each input indicator range and different operation situations / trends / events; Step 3.4: According to the current scene operation status, retrieve the possible association rule set that matches the input conditions and evaluate its probability of occurrence, and output the results in descending order of probability.

5. The digital parallel simulation airport aircraft taxiing scheduling method according to claim 1 is characterized in that: The autonomous learning-based aircraft taxiing scheduling strategy generation model evaluates the inadequacies of different traffic events on the current aircraft pushback time and taxiing path plan based on the real-time and future traffic event detection results of the data-driven airport situation awareness model, and dynamically generates a feasible adjustment plan based on the following steps: Step 4.1: Data preparation, including: (1) aircraft positions, runway gates, and paths; (2) real-time and future traffic event detection results based on the data-driven airport situational awareness model; (3) taxi network data; (4) operation rules; Step 4.2: Construct autonomous learning-based aircraft taxiing scheduling strategy generation model scenarios for different user scenarios, including: Step 4.2.1: User scenario: separation of arriving and departing flights; mixed arrival and departure; Step 4.2.2: Scheduling objectives: carbon emissions, taxiing distance and waiting time; Step 4.2.3: Autonomous learning objects: taxi time, push-off time, runway incursion; Step 4.2.4: Constraints: taxi path conflict, maximum taxi waiting time, maximum push-out delay time, path network flow constraints; Step 4.2.5: Decision variables: taxi path, aircraft pushback time, aircraft arrival and departure time through taxi nodes; Step 4.3: According to different scenarios, use PyEPO to build an aircraft taxiing scheduling strategy generation model engine based on autonomous learning, integrating data, knowledge, models and algorithms to generate aircraft pushback time and taxiing path plans; Step 4.3.1: Use the deep learning module of PyEPO to autonomously predict the evolution of taxi time, pushback time, and runway incursion; Step 4.3.2: Use the operations research optimization solver module of PyEPO to establish a distributed robust strategy generation model for aircraft taxiing scheduling, and bring the prediction results of step 4.3.1 into the solver for solution; Step 4.3.3: If the deviation of the prediction optimization result is less than the threshold, output the result; otherwise, repeat steps 4.3.1 and 4.3.2 continuously; Step 4.4: Import the generated aircraft pushback time and taxiing path plan into the airport aircraft taxiing parallel simulation model based on virtual-real integration constructed in step 2, evaluate the impact of different traffic events on it in multiple dimensions, and generate constraints for illegal operation rules. If the constraints exist, return to step 4.2 and update the constraints of step 4.2.4; otherwise, return to step 4.5; Step 4.5: Compare the historical and generated aircraft pushback times and taxi path plans, and dynamically make a feasible adjustment plan based on them.

6. A digital parallel simulation airport aircraft taxiing simulation dispatching system, characterized in that: The system includes an airport aircraft taxiing physical model, an airport aircraft taxiing parallel simulation model, an execution object, an intelligent network sensor, an intelligent computing platform and a server; wherein the airport aircraft taxiing physical model is an entity, and the airport aircraft taxiing parallel simulation model is installed on the server, which has a 1:1 real correspondence with the airport aircraft taxiing physical model; the intelligent network sensor is installed on the execution object and the airport aircraft taxiing physical model, and tracks the status of micro-macro indicators of the field operation in real time, and stores the data on the server; the intelligent computing platform is installed on the server, and is mainly executed by a data-driven airport situation awareness model and an autonomous learning-based aircraft taxiing scheduling strategy generation model; Airport aircraft taxiing physical model, which tracks the status of surface operation micro-macro indicators in real time based on intelligent network connection technology, integrates the status of surface operation micro-macro indicators, monitors the operation status of the taxiing network, warns of abnormal information and diagnoses the cause, wherein the surface operation micro-macro indicator status includes the aircraft's operation trajectory, runway operation status and meteorological environment data; The airport aircraft taxiing parallel simulation model builds a virtual reality environment of the airport aircraft taxiing physical model based on high-precision GIS, including parking positions, taxiway networks and runway gate configurations, accesses real-time field operation multi-source heterogeneous data based on intelligent network technology, and visualizes the historical and current aircraft taxiing process and the actual operation status of the runway; The data-driven airport situation awareness model is connected to the operation data of the airport aircraft taxiing parallel simulation model. It uses large model technology to dynamically evaluate the current scene operation situation and predict the evolution trend of the future scene operation situation. Based on this, it detects real traffic events or upcoming events in the airport aircraft taxiing physical model. An aircraft taxiing scheduling strategy generation model based on autonomous learning evaluates the deficiencies of different traffic events on the current aircraft pushback time and taxiing path plan in multiple dimensions according to the real-time and future traffic event detection results of the data-driven airport situation awareness model, and dynamically proposes a feasible adjustment plan based on this.

7. The digital parallel simulation airport aircraft taxiing simulation dispatching system according to claim 6 is characterized in that: The airport aircraft taxiing physical model includes: Micro-indicator monitoring unit, used to monitor the status of micro-indicators of field operation: Monitor the flight trajectory of aircraft based on video technology, GPS, and ADS-B, including the aircraft's location and time, speed, energy consumption, and plan execution deviation; Monitor the status of whether the aircraft has left the parking position based on RFID; Monitor the operation status of the runway based on video technology and RFID, including runway occupancy and idle time; Monitoring of abnormal events based on video and radar technology, including the time, location, type, degree and duration of abnormal events; Monitoring of meteorological environment data based on meteorological radar sensing technology, involving the time, location, type, degree and duration of extreme weather events; Macro indicator monitoring unit, used to monitor the status of macro indicators of field operation: Monitor the operation status of each taxiway network node and section based on video technology, including queue length, density, and waiting time; Monitor the operation status of the taxiway network based on video technology, including the number of taxiing aircraft and the number of stops; The abnormal information warning unit is used to integrate the micro- and macro-indicator status of the field operation, monitor the operation status of the taxiing network, and when a certain indicator is abnormal, it warns of the abnormal information and diagnoses the cause based on the causal correlation between their operations.

8. The digital parallel simulation airport aircraft taxiing simulation dispatching system according to claim 6 is characterized in that: The data-driven airport situation awareness model includes: The information association unit is used to build a data-driven library of multiple deep learning algorithms based on time series or causal analysis according to the situation awareness targets, objects and time ranges of different airports, and associate each deep learning algorithm with the target, object and time range one by one; An output label selection unit is used to select different output labels according to different airport situation awareness targets based on the multi-input-multi-output causal relationship label history data of the micro-macro indicator status of the surface operation, and to determine the training samples and test data sets using different airport situation awareness objects and time ranges; The association rule acquisition unit is used to select the best deep learning algorithm in the deep learning algorithm library according to different airport situation awareness targets, run various training samples and test data sets on the field, and obtain a series of association rules, that is, the one-to-one correspondence between each input indicator range and different operation situations / trends / events; The result output unit is used to retrieve the possible association rule set matching the input conditions and evaluate its occurrence probability according to the current scene operation status, and output the results in descending order of probability.

9. The digital parallel simulation airport aircraft taxiing simulation dispatching system according to claim 6, characterized in that: The airport aircraft taxiing parallel simulation model includes: Environment construction unit, used to build digital virtual reality environment based on high-precision GIS, including GIS of sliding network physical model; Data access unit, used to add different types of equipment layer GIS in the taxiing network physical model GIS, and interact with the intelligent network technology to access various micro-macro-indicator status data of the field operation; Data calibration unit, used to accurately calibrate the aircraft's trajectory, weather information, runway status, and abnormal event data using multi-source data fusion technology; Data recording unit, used to automatically record multi-input and multi-output causal relationship label time series data of various micro-macro-indicator states of scene operation; A data generation unit is used to associate the aircraft's trajectory with meteorological information, runway status, and abnormal events to automatically form a multi-input and multi-output causal relationship label data; The current taxi tracking unit is used to dynamically track the current taxiing process of the aircraft, the current operating status of the runway, and the weather conditions at a 1:1 ratio, using the taxiing network physical model GIS as a visualization base; The historical taxiing reproduction unit is used to reproduce the historical taxiing process of the aircraft, the current and historical operating status of the runway, and the weather status in a 1:1 scale multi-dimensional visualization based on the taxiing network physical model GIS.

Images