Training flight path optimization method and system under combination of low-altitude training and airport, and medium

Abstract

The invention discloses a training flight path optimization method and system in combination with low-altitude training and transportation in an airport and a medium. Relates to the technical field of air transportation and flight training. Improvement is carried out on the basis of the prior art, a built training flight airspace grid model finely expresses training flight airspace information, the fine degree of airspace resource management is improved, a multi-target optimization model comprehensively considering safety, economy and green targets is built, and through reasonable setting of constraint conditions, the optimization efficiency of airspace resource management is improved. The optimization result better meets the actual flight training requirement, and the risk, oil consumption and pollutant emission of training flight are effectively reduced; based on the improved artificial bee colony algorithm, the solving efficiency and quality are improved, a better solution can be quickly found in a complex multi-objective optimization problem, and reliable algorithm support is provided for training flight trajectory optimization.

Description

Technical Field

[0001] This invention relates to the field of air transport and flight training technology, specifically to a method, system, and medium for optimizing training flight trajectories at low-altitude training and transport combined airports. Background Technology

[0002] Low-altitude airspace generally refers to airspace below 3,000 meters in altitude. Training and operational airports, as special locations capable of simultaneously undertaking flight training missions for pilots and regular passenger transport missions, play a vital role in the aviation industry. However, with the booming development of the global aviation industry, the demand for pilot training is increasing daily, and passenger travel frequency is rising significantly, making the challenges faced by training and operational airports increasingly severe.

[0003] From the perspective of airspace resource management, the vertical range of low-altitude airspace is limited, and competition for airspace between training and transport flights is becoming increasingly fierce. Traditional static airspace segmentation models based on latitude and longitude are difficult to adapt to the dynamic changes in training flight scenarios and the frequent adjustments in airspace occupancy demands. Training flights involve numerous subjects, such as eight basic training flight path subjects including takeoff and landing procedures and right-angle procedures. Each subject has different requirements for flight trajectory, altitude, and duration, and must avoid transport flight routes. However, existing models cannot accurately match these complex requirements, resulting in unreasonable allocation of airspace resources and low utilization rates.

[0004] In terms of flight safety, low-altitude flight speeds create significant risks of potential conflicts between training flights and civil aviation transport flights. Some training routes intersect with busy arrival and departure routes, greatly increasing the difficulty of air traffic control and the workload of controllers, and seriously threatening flight safety. The consequences of a flight conflict would be unimaginable, potentially leading to tragedies of aircraft destruction and loss of life, causing enormous casualties and property damage.

[0005] In terms of environmental impact, the slow convection velocity at low altitudes makes the fuel consumption and pollutant emissions from training flights a significant concern. Training flights are mostly visual flight maneuvers, resulting in redundant flight paths and increased fuel waste and pollutant emissions. SO2 and NO emissions during aircraft flight... x Pollutants such as HC and CO can pollute the atmospheric environment, affect the quality of life of surrounding residents, and hinder the sustainable development of the aviation industry.

[0006] From the perspective of flight training quality, in low-altitude environments, trainees need to handle multiple tasks simultaneously, including terrain avoidance, obstacle detection, and conflict resolution with transport flights. However, existing training models lack simulation tools for low-altitude special situations (such as low-altitude wind shear and terrain approach), making it difficult for trainees to develop accurate trajectory prediction capabilities. During the ground preparation phase, trainees cannot fully understand the trajectory details during flight, making it difficult to achieve the expected flight results, leading to higher fuel consumption and emissions. Furthermore, traditional training models focus on basic flight skills training, insufficiently cultivating trainees' ability to cope with complex flight environments, and failing to meet the comprehensive quality requirements of modern aviation technology for pilots.

[0007] Furthermore, existing flight procedure design and trajectory optimization algorithms also have shortcomings. Although a certain standard system for flight procedure design has been established internationally and domestically, and numerous auxiliary design software programs exist, there are still deficiencies in integrating flight procedure design with trajectory optimization algorithms, making it difficult to achieve optimal trajectory planning. Regarding flight training quality assessment, while existing assessment methods are diverse, some tools are not closely integrated with actual flight training, failing to provide real-time and comprehensive feedback on training dynamics. This affects the timeliness and accuracy of assessments, hindering the improvement of trainees' flight skills.

[0008] In summary, airports that integrate training and operations face numerous challenges in planning and managing training flight trajectories. There is an urgent need for an innovative method and system to achieve multi-objective optimization of training flight trajectories, improve airport operational efficiency, ensure flight safety, reduce environmental impact, and enhance the quality of flight training. Summary of the Invention

[0009] The technical problem this invention aims to solve is that conventional flight trajectory planning methods have many issues in the planning and management of training flight trajectories at low-altitude training and operation airports. This invention provides a method, system, and medium for optimizing training flight trajectories at low-altitude training and operation airports. Based on existing technologies, it improves upon existing techniques by establishing a training flight airspace grid model that precisely represents training flight airspace information, improving the precision of airspace resource management. It constructs a multi-objective optimization model that comprehensively considers safety, economy, and environmental goals. By rationally setting constraints, the optimization results better meet actual flight training needs, effectively reducing the risks, fuel consumption, and pollutant emissions of training flights. Based on an improved artificial bee colony algorithm, it improves solution efficiency and quality, enabling rapid finding of optimal solutions in complex multi-objective optimization problems, providing reliable algorithmic support for training flight trajectory optimization, and realizing knowledge construction.

[0010] This invention is achieved through the following technical solution:

[0011] This solution provides a method for optimizing training flight trajectories at low-altitude training and transport combined airports, characterized by including:

[0012] Determine the training flight airspace and construct a grid model of the training flight airspace;

[0013] A multi-objective optimization model for training flight trajectories is constructed. The multi-objective optimization model simultaneously considers the influencing factors of the number of flight conflicts during the simultaneous operation of transport and training flights, the influencing factors of fuel consumption during the training phase (decomposed into the turning phase and the level flight phase), and the influencing factors of the main pollutant emissions of the training aircraft during the training process.

[0014] Solving the multi-objective optimization model yields the optimal training flight trajectory.

[0015] A further optimized solution involves determining the training flight airspace and constructing a training flight airspace grid model, including the following methods:

[0016] Based on GeoSOT theory, training flight airspace is decomposed from low-altitude training and operation airport. In the vertical direction, the airspace is divided into layers with each n-meter height layer, and in the horizontal direction, multiple airspace grid cells are obtained by using regular hexagons as the basic grid cells.

[0017] The address, height, and attributes of each spatial raster cell are encoded, and each spatial raster cell is modeled.

[0018] A further optimization scheme is that the method for constructing the multi-objective optimization model includes:

[0019] The influencing factors of the multi-objective optimization model are obtained and normalized. The influencing factors include: the number of flight conflicts during the operation of transport and training flights in the same field, Z1; the fuel consumption of the training subject phase decomposed into the turning phase and the level flight phase, Z2; and the main pollutant factor of the training aircraft during the training process, Z3.

[0020] Weights were assigned to each of the normalized influencing factors.

[0021] The objective function of the multi-objective optimization model is to minimize the sum of all influencing factors after weighting.

[0022] A further optimization scheme is that the method for obtaining the flight conflict frequency factor Z1 during the simultaneous operation of transport and training flights includes:

[0023] The flight conflict frequency factor Z1 during simultaneous transport and training flights is calculated using the following formula:

[0024] ;

[0025] Where N represents the total number of flights in the training airspace within a certain time period, and c iThis indicates that i flight conflicts occurred in the training airspace within a certain time period; This represents the horizontal coordinate of the j-th path node in the training flight trajectory, used for initializing trajectory nodes and limiting the search space in the artificial bee colony algorithm.

[0026] A further optimization scheme is that the method for obtaining the fuel consumption factor Z2 of the turning phase and the level flight phase in the fuel consumption decomposition of the training subject phase includes:

[0027] ;

[0028] Among them, t s Indicates the duration of each flight phase; 's' represents the takeoff phase, climb phase, level flight phase, turn phase, or approach phase; FF s N1 represents the fuel consumption of the aircraft during flight phase s; N2 represents the number of engines on the training aircraft.

[0029] A further optimized solution is that the method for obtaining the main pollutant factor Z3 during the training process of the training aircraft includes:

[0030] ;

[0031] Among them, EI p,s This represents the emission factor p of pollutants during flight phase s; p represents SO2, NO. x HC and CO; T s This represents the duration of the s-th flight phase.

[0032] A further optimization scheme involves solving the multi-objective optimization model to obtain the optimal training flight trajectory; including the following methods:

[0033] An artificial bee colony algorithm is introduced to search for training flight trajectory strategies in a training flight airspace grid model; path nodes are determined by a bidirectional search mechanism consisting of one-way search of odd-numbered path nodes and one-way search of even-numbered path nodes.

[0034] In the initialization phase of the artificial bee colony algorithm, the horizontal coordinates of the path nodes are determined at fixed intervals based on the distance between the starting point and the target point, thus imposing a dimensional constraint. This constraint limits the range of values ​​for the path nodes in the horizontal coordinate (x-direction) when the artificial bee colony algorithm searches for training flight trajectories. This compresses the full-space search that originally needed to be performed in two or three-dimensional space into a "straight line" or "narrow band", reducing the search dimensionality and improving efficiency.

[0035] During the search phase of the artificial bee colony algorithm, the generated flight training trajectory point set is evaluated by the fitness function, and a greedy algorithm is used to select a better nectar source.

[0036] A further optimization scheme is that the rules for the dimensional constraints include:

[0037] When cou≤n / 2 ;

[0038] When n / 2 < cou ≤ r ;

[0039] When r < cou ≤ r + n / 2r ;

[0040] When r+n / 2r<cou≤2r ;

[0041] ; ; ;

[0042] Where i = 1, 2, 3, ..., n are the honey source numbers, and cou represents the number of algorithm iterations; The initialization path node array represents a component of the D-dimensional solution vector; This represents the initial value of the x-coordinate of the first path node in the i-th candidate trajectory; This indicates the first stage of flight, namely the takeoff stage; This indicates the first aircraft to participate in the optimization calculation; The expression represents the initial value of the x-coordinate of the second path node in the i-th candidate trajectory; The overall value is a scale normalization coefficient used to limit the dimension of the x-coordinate of the path nodes; the value of r is usually taken as half of the total number of iterations or an empirically set value.

[0043] This solution also provides a training flight trajectory optimization system for low-altitude training and transport combined airports, used to implement the aforementioned training flight trajectory optimization method for low-altitude training and transport combined airports; the system includes:

[0044] The first building module is used to determine the training flight airspace and build a training flight airspace grid model;

[0045] The second construction module is used to construct a multi-objective optimization model for training flight trajectories. The multi-objective optimization model simultaneously considers the influencing factors of the number of flight conflicts during the simultaneous operation of transportation and training flights, the influencing factors of fuel consumption during the training phase (decomposed into the turning phase and level flight phase), and the influencing factors of major pollutant emissions from the training aircraft during the training process.

[0046] The solution module is used to solve the multi-objective optimization model to obtain the optimal training flight trajectory.

[0047] This solution also provides a computer-readable medium storing a computer program, which, when executed by a processor, can implement the training flight trajectory optimization method under the low-altitude training and operation combined airport as described above.

[0048] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0049] 1. The present invention provides a method, system, and medium for optimizing training flight trajectories at low-altitude training and operation combined airports. Based on existing technologies, improvements are made to establish a training flight airspace grid model that precisely expresses training flight airspace information, improving the precision of airspace resource management. A multi-objective optimization model comprehensively considers safety, economy, and environmental goals. By rationally setting constraints, the optimization results better meet actual flight training needs, effectively reducing the risks, fuel consumption, and pollutant emissions of training flights. Based on an improved artificial bee colony algorithm, the solution efficiency and quality are improved, enabling rapid finding of optimal solutions in complex multi-objective optimization problems, providing reliable algorithmic support for training flight trajectory optimization.

[0050] 2. The present invention provides a method, system, and medium for optimizing training flight trajectories at low-altitude training and operation combined airports; the model and algorithm are verified by AirTOp simulation software, ensuring the feasibility and effectiveness of the proposed method in practical applications and reducing the cost and risk of actual flight tests. Attached Figure Description

[0051] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0052] Figure 1 A schematic diagram of the process for optimizing training flight trajectories at airports that combine low-altitude training and operations;

[0053] Figure 2 A schematic diagram illustrating the process of constructing a raster model of the flight airspace;

[0054] Figure 3 This is a schematic diagram illustrating the process of partitioning the training flight airspace based on GeoSOT theory.

[0055] Figure 4 This is a schematic diagram of a point-based spatial grid model;

[0056] Figure 5 This is a schematic diagram of a linear spatial grid model;

[0057] Figure 6 This is a schematic diagram of a planar spatial grid model;

[0058] Figure 7 A schematic diagram illustrating the solution process of the improved artificial bee colony algorithm;

[0059] Figure 8 A schematic diagram of the flight trajectory for angle correction training in the unoptimized AirTOp simulation software;

[0060] Figure 9 This is a schematic diagram of the training flight trajectory for the corrected angle subject in the optimized training flight grid airspace. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0062] Conventional flight trajectory planning methods face numerous challenges in planning and managing training flight trajectories at low-altitude training and operational airports. Therefore, this solution provides the following embodiments to address these technical problems:

[0063] Example 1

[0064] This embodiment provides a method for optimizing training flight trajectories at low-altitude training and transport combined airports, such as... Figure 1 As shown, it includes:

[0065] Step 1: Determine the training flight airspace and construct a grid model of the training flight airspace; for example... Figures 2-6 As shown, this step specifically includes the following methods:

[0066] S11, based on GeoSOT theory, training flight airspace is partitioned from a low-altitude training and operation combined airport. Vertically, the airspace is divided into layers with an altitude of n meters, and horizontally, it is divided into multiple airspace grid cells using regular hexagonal grid units as the basic grid cells. Specifically, in this embodiment, vertically, the airspace is divided into layers with an altitude of 300 meters. A schematic diagram of the grid partitioning is shown below. Figure 3 As shown;

[0067] S12 encodes the address, altitude, and attributes of each airspace grid cell and models each airspace grid cell. Considering key elements of training flights, such as training airspace boundary coordinates, return point, turning point, airport ARP point, and training flight path, the addresses, altitudes, and attributes of the subdivided airspace grid cells are encoded. Point, line, and area airspace grid models are established respectively. The constructed point, line, and area airspace grid models are as follows: Figure 4-6As shown, this provides a precise airspace representation for subsequent flight trajectory optimization.

[0068] Step 2: Construct a multi-objective optimization model for the training flight trajectory. The multi-objective optimization model simultaneously considers the influencing factors of the number of flight conflicts during the simultaneous operation of transport and training flights, the influencing factors of fuel consumption during the training phase (decomposed into the turning phase and level flight phase), and the influencing factors of the main pollutant emissions of the training aircraft during the training process.

[0069] Methods for constructing multi-objective optimization models include:

[0070] S21, obtain the influencing factors of the multi-objective optimization model and normalize the influencing factors; the influencing factors include: the number of flight conflicts during the operation of transport and training flights in the same field, factor Z1; the fuel consumption of the training subject phase decomposed into the turning phase and the level flight phase, factor Z2; and the main pollutant factor Z3 of the training aircraft during the training process.

[0071] The flight conflict frequency factor Z1 during simultaneous transport and training flights is calculated using the following formula:

[0072] ;

[0073] Where N represents the total number of flights in the training airspace within a certain time period, and c i This indicates that i flight conflicts occurred in the training airspace within a certain time period; express This represents the horizontal coordinate of the j-th path node in the training flight trajectory, used for initializing trajectory nodes and limiting the search space in the artificial bee colony algorithm.

[0074] The methods for obtaining the fuel consumption factor Z2 for the training phase, which is decomposed into the turning phase and the level flight phase, include:

[0075] ;

[0076] Among them, t s Indicates the duration (min) of each flight phase; s represents the takeoff phase, climb phase, level flight phase, turn phase, or approach phase; FF s N1 represents the aircraft's fuel consumption (kg / s) during flight phase s; N2 represents the number of engines on the training aircraft.

[0077] The methods for obtaining the main pollutant factor Z3 during training aircraft include:

[0078] ;

[0079] Among them, EI p,sThis represents the emission factor (g / kg) of pollutant p during flight phase s; p represents SO2, NO... x HC and CO; T s This represents the duration of the s-th flight phase.

[0080] This scheme comprehensively considers safety, economy, and environmental friendliness, and designs a comprehensive optimization target for training flight trajectories. The safety target is measured by the number of flight conflicts occurring during simultaneous transport and training flight operations, i.e., the flight conflict frequency factor Z1. The economy target is calculated by improving the LTO method proposed by ICAO, decomposing the fuel consumption of the training phase into the fuel consumption of the turning phase and the level flight phase, i.e., the fuel consumption of the training phase is decomposed into the fuel consumption factor Z2 of the turning and level flight phases. The environmental friendliness target calculates the main pollutants (SO2, NO) emitted by the training aircraft during training based on the emission factor method specified by ICAO. x The three influencing factors (HC and CO, etc.) are the main pollutant factors Z3 of the training aircraft during training. The three influencing factors are processed into a multi-objective optimization function model of the training flight trajectory by weighted normalization method. According to the training outline and the special characteristics of training flight, conditions such as mandatory point constraints, training flight altitude constraints, training flight time constraints, engine operating status constraints, and flight turning constraints are set.

[0081] To improve the efficiency and accuracy of the multi-objective optimization model, the influencing factors are normalized. The flight conflict frequency factor c is used as an example. i For example, the normalization formula is:

[0082] ;

[0083] Among them, c min c represents the minimum risk of airborne operations. max This indicates the maximum risk level for air operations. This indicates the standard value for airborne operational risks.

[0084] After normalization, all optimization objectives are mapped to the [0,1] interval, eliminating the influence of dimensions, which is conducive to the rapid convergence of the multi-objective optimization model, finding the training flight trajectory that meets the needs of multi-objective optimization, and achieving the overall minimization of air operation risks, fuel consumption and pollutant emissions.

[0085] S22, assign weights to each of the normalized influencing factors respectively;

[0086] S23, the objective function of the multi-objective optimization model is to minimize the sum of all influencing factors after weighting.

[0087] The optimization objective is to minimize the overall flight operation risks, fuel consumption, and pollutants generated by the training aircraft during training. Based on constraints such as fixed start and end points and fixed altitude, a multi-objective optimization model for the training flight trajectory is established, with the objective function as follows: ;in, Indicates the weight of Z1; This represents the weight of Z2; Indicates the weight of Z3;

[0088] Step 3: Solve the multi-objective optimization model to obtain the optimal training flight trajectory. This step specifically includes the following methods:

[0089] An artificial bee colony algorithm is introduced to search for training flight trajectory strategies in a training flight airspace grid model; path nodes are determined by a bidirectional search mechanism consisting of one-way search of odd-numbered path nodes and one-way search of even-numbered path nodes.

[0090] In the initialization phase of the artificial bee colony algorithm, the horizontal coordinates of the path nodes are determined at fixed intervals based on the distance between the starting point and the target point, thus imposing a dimensional constraint. This constraint limits the range of values ​​for the path nodes in the horizontal coordinate (x-direction) when the artificial bee colony algorithm searches for training flight trajectories. This compresses the full-space search that originally needed to be performed in two or three-dimensional space into a "straight line" or "narrow band", reducing the search dimensionality and improving efficiency.

[0091] During the search phase of the artificial bee colony algorithm, the generated flight training trajectory point set is evaluated by the fitness function, and a greedy algorithm is used to select a better nectar source.

[0092] Dimensional constraint rules include:

[0093] When cou≤n / 2 ;

[0094] When n / 2 < cou ≤ r ;

[0095] When r < cou ≤ r + n / 2r ;

[0096] When r+n / 2r<cou≤2r ;

[0097] ; ; ;

[0098] Where i = 1, 2, 3, ..., n are the honey source numbers, and cou represents the number of algorithm iterations; The initialization path node array represents a component of the D-dimensional solution vector; This represents the initial value of the x-coordinate of the first path node in the i-th candidate trajectory; This indicates the first stage of flight, namely the takeoff stage; This indicates the first aircraft to participate in the optimization calculation; The expression represents the initial value of the x-coordinate of the second path node in the i-th candidate trajectory; The overall value is a scale normalization coefficient used to limit the dimension of the x-coordinate of the path nodes; the value of r is usually taken as half of the total number of iterations or an empirically set value.

[0099] This embodiment introduces the artificial bee colony algorithm to solve the multi-objective optimization model and improves its search strategy; a bidirectional search mechanism consisting of one-way search of odd-numbered path nodes and one-way search of even-numbered path nodes is used to determine path nodes, thereby enhancing the algorithm's global search capability.

[0100] During the search process, the generated flight training trajectory point set is evaluated using a fitness function, and a greedy algorithm is used to select a better honey source (i.e., the flight trajectory point set). When the algorithm meets the preset maximum number of iterations, or when the fitness value of the optimal route does not improve significantly after multiple consecutive iterations, the route with the best fitness value is output, which is the optimized training flight trajectory.

[0101] Finally, the multi-objective optimization model and algorithm were validated. Taking the simultaneous operation of training and transport flights on a typical weekday at a combined training and transport airport as an example, the established multi-objective optimization model and algorithm for training flight trajectories were used to obtain relevant operational results (airborne operational risk, fuel consumption, and pollutant emission data). Using AirTOp simulation software, a ground and airspace simulation model of the airport was established based on aeronautical chart data and training airspace maps. Flight schedules were compiled and imported, and the same constraints as actual operations were set. The simulation model was run, and operational data were collected. The optimization calculation results were compared with the simulation operational data collected based on the AirTOp airspace simulation model to verify the rationality of the proposed method.

[0102] Example 2

[0103] This embodiment provides a training flight trajectory optimization system for low-altitude training and transport combined airports, used to implement the training flight trajectory optimization method for low-altitude training and transport combined airports described in Embodiment 1; the system includes:

[0104] The first building module is used to determine the training flight airspace and build a training flight airspace grid model;

[0105] The second construction module is used to construct a multi-objective optimization model for training flight trajectories. The multi-objective optimization model simultaneously considers the influencing factors of the number of flight conflicts during the simultaneous operation of transportation and training flights, the influencing factors of fuel consumption during the training phase (decomposed into the turning phase and level flight phase), and the influencing factors of major pollutant emissions from the training aircraft during the training process.

[0106] The solution module is used to solve the multi-objective optimization model to obtain the optimal training flight trajectory.

[0107] Example 3

[0108] This embodiment provides a computer-readable medium storing a computer program thereon. The computer program, executed by a processor, can implement the training flight trajectory optimization method for low-altitude training and operation combined airports as described in Embodiment 1; specifically, it performs the following steps:

[0109] Step 1: Determine the training flight airspace and construct a training flight airspace grid model; specifically, determine the training flight airspace boundary: based on the actual range of the training flight, determine the boundary coordinates of the training flight airspace.

[0110] According to GeoSOT theory, the grid is divided into layers at 300-meter intervals in the vertical direction and subdivided into regular hexagonal grid cells in the horizontal direction. For example, starting from the 0° latitude and longitude position, the grid is divided into regular hexagonal grids with a fixed side length (e.g., 32 meters) based on the center of the basic grid cell, and adjacent regular hexagonal grids are connected by a common edge.

[0111] The segmented spatial grid cells are then encoded, including altitude encoding, address encoding, and attribute encoding. Altitude encoding calculates the altitude level based on the feature's elevation and is represented by a combination of the capital letter H and the altitude level. Address encoding is calculated using a specific formula based on latitude and longitude information and the location level. Attribute encoding is represented by two digits for different key elements in training flights.

[0112] Point, linear, and area spatial grid models were established separately. For the point spatial model, the improved GeoSOT level 8 (corresponding to 1″×1″ grid precision) was selected for encoding based on the positioning accuracy requirements of key elements. The linear spatial model is composed of multiple continuous grid cells, and the subdivision level was selected for encoding according to the accuracy matching criterion. When representing the area spatial model, 16 sub-grids belonging to the same parent level were clustered together, and the parent level grid was used for encoding.

[0113] Step 2: Construct a multi-objective optimization model for the training flight trajectory. The multi-objective optimization model simultaneously considers the influencing factors of the number of flight conflicts during the simultaneous operation of transport and training flights, the influencing factors of fuel consumption during the training phase (decomposed into the turning phase and level flight phase), and the influencing factors of the main pollutant emissions of the training aircraft during the training process.

[0114] The three optimization goals are clearly defined as safety, economy, and environmental friendliness, which are measured by the number of flight conflicts, fuel consumption, and pollutant emissions, respectively.

[0115] Set constraints: Based on the training syllabus and actual flight conditions, set constraints such as mandatory arrival point constraints, training flight altitude constraints, training flight time constraints, engine operating status constraints, and flight turning constraints.

[0116] Constructing the objective function: By using a weighted normalization method, the three optimization objectives are processed into a multi-objective optimization function model for training flight trajectories.

[0117] Step 3: Solve the multi-objective optimization model to obtain the optimal training flight trajectory. For example... Figure 7 As shown, the specific steps include:

[0118] Initialize algorithm parameters: Set parameters such as starting point, target point, self-cost constraints and threat-related information, weight size, number of algorithm iterations, number of nectar sources, number of algorithm runs, number of bees, upper and lower limits of the environment, environment dimension, upper limit of the number of times nectar sources have not been updated, and number of path nodes.

[0119] Initialize the honey source: Randomly generate a set of available flight training trajectory points (honey source) using the following formula, and simplify the search range based on dimensional constraints. Calculate fitness: Calculate the total cost value of each feasible solution using the cost function formula, and then calculate the fitness using the fitness calculation formula.

[0120] ;

[0121] Among them, H ij U represents the j-th dimension variable of the i-th nectar source; ij D ij H represents ij The upper and lower bounds of r1; r1 represents a random number within (0,1).

[0122] Finding new nectar sources: Find new nectar sources according to the following formula, and use a greedy algorithm to compare the fitness of new and old nectar sources and select the better nectar source.

[0123] ;

[0124] Among them, H newLet r be the new nectar source found through the search; r2 be a random number; and i ≠ k. By comparing the fitness of the new and old nectar sources, the higher-quality nectar source is retained based on the greedy principle.

[0125] Honey Source Selection: A roulette wheel algorithm is used to select honey sources. If a source is updated, it is retained; otherwise, appropriate processing is performed. The roulette wheel algorithm is used in the honey source selection phase of the artificial bee colony algorithm. Specifically, it allocates selection probabilities based on the fitness value of the honey source (i.e., the quality of the trajectory optimization target), prioritizing the selection of better honey sources. Its core logic is similar to a "roulette wheel lottery": the fitness value of each honey source (the set of training flight trajectory points) is converted into a "sector area" on the roulette wheel; the higher the fitness of the honey source, the larger the corresponding sector area. During the selection process, a numerical simulation "wheel" is randomly generated. When the value falls into the sector corresponding to a honey source, that honey source is selected. This method ensures that honey sources with better fitness (i.e., trajectories that better meet safety, economy, and environmental goals) have a higher probability of being retained and iteratively optimized, thereby improving the efficiency of the algorithm's convergence to the optimal solution and ensuring that the final output training flight trajectory is superior in terms of conflict risk, fuel consumption, and pollutant emissions.

[0126] Assess the status of nectar source exploitation: Based on the comparison between the number of times a nectar source has not been updated and its upper limit, decide whether to abandon the nectar source and search for a new one.

[0127] Iterative optimization: The counter is incremented, and the steps of calculating fitness, finding new nectar sources, selecting nectar sources, and judging the mining status of nectar sources are repeated until the number of iterations reaches the set value.

[0128] Output: The path nodes are optimized and smoothed, and the optimized training flight trajectory is output.

[0129] Finally, a simulation model was built: the unoptimized AirTOp simulation software correction angle training flight trajectory is shown in the figure below. Figure 8 As shown, using AirTOp simulation software, based on aeronautical chart data and training airspace maps, a ground and airspace simulation model of this training and operation combined airport was established. The optimized training flight trajectory diagram for the corrected angle subject in the training flight grid airspace is shown below. Figure 9The process includes generating approach, departure, and training airspace procedures, and setting up the airport surface structure and operating rules. Flight plans are compiled and imported: Based on typical weekday flight plans and training syllabus requirements, transport flight plans and training plans are compiled and imported into the AirTOp simulation software as CSV files. Simulation parameters are set: Runway operation mode, aircraft type, operating interval, training subject operations, number of training aircraft, and flight duration are set. The simulation model is run and the results are analyzed: The simulation model is run, and the unoptimized and optimized corrected angle procedure training flight operation data (airborne operational risks, fuel consumption, and pollutants) are statistically analyzed and compared to verify the scientific validity and rationality of the model and algorithm.

[0130] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing training flight trajectories at low-altitude training and operation combined with airport operations, characterized in that: include: Determine the training flight airspace and construct a grid model of the training flight airspace; A multi-objective optimization model for training flight trajectories is constructed. The multi-objective optimization model simultaneously considers the influencing factors of the number of flight conflicts during the simultaneous operation of transport and training flights, the influencing factors of fuel consumption during the training phase (decomposed into the turning phase and the level flight phase), and the influencing factors of the main pollutant emissions of the training aircraft during the training process. Solving the multi-objective optimization model yields the optimal training flight trajectory.

2. The method for optimizing training flight trajectories at low-altitude training and transport combined airports according to claim 1, characterized in that, The method for determining the training flight airspace and constructing a training flight airspace grid model includes: Based on GeoSOT theory, training flight airspace is decomposed from low-altitude training and operation airport. In the vertical direction, the airspace is divided into layers with each n-meter height layer, and in the horizontal direction, multiple airspace grid cells are obtained by using regular hexagons as the basic grid cells. The address, height, and attributes of each spatial raster cell are encoded, and each spatial raster cell is modeled.

3. The method for optimizing training flight trajectories at low-altitude training and transport combined airports according to claim 1, characterized in that, The method for constructing the multi-objective optimization model includes: The influencing factors of the multi-objective optimization model are obtained and normalized. The influencing factors include: the number of flight conflicts during the operation of transport and training flights in the same field, Z1; the fuel consumption of the training subject phase decomposed into the turning phase and the level flight phase, Z2; and the main pollutant factor of the training aircraft during the training process, Z3. Weights were assigned to each of the normalized influencing factors. The objective function of the multi-objective optimization model is to minimize the sum of all influencing factors after weighting.

4. The method for optimizing training flight trajectories at low-altitude training and transport combined airports according to claim 3, characterized in that, The method for obtaining the flight conflict frequency factor Z1 during the simultaneous operation of transport and training flights includes: The flight conflict frequency factor Z1 during simultaneous transport and training flights is calculated using the following formula: ； Where N represents the total number of flights in the training airspace within a certain time period, and c i This indicates that i flight conflicts occurred in the training airspace within a certain time period; This represents the horizontal coordinate of the j-th path node in the training flight trajectory.

5. The method for optimizing training flight trajectories at low-altitude training and transport combined airports according to claim 4, characterized in that, The method for obtaining the fuel consumption factor Z2 for the turning and level flight phases in the training subject phase includes: ； Among them, t s Indicates the duration of each flight phase; 's' represents the takeoff phase, climb phase, level flight phase, turn phase, or approach phase; FF s N1 represents the fuel consumption of the aircraft during flight phase s; N2 represents the number of engines on the training aircraft.

6. The method for optimizing training flight trajectories at low-altitude training and transport combined airports according to claim 5, characterized in that, The method for obtaining the main pollutant factor Z3 during the training process of the training aircraft includes: ； Among them, EI p,s This represents the emission factor p of pollutants during flight phase s; p represents SO2, NO. x HC and CO; T s This represents the duration of the s-th flight phase.

7. The method for optimizing training flight trajectories at low-altitude training and transport combined airports according to claim 2, characterized in that, The method for solving the multi-objective optimization model to obtain the optimal training flight trajectory includes: An artificial bee colony algorithm is introduced to search for training flight trajectory strategies in a training flight airspace grid model; path nodes are determined by a bidirectional search mechanism consisting of one-way search of odd-numbered path nodes and one-way search of even-numbered path nodes. In the initialization phase of the artificial bee colony algorithm, the horizontal coordinates of the path nodes are determined at fixed intervals based on the distance between the starting point and the target point, and the values ​​of the path nodes on the horizontal coordinates are subject to dimensional constraints. During the search phase of the artificial bee colony algorithm, the generated flight training trajectory point set is evaluated by the fitness function, and a greedy algorithm is used to select a better nectar source.

8. The method for optimizing training flight trajectories at low-altitude training and transport combined airports according to claim 7, characterized in that, The rules for the dimensional restrictions include: When cou≤n / 2 ; When n / 2 < cou ≤ r ; When r < cou ≤ r + n / 2r ; When r+n / 2r<cou≤2r ; ； ； ； Where i = 1, 2, 3, ..., n are the honey source numbers, and cou represents the number of algorithm iterations; The initialization path node array represents a component of the D-dimensional solution vector; This represents the initial value of the x-coordinate of the first path node in the i-th candidate trajectory; This indicates the first stage of flight, namely the takeoff stage; This indicates the first aircraft to participate in the optimization calculation; The expression represents the initial value of the x-coordinate of the second path node in the i-th candidate trajectory; The overall value is a scale normalization coefficient used to limit the dimension of the x-coordinate of the path nodes; the value of r is usually taken as half of the total number of iterations or an empirically set value.

9. A training flight trajectory optimization system for low-altitude training and operations combined with airport operations, characterized in that: The system is used to implement the training flight trajectory optimization method under a low-altitude training and transport combined airport as described in any one of claims 1-8; the system includes: The first building module is used to determine the training flight airspace and build a training flight airspace grid model; The second construction module is used to construct a multi-objective optimization model for training flight trajectories. The multi-objective optimization model simultaneously considers the influencing factors of the number of flight conflicts during the simultaneous operation of transportation and training flights, the influencing factors of fuel consumption during the training phase (decomposed into the turning phase and level flight phase), and the influencing factors of major pollutant emissions from the training aircraft during the training process. The solution module is used to solve the multi-objective optimization model to obtain the optimal training flight trajectory.

10. A computer-readable medium having a computer program stored thereon, characterized in that, The computer program, when executed by a processor, can implement the training flight trajectory optimization method for low-altitude training and transport combined airports as described in any one of claims 1-8.

Images