A green real-time planning method for four-dimensional flight trajectory in aircraft terminal area

Abstract

The application provides a green real-time planning method for four-dimensional flight trajectories in an aircraft terminal area, comprising: using radar track data, identifying track patterns in the terminal area by means of a DBSCAN clustering algorithm in machine learning, and limiting the planning space of a track planning algorithm by the track patterns; using a grid method to model the terminal area airspace in real time, proposing a limited airspace probability boundary, dynamically representing the terminal area airspace situation, and supplementing the shortcomings of the track pattern in limiting the planning space while reducing the algorithm planning space again; aiming at the limited algorithm planning space, proposing a track planning model and an improved genetic algorithm, and planning a green four-dimensional track in real time. The application takes the green target and operation efficiency as the breakthrough point, solves the problem that the existing method cannot consider the solving efficiency and solving quality by reducing the track planning space and improving the genetic algorithm, and provides a more scientific and effective tool for the green real-time planning of four-dimensional flight trajectories in the aircraft terminal area in practical applications.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of aircraft trajectory planning, and particularly relates to a green real-time planning method for a four-dimensional flight trajectory of an aircraft in a terminal area. BACKGROUND

[0002] In recent years, the civil aviation industry has developed rapidly, and the national flight transportation volume has rapidly increased, which has brought about increasingly serious problems such as air traffic operation and flight delay. The reason is that the flight operation process is complex and changeable, and the terminal area as a relatively busy airspace in the aircraft operation process is more widely affected and difficult to predict.

[0003] In order to solve this problem, the effective way is to make the aircraft fly on a reasonable track planned by a system. Due to the reasons such as many constraint conditions of the track planning problem and high requirement for planning timeliness, the method used determines the quality of problem solving. At present, the track planning algorithm usually includes two methods: traditional planning method and intelligent planning method. The traditional method includes artificial potential field method, A* method, D* method, etc.; the intelligent planning method includes genetic algorithm, ant colony algorithm, neural network, particle swarm algorithm, etc. An excellent track planning can greatly improve the aircraft operation efficiency on the basis of ensuring safety.

[0004] At present, the research on the direction of track planning is mostly based on the innovation of algorithms such as A* algorithm and improved ant colony algorithm, and most of the existing researches are for unmanned aerial vehicle flight rather than civil aviation transportation task. The research on civil aviation transportation mostly focuses on traditional directions such as conflict detection and removal, and there is still a lack of research on green four-dimensional track planning. SUMMARY

[0005] In view of the above deficiencies, the present application provides a green dynamic real-time planning method for a flight trajectory of an aircraft in a terminal area, which aims to alleviate the current situation of busy terminal airspace and lack of consideration of environmental impact, and simultaneously identify the restricted airspace in real time to plan a real-time optimal green four-dimensional track.

[0006] To achieve the above purpose, the technical scheme adopted by the present application is as follows:

[0007] A green real-time planning method for a four-dimensional flight trajectory of an aircraft in a terminal area, the method comprising:

[0008] acquiring historical radar track data in a target terminal airspace;

[0009] performing clustering processing on the historical radar track data to obtain a center track of each track mode, and acquiring a track planning space corresponding to each center track;

[0010] The flight path planning space is rasterized to obtain a raster flight path planning space;

[0011] According to the real-time meteorological data of the target terminal area airspace, the real-time restricted space in the raster flight path planning space is determined;

[0012] The decision variables, decision boundaries and state values in each flight path planning space are obtained, and an optimization model with a target function of minimizing the state value and a constraint condition of the decision boundary is constructed; wherein the state value is the aviation emission amount corresponding to the operation of the aircraft in the terminal area airspace;

[0013] Based on the raster flight path planning space and the real-time restricted airspace, the optimization model is optimized and solved by using an improved genetic algorithm, to obtain an optimal terminal area four-dimensional green real-time planning flight path.

[0014] In some embodiments, the historical radar flight path data in the target terminal area airspace includes longitude, latitude, height, speed and flight number dimensions of all aircrafts at any time within a specified time period in the target terminal area airspace.

[0015] In some embodiments, the flight path planning space corresponding to the center flight path is obtained, including:

[0016] Step 1: selecting a certain center flight path h j as a target center flight path, and extracting the flight path points p1, p2, …, p n , by default, taking p1 as the starting point;

[0017] Step 2: extracting the longitude long, latitude lat and height alt of each flight path point in sequence;

[0018] Step 3: dividing the rectangular region within dx±k, dy±k and dz±k on the three coordinate axes of the flight path point into the flight path planning space of the flight path point; wherein dx, dy and dz respectively represent the three directions of the horizontal axis, the vertical axis and the vertical axis, and k is the expansion range of the planning space;

[0019] Step 4: traversing the flight path points of the target center flight path to complete the flight path planning space restriction, to obtain the flight path planning space corresponding to the target center flight path.

[0020] In some embodiments, the flight path planning space is rasterized to obtain a raster flight path planning space, including:

[0021] Step 1: obtaining the terminal area airspace boundary longitude and latitude coordinates, and the maximum longitude long max , the maximum latitude lat max , the minimum longitude long min , and the minimum latitude lat min ;

[0022] Step 2: Convert the latitude and longitude coordinates of the terminal airspace boundary into Cartesian coordinate system coordinates:

[0023] x = R * (long i -long0)

[0024]

[0025] where lat i , lat0 and long i , long0 represent the longitude and latitude of point i and the coordinate origin, respectively, R represents the Earth's radius, and x and y are the Cartesian coordinate system coordinates of the point;

[0026] Step 3: Based on the maximum longitude long max , maximum latitude lat max , minimum longitude long min , and minimum latitude lat min of the terminal airspace boundary, according to the selected grid horizontal size L and vertical size H, the number of grids in the x, y, and z directions of the terminal airspace is calculated, and the occupied grid of the flight path planning space is calculated to obtain the grid flight path planning space.

[0027] In some embodiments, according to the real-time weather data of the target terminal airspace, the real-time restricted airspace in the grid flight path planning space is determined, including:

[0028] According to the real-time weather data of the target terminal airspace, the grids of the grid flight path planning space are classified into forbidden zone grids, buffer zone grids, and operating zone grids.

[0029] The real-time restricted airspace is determined according to the probability boundary of the forbidden zone grids.

[0030] Further, wherein the real-time weather data of the terminal airspace includes the centroid C w (X, Y, Z) and the influence radius R w of the dangerous weather area, when there is a dangerous weather area in the grid airspace, it is defined as a forbidden zone grid, a layer of grids outside the forbidden zone grid is set as a buffer zone grid, and the remaining part is set as an operating zone grid.

[0031] Assuming that there is a dynamic dangerous weather in the terminal area at a certain time, the centroid of the dangerous weather area is C w (X, Y, Z), and the influence radius is R w , after the random change based on the probability boundary, the centroid position C w '(X, Y, Z) and the influence radius R w ' of the dangerous weather area are respectively:

[0032] C w(X, Y, Z) = C w (X+Y2, Y+Y2, Z+Y2)

[0033] R w ' = R w +Y1;

[0034] Y n = k n *X, X ~ N(0, σ 2 ), n ∈ (1, 2)

[0035] where Y n is the radius or centroid change amount, k n is a scaling factor to move the probability boundary within a reasonable range, X is subject to a standard normal distribution, and n ∈ (1, 2) represents radius change or centroid change.

[0036] The probability boundary of the exclusion zone grid is to convert the determined restricted airspace boundary into a probability boundary. Within a certain confidence interval, the restricted airspace boundary follows a certain probability distribution and is divided into two cases: random change of radius of dangerous weather area and random movement of centroid of dangerous weather area.

[0037] In some embodiments, the decision variable is the flight strategy of the aircraft at each stage, including the running grid, the turning angle, and the speed.

[0038] In some embodiments, the decision boundary includes a grid decision boundary, a turning angle decision boundary, and a speed decision boundary.

[0039] The grid decision boundary is represented as:

[0040]

[0041] i ∈ U S

[0042] where, for the planned flight path where i I , i O represent the start and end points of the flight path, respectively; the flight point i within the planned flight path cannot exist within the restricted airspace U P , and any central flight path X all central flight path points i X corresponding to the grid flight path planning space S x,i form a set U S , which limits the next flight point i planning space.

[0043] The turning angle decision boundary is that the aircraft cannot exceed the turning angle of the aircraft performance when turning in flight. The turning angle decision boundary is represented as:

[0044]

[0045] wherein, a max is the maximum turning angle of the corresponding aircraft, l1, l2, l3 represent the length of the current flight segment, the length of the next flight segment, and the length from the start of the current flight segment to the end of the next flight segment, respectively;

[0046] The speed decision boundary is that the aircraft cannot exceed the speed range in the corresponding control airspace when flying at each altitude layer, and the turning angle decision boundary is expressed as:

[0047] v∈[V cr,Z,min ,V cr,Z,max ]

[0048] v∈[V cli,Z,min ,V cli,Z,max ]∩[V cli,Z+1,min ,V cli,Z+1,max ]

[0049] v∈[V des,Z,min ,V des,Z,max ]∩[V des,Z-1,min ,V des,Z-1,max ]

[0050] The three formulas respectively represent that the speed v of the aircraft itself cannot exceed the speed limit of the current and destination altitude layer when flying straight, climbing, and descending, respectively. cr,Z,min ,V cr,Z,max , V cli,Z,min ,V cli,Z,max , V des,Z,min ,V des,Z,max respectively represent the maximum and minimum true airspeeds of the altitude layer Z corresponding to the cruising, climbing, and descending states.

[0051] In some embodiments, the aviation emission discharge amount is represented by the following formula:

[0052]

[0053] wherein, EM represents the aviation emission discharge amount, E j,N represents the discharge amount of each type of aviation emission after processing by the normalized extreme value method, and n represents the total number of types of aviation emissions.

[0054] The normalized extreme value method is represented by the following formula:

[0055]

[0056] wherein, E j , maxE j , and minE j are the discharge amount, maximum discharge amount, and minimum discharge amount of the pollutant j, respectively.

[0057]

[0058] wherein Eij j is the emission index of pollutant j, representing the mass of the corresponding emission produced by burning 1 kg of fuel, Fij i represents the fuel flow rate of the aircraft at waypoint i; Tij i represents the time of flight of the aircraft within waypoint i;

[0059]

[0060] wherein Fij CR,i , Fij min,i , Fij nom,i are the fuel flow rates of different parts in the BADA model of the aircraft performance database, PHSij i represents the flight state at waypoint i, CFGij i represents the aircraft configuration at waypoint i.

[0061] In some embodiments, the various types of aviation emissions include CO2, SO2, NO x , CO.

[0062] In some embodiments, the optimization model is solved by using an improved genetic algorithm, including:

[0063] Step 1: generating individuals in a preset population number according to an individual generation rule;

[0064] Step 2: completing the crossover and mutation operations of the individuals according to the designed crossover operator and mutation operator;

[0065] Step 3: selecting individuals with better fitness from the individuals after the crossover and mutation operations;

[0066] Step 4: repeating Step 2-Step 3 until the maximum number of iterations or the target fitness is reached, to obtain the optimal four-dimensional green real-time planning route of the terminal area;

[0067] The improved genetic algorithm is an improved version of the basic genetic algorithm for solving the route planning problem, including:

[0068] a. The individual generation rule is that the individual can only search for the next waypoint within the adjacent grid in the random generation process, and a hyperparameter a is introduced. When searching for the adjacent grid as the next waypoint, the distance of all adjacent grids to the end point is calculated, wherein the closer the grid is, the greater the probability of being selected is. The probability formula is as follows:

[0069]

[0070] Wherein prob represents the probability of the grid being selected, dis represents the Euclidean distance of the grid from the terminal grid;

[0071] b, the crossover operator is designed: when requiring the parent individuals to be paired, the individuals with better fitness are paired, the individuals with lower fitness are paired, and among the two parent individuals to be crossed, the same genes of the two individuals are searched, and the same point is taken as the crossover point to complete the crossover operation;

[0072] c, the adaptive genetic algorithm, the crossover probability P c And the mutation probability P m Can change with the change of population fitness, and is represented by the following formula:

[0073]

[0074]

[0075] Wherein, k1 and k2 represent the maximum probability in the crossover and mutation operation, the value is between 0 and 1, and β is a hyperparameter for controlling the change range, S 2 Is the variance of the fitness of the entire population.

[0076] A green real-time planning system for four-dimensional flight trajectories in an aircraft terminal area, the system comprising a processor and a storage medium;

[0077] The storage medium is used for storing instructions;

[0078] The processor is used for operating according to the instructions to perform the steps of the above-mentioned method.

[0079] A computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the steps of the above-mentioned method.

[0080] Compared with the prior art, the beneficial effects achieved by the present application are:

[0081] The present application processes radar track data, uses a DBSCAN clustering algorithm to identify track patterns in the terminal area airspace, uses track patterns to reduce track planning space, and avoids the problem of low search efficiency due to excessively large track planning space; the present application establishes a real-time terminal area grid airspace based on the grid method for the terminal area airspace, performs gridding on the interior of the airspace based on the demarcation of the airspace range, simultaneously introduces a terminal area restricted airspace with a probability boundary, and completes optimization of the planning space; the present application uses an improved genetic algorithm to respectively plan CO2, SO2, NO xThe emission modeling is performed for environmental protection targets such as CO, and the comprehensive emissions of the four types of emissions after extreme value standardization are considered, the operation is performed in the planning space after limitation, and the green four-dimensional flight path in the terminal area is obtained in real time. BRIEF DESCRIPTION OF DRAWINGS

[0082] Figure 1 A method flowchart of an embodiment of the present application is shown in the figure.

[0083] Figure 2 A flight path planning space discretization diagram of an embodiment of the present application is shown in the figure.

[0084] Figure 3 (a) and (b) are respectively a green flight path diagram planned by an embodiment of the present application and a green flight path obstacle avoidance diagram planned in real time. DETAILED DESCRIPTION

[0085] The present application will be further described below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application, and cannot be used to limit the protection scope of the present application.

[0086] The present application establishes a green four-dimensional flight path real-time planning model based on economic and environmental protection targets, reduces the algorithm planning space through flight path mode and grid processing, improves the algorithm calculation speed, provides a model and method basis for the green four-dimensional flight path real-time planning of an aircraft in a terminal area airspace, and provides strong support for improving the sustainable development capability of civil aviation.

[0087] A dynamic memory planning method of a green cruise trajectory of an aircraft, comprising:

[0088] Obtaining historical radar flight path data in a target terminal area airspace;

[0089] Performing clustering processing on the historical radar flight path data to obtain a center flight path of each flight path mode, and obtaining a flight path planning space corresponding to each center flight path;

[0090] Performing grid processing on the flight path planning space to obtain a grid flight path planning space;

[0091] Determining a real-time restricted airspace in the grid flight path planning space according to real-time weather data of the target terminal area airspace;

[0092] Obtaining decision variables, decision boundaries and state values in each flight path planning space, and constructing an optimization model with a minimum state value as an objective function and the decision boundaries as constraint conditions; wherein the state value is an aviation emission amount corresponding to the operation of the aircraft in the terminal area airspace;

[0093] Based on the grid flight path planning space and the real-time restricted airspace, an improved genetic algorithm is used to optimize and solve the optimization model, and an optimal terminal area four-dimensional green real-time planning flight path is obtained.

[0094] As shown in Figure 1 A dynamic memory planning method for an aircraft green cruise trajectory, the method comprising:

[0095] Step 1: Obtain historical radar track data in the target terminal area airspace;

[0096] Step 1.1: Process default values and outliers of radar track data;

[0097] Step 1.2: Delete duplicate values of radar track data;

[0098] Step 1.3: Dimensionality reduction processing of radar track data;

[0099] Step 2: DBSCAN clustering processing of historical radar track data, dividing historical radar track data into multiple clusters, obtaining center tracks of each cluster, and obtaining track planning space corresponding to each center track;

[0100] Step 2.1: After data cleaning, default value processing and data dimensionality reduction of historical radar track data, DBSCAN algorithm clustering analysis is adopted to identify the center track of each cluster;

[0101] Step 2.2: Expand the center track to form a track planning space, and the track planning space calculation method comprises:

[0102] Step 2.2.1: Select a center track h j as a target center track, and extract the track points p1, p2, …, p n , by default, p1 is the starting point;

[0103] Step 2.2.2: Extract the longitude long, latitude lat, and height alt of each track point in order;

[0104] Step 2.2.3: Divide the rectangular region within dx±k, dy±k, dz±k in the three coordinate axes of the track point into the track point planning space; wherein dx, dy, and dz represent the three directions of the horizontal axis, the vertical axis, and the vertical axis, respectively, and k is the expansion range of the planning space;

[0105] Step 2.2.4: Traverse the track points of the target center track to complete the planning space restriction of the track points, and obtain the track planning space corresponding to the target center track.

[0106] Step 3: In the track planning space, perform rasterization processing to obtain a raster track planning space, update the raster division in real time according to the real-time weather data of the target terminal area airspace, and determine the real-time restricted airspace U P in the raster track planning space.

[0107] Step 3.1: Conversion from continuous space represented by latitude and longitude to discrete space represented by grid, the steps are as follows:

[0108] Step 3.1.1: Obtain the terminal area boundary latitude and longitude coordinate points, confirm their geographical location, and obtain the maximum longitude long max , maximum latitude lat max and minimum longitude long min , minimum latitude lat min boundary point values;

[0109] Step 3.1.2: The calculation formula for converting latitude and longitude coordinates to rectangular coordinate system coordinates is:

[0110] x = R * (long i -long0)

[0111]

[0112] Where lat i , lat0 and long i , long0 represent the longitude and latitude of the i point and the coordinate origin respectively, R represents the radius of the earth, and x and y are the rectangular coordinate system coordinates of the point;

[0113] Step 3.1.3: Select the horizontal size L and vertical size H of the grid, calculate the number of grids in the x, y, z directions of the terminal area airspace through the maximum and minimum points obtained in step 3.1.1, and calculate the track planning space occupied by the grid to obtain the grid track planning space, as shown in Figure 2 ;

[0114] Step 3.2: Meteorological data of terminal area airspace, including the centroid C w (X, Y, Z) and radius R w of the dangerous weather area, when there is a dangerous weather area in the grid airspace, even if it is not completely occupied, it will be defined as a forbidden grid, a layer of grid outside the forbidden grid is set as a buffer grid, and the rest is set as a running grid. Among them, the forbidden grid has a probability boundary to simulate the randomness of weather in actual operation;

[0115] Step 3.3: Forbidden grid probability boundary, which converts the determined restricted airspace boundary into a probability boundary, that is, within a certain confidence interval, the restricted airspace boundary follows a certain probability distribution, which is divided into two cases: random change of dangerous weather radius and random movement of dangerous weather centroid;

[0116] Assume that there is a dynamic dangerous weather in the terminal area at a certain time, the dangerous weather area centroid is C w (X, Y, Z), and the influence radius is Rw , after the random change based on the probability boundary, the dangerous weather area centroid position C w '(X, Y, Z) and the influence radius R w ' are respectively:

[0117] C w '(X, Y, Z) = C w (X + Y2, Y + Y2, Z + Y2)

[0118] R w ' = R w + Y1;

[0119] Y n = k n *X, X ~ N(0, σ 2 ), n ∈ (1, 2)

[0120] where Y n is the radius or centroid change amount, k n is a scaling factor to move the probability boundary within a reasonable range, X is subject to a standard normal distribution, and n ∈ (1, 2) represents the radius change or centroid change. In the above formula, if the change amount obtained is positive, it means moving in the positive direction of the coordinate axis, otherwise it means moving in the negative direction.

[0121] Step 4: Define the decision variables, decision boundaries and state value set of each stage, the specific steps are:

[0122] Step 4.1: The decision variable is the flight strategy of the aircraft in each stage, including the running grid, the turning angle and the speed; the decision boundary includes the grid decision boundary, the turning angle decision boundary and the speed decision boundary;

[0123] Step 4.1.1: The grid decision boundary is that the aircraft cannot fly in the corresponding forbidden zone grid when running, and the grid decision boundary is represented by the following formula:

[0124]

[0125] i ∈ U S

[0126] where, for the flight path , the flight path point i within it cannot exist in the restricted airspace U P , the set U S is formed by all the center flight path points i X of any center flight path X x,i ;

[0127] Step 4.1.2: The turning angle decision boundary is the maximum turning angle of the aircraft when the aircraft is turning in flight, and the turning angle decision boundary is expressed by the following formula:

[0128]

[0129] wherein, a max is the corresponding maximum turning angle of the aircraft, and l1, l2, l3 represent the current leg length, the next leg length, and the length from the start of the current leg to the end of the next leg, respectively;

[0130] Step 4.1.3: The speed decision boundary is the speed range in the corresponding control airspace that the aircraft cannot exceed when flying at each altitude, and the turning angle decision boundary is expressed by the following formula:

[0131] v∈[V cr,Z,min ,V cr,Z,max ]

[0132] v∈[V cli,Z,min ,V cli,Z,max ]∩[V cli,Z+1,min ,V cli,Z+1,max ]

[0133] v∈[V des,Z,min ,V des,Z,max ]∩[V des,Z-1,min ,V des,Z-1,max ]

[0134] The three formulas respectively represent that the speed of the aircraft itself cannot exceed the speed limit of the current and destination altitude when flying, ascending, and descending;

[0135] Step 4.2: The state value is the amount of aviation emissions generated by the aircraft when operating in the terminal area, and the steps are as follows:

[0136] Step 4.2.1: The amount of aviation emissions is expressed by the following formula:

[0137]

[0138] wherein, EM represents the amount of aviation emissions, measured in grams, and E j,N represents the amount of each type of aviation emissions after processing by the standardized extreme value method.

[0139] The standardized extreme value method is expressed by the following formula:

[0140]

[0141] wherein E j is the amount of pollutant j (kg);

[0142] Step 4.2.2: Emissions of various pollutants are CO2, SO2, and NO. x The emissions of CO and other pollutants during flight can be expressed by the following formulas:

[0143]

[0144] Where E j EI represents the emission amount (kg) of pollutant j. j The emission index (g / kg) for pollutant j represents the mass of emissions produced by burning 1 kg of fuel oil. FF i T represents the fuel flow rate of the aircraft at trackpoint i; i This indicates the flight time of the aircraft within waypoint i.

[0145] For CO2 and SO2 emissions and The values ​​will be constant, at 3115 g / kg and 0.8 g / kg respectively; for NO x The relationship between CO and these two emissions indices and total fuel consumption is also related, but it is also related to engine operating conditions and atmospheric conditions, corresponding to different values ​​under different conditions. Since the emission indices for these two types of emissions in the EEDB database only apply to International Standard Atmosphere and sea level (0 meters, ISA) conditions, it is necessary to correct and interpolate the emission indices during operation. The steps are as follows:

[0146] Step 4.2.2.1: Calculate the fuel flow rate FF at any given time. i The method uses fuel flow rate and emission index for fitting, and interpolates the corresponding emission index baseline value E using the baseline fuel flow rate value in the database. j '.

[0147] Step 4.2.2.2: Based on the atmospheric conditions at the aircraft's current flight altitude, adjust the emission index baseline value E... j Correct to the actual value using the following formula:

[0148]

[0149] Where δ is the ratio of the atmospheric pressure at the aircraft's current location to the standard sea-level pressure, θ is the ratio of the current atmospheric temperature to the standard sea-level atmospheric temperature, and P is the current atmospheric pressure. The relative humidity (%) is p v The saturated vapor pressure is related to the atmospheric temperature as follows:

[0150]

[0151] Therefore, the established optimization model is:

[0152] The objective function of the optimization model is:

[0153] The constraint condition of the optimization model is the decision boundary, including the grid decision boundary, the steering angle decision boundary and the speed decision boundary.

[0154] Step 4.3: The improved genetic algorithm is an improved version proposed on the basis of the genetic algorithm for solving the path planning problem, which includes:

[0155] a. The individual generation rule design requires that the individual can only search for the next path point in the adjacent grid during the random generation process. The super parameter a is introduced, and when searching for the adjacent grid as the next path point, the distance of all adjacent grids from the end point will be calculated. The grid with a shorter distance has a higher probability of being selected. The probability formula is as follows:

[0156]

[0157] Where prob represents the probability of the grid being selected, and dis represents the Euclidean distance of the grid from the end grid;

[0158] b. The crossover operator design requires that the parent individuals are paired between the individuals with better fitness and the individuals with lower fitness, and in the two parent individuals about to be crossed, the same genes of the two individuals are found to complete the crossover operation at the same point as the crossover point.

[0159] c. Adaptive genetic algorithm, crossover probability P c and mutation probability P m can change with the change of population fitness, which can be represented by the following formula:

[0160]

[0161]

[0162] Where k1 and k2 represent the maximum probability in the crossover and mutation operations, taking values between 0 and 1, and β is a super parameter that controls the change amplitude, S 2 is the variance of the entire population fitness.

[0163] Step 5: Green four-dimensional real-time planning path includes:

[0164] Step 5.1: Generate individuals of a preset population size according to the individual generation rule. The individual must meet the requirements of steps 2.2, 3 and 4.1 during the generation process.

[0165] Step 5.2: according to the design of the crossover operator and the mutation operator, the crossover and mutation operations of the individual are completed;

[0166] Step 5.3: among the individuals after the crossover and mutation operations, the individual with better fitness is selected;

[0167] Step 5.4: steps 5.2 to 5.3 are repeated until the maximum number of iterations or the target fitness is reached, and a green four-dimensional flight path is planned in real time, as shown in Figure 3 .

[0168] A green real-time planning device for a four-dimensional flight path in an aircraft terminal area, comprising:

[0169] A data acquisition module: for acquiring historical radar track data in a target terminal area airspace;

[0170] A first processing module: for clustering the historical radar track data to obtain a center track of each track mode, and acquiring a track planning space corresponding to each center track;

[0171] A second processing module: for rasterizing the track planning space to obtain a raster track planning space U S ; and determining a real-time restricted airspace U P in the raster track planning space according to real-time weather data of the target terminal area airspace;

[0172] An acquisition module: for decision variables, decision boundaries and state values in each track planning space, and constructing an optimization model with a minimum state value as an objective function and a decision boundary as a constraint condition; wherein the state value is the amount of aviation emissions generated when the aircraft operates in the terminal area airspace;

[0173] A real-time calculation module: for optimizing and solving the optimization model based on the raster track planning space and the real-time restricted airspace using an improved genetic algorithm, to obtain an optimal terminal area four-dimensional green real-time planning track.

[0174] A green real-time planning system for a four-dimensional flight path in an aircraft terminal area, the system comprising a processor and a storage medium;

[0175] The storage medium is used to store instructions;

[0176] The processor is used to operate according to the instructions to perform the steps of the above-mentioned method.

[0177] A computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the steps of the above-mentioned method.

[0178] Compared with the prior art, the present application has the following beneficial effects:

[0179] The application identifies the track mode in the terminal area airspace by processing the radar track data using the DBSCAN clustering algorithm, and reduces the track planning space using the track mode, thereby avoiding the problem of low search efficiency due to excessively large track planning space; the application establishes a real-time terminal area grid airspace based on the grid method for the terminal area airspace, and performs gridding on the interior of the airspace on the basis of the demarcation of the airspace range, and introduces a terminal area restricted airspace with a probability boundary, thereby completing optimization of the planning space; the application adopts an improved genetic algorithm to perform emission modeling on environmental protection targets such as CO2, SO2, NO x , CO, considers the comprehensive emission of the four types of emissions after extreme value standardization, and performs operation in the limited planning space, thereby obtaining a green four-dimensional track in the terminal area in real time.

[0180] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus generate an apparatus that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in the flowcharts and / or block diagrams.

[0181] These computer program instructions can also be stored in a computer-readable memory that can direct the computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction apparatus, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in the flowcharts and / or block diagrams.

[0182] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable data processing apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable data processing apparatus provide steps for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in the flowcharts and / or block diagrams.

[0183] The above merely illustrates the embodiments of the present application, but should not be taken as limitations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall fall into the protection scope of the present application.

Claims

1. A green real-time planning method for four-dimensional flight trajectories in an aircraft terminal area, characterized in that, The method comprises: acquiring historical radar track data in a target terminal airspace; clustering the historical radar track data to obtain a center track of each track mode, and acquiring a track planning space corresponding to each center track; performing rasterization processing on the track planning space to obtain a raster track planning space; determining a real-time restricted airspace in the raster track planning space according to real-time meteorological data of the target terminal airspace; acquiring decision variables, decision boundaries and state values in each track planning space, and constructing an optimization model with a minimum state value as an objective function and the decision boundaries as constraint conditions; wherein the state value is an aviation emission amount corresponding to the operation of an aircraft in the terminal airspace; based on the raster track planning space and the real-time restricted airspace, performing optimization and solving on the optimization model by using an improved genetic algorithm to obtain an optimal terminal four-dimensional green real-time planning track, including: step 1: generating individuals in a preset population number according to an individual generation rule; step 2: completing the crossover and mutation operations of the individuals according to a designed crossover operator and a mutation operator; step 3: selecting individuals with better fitness from the individuals after the crossover and mutation operations; and step 4: repeating steps 2-3 until a maximum iteration number or a target fitness is reached to obtain the optimal terminal four-dimensional green real-time planning track; wherein the improved genetic algorithm is an improved type proposed for solving the track planning problem based on a basic genetic algorithm, including: a. The individual generation rule is: require individuals to search for the next track point in the adjacent grid in the random generation process, and introduce a super parameter When searching for adjacent grids as the next track point, the distance of all adjacent grids from the end point will be calculated, and the grid with a shorter distance will have a higher probability of being selected. The probability formula is as follows: ； wherein represents the probability that the grid is selected, represents the Euclidean distance of the grid from the end grid; b. the crossover operator is designed to require the parent individuals to be paired, the pairing being performed between individuals with better fitness and between individuals with lower fitness, and in the two parent individuals about to be crossed, the same genes of the two individuals are found to complete the crossover operation at the same point as the crossover point; c. Adaptive Genetic Algorithm, crossover probability and mutation probability which varies with the population fitness, and is represented by the following equation: ； ； wherein, and represents the maximum probability in the crossover and mutation operations, and has a value between 0 and 1, is a hyperparameter that controls the amplitude of its variation, is the variance of the fitness of the entire population.

2. The method for green real-time planning of four-dimensional flight trajectories in an aircraft terminal area according to claim 1, characterized in that, the historical radar track data in the target terminal airspace includes longitude, latitude, height, speed and flight number dimensions of all aircrafts at any time within a specified time period in the target terminal airspace.

3. The method for green real-time planning of four-dimensional flight trajectories in an aircraft terminal area according to claim 1, characterized in that, acquiring the track planning space corresponding to the center track, including: Step 1: Select a central track As the target center track, extract the track points of the target center track one by one. , default As the starting point; Step 2: Extract the longitude of each track point in sequence , latitude , altitude ; Step 3: Place the corresponding track points on the three coordinate axes , , The rectangular area within is divided into a planning space for waypoints; where , , These represent the horizontal, vertical, and center axes, respectively. To plan the expansion of the space; step 4: traversing the track points of the target center track to complete the planning space limitation of the track points, and obtaining the track planning space corresponding to the target center track.

4. The method for green real-time planning of four-dimensional flight trajectories in an aircraft terminal area according to claim 1, characterized in that, performing rasterization processing on the track planning space to obtain a raster track planning space, including: Step 1: Obtain the latitude and longitude coordinates of the terminal area airspace boundary and the maximum longitude , maximum latitude , minimum longitude , and minimum latitude ; step 2: converting the longitude and latitude coordinates of the terminal airspace boundary into coordinates of an orthogonal coordinate system: ； ； in, , and , Represent The longitude and latitude of the point and the origin of the coordinate system. Represents the Earth's radius. and The coordinates of this point are in a rectangular coordinate system. Step 3: maximum longitude based on terminal area airspace boundary , maximum latitude , minimum longitude , minimum latitude , according to the selected grid horizontal size and vertical size , the number of grids in , , three directions of the terminal area airspace is calculated, and the track planning space occupied grid is calculated to obtain the grid track planning space.

5. The method for green real-time planning of four-dimensional flight trajectories in an aircraft terminal area according to claim 1, characterized in that, determining the real-time restricted airspace in the raster track planning space according to the real-time meteorological data of the target terminal airspace, including: according to the real-time meteorological data of the target terminal airspace, classifying the grids of the raster track planning space into forbidden grid, buffer grid and running grid; determining the real-time restricted airspace according to the probability boundary of the forbidden grid.

6. The method for green real-time planning of four-dimensional flight trajectories in an aircraft terminal area according to claim 5, characterized in that, The real-time weather data of the terminal area airspace includes a dangerous weather area centroid and an influence radius When there is a dangerous weather area in the grid airspace, the grid is defined as a forbidden grid, a layer of grids outside the forbidden grid is set as a buffer grid, and the remaining part is set as a running grid. Assume that there is a dynamic dangerous weather at some time in the terminal area, the dangerous weather area centroid is , the influence radius is , after the random change based on the probability boundary, the dangerous weather area centroid position and the influence radius are respectively: ； ； ； wherein, is a radius or centroid change amount, is a scaling factor to move the probability bounds within a reasonable range, then the probability is subject to a standard normal distribution, represents a radius change or a centroid change; The probability boundary of the forbidden grid is to convert the determined restricted airspace boundary into a probability boundary. Within a certain confidence interval, the restricted airspace boundary follows a certain probability distribution, which includes random changes in the radius of the dangerous weather area and random movements of the dangerous weather area centroid.

7. The green real-time 4D flight trajectory planning method for aircraft in terminal area according to claim 1, characterized in that, the decision variables are the flight strategies of the aircraft at each stage, including the running grid, the turning angle and the speed.

8. The method for green real-time planning of four-dimensional flight trajectories in an aircraft terminal area according to claim 1, characterized in that, The decision boundary comprises a grid decision boundary, a turning angle decision boundary and a speed decision boundary; The grid decision boundary is expressed as: ； ； wherein, for a planned trajectory wherein respectively denote the start and end points of the trajectory; trajectory points within the planned trajectory cannot exist within the restricted airspace any center trajectory all center trajectory points corresponding to the grid trajectory planning space form a set restricting the next trajectory point planning space; The turning angle decision boundary is that the aircraft cannot exceed the turning angle of the aircraft performance when turning in operation, and the turning angle decision boundary is expressed as: ； wherein, is the respective maximum turning angle of the aircraft, , , respectively represent the current leg length, the next leg length and the length from the current leg start point to the next leg end point. The speed decision boundary is that the aircraft cannot exceed the speed range in the corresponding control airspace when flying in each altitude layer, and the turning angle decision boundary is expressed as: ； ； ； These three formulas represent the aircraft's speed during level flight, ascent, and descent, respectively. Neither is allowed to exceed the speed limits of the current and destination altitude layers. , , These represent the altitude levels during cruise, climb, and descent, respectively. The corresponding maximum and minimum vacuum velocities.

9. The green real-time 4D flight trajectory planning method for aircraft in terminal area according to claim 1, characterized in that, The aircraft emission discharge amount is expressed by the following formula: ； wherein, represents the aviation emission discharge amount, represents the aviation emission discharge amount processed by the normalized extreme value method, represents the total category of aviation emissions; The normalized extreme value method is expressed by the following formula: ； in , , Pollutants The emissions, maximum emissions, and minimum emissions; ； wherein is the emission index of the pollutant representing the mass of the corresponding emission produced by the combustion of 1 kg of fuel, represents the fuel flow rate of the aircraft at the waypoint ; represents the time of flight of the aircraft within the waypoint ; ； wherein , , is the fuel flow rate for different parts of the BADA model for aircraft performance, represents the flight state at the waypoint , represents the aircraft configuration at the waypoint .

10. The method for green real-time planning of four-dimensional flight trajectories in an aircraft terminal area according to claim 9, characterized in that, The various types of aviation emissions include CO2, SO2, NO x , CO.

Images