A 4D trajectory planning method for aircraft taxiing based on spatiotemporal network

Abstract

The present invention provides a method for planning a 4D trajectory of aircraft taxiing on the ground based on a space-time network, including constructing an aircraft taxiing model; establishing an objective function and constraint conditions for the constructed model; solving the model using a dynamic programming method to obtain a planned 4D trajectory of aircraft taxiing on the ground. The present invention plans the taxiing path of the ground using a space-time network, plans the 4D trajectory of aircraft taxiing on the ground based on the time when the aircraft passes through the node, and takes conflict into account when planning the path. Example data shows that the method can effectively avoid conflict and plan the 4D trajectory of aircraft taxiing on the ground. The 4D trajectory planning method of the present invention takes into account the taxiing time and fuel consumption rate at the same time, and can optimize the operating costs of airlines.

Description

Technical Field

[0001] The invention belongs to the field of aviation technology, and in particular relates to a 4D trajectory planning method for aircraft taxiing based on a spatiotemporal network. Background Art

[0002] In China, Liu Fan et al. proposed methods to delay the push-out time of aircraft with lower priority, slow down, wait at the leading point, and re-plan a new taxiing path in terms of conflict resolution. The model was processed with the A* algorithm to complete the conflict avoidance and taxiing time reduction of the airport taxiing path planning within a certain period of time. Jiang Yu et al. discretized time and established a time step and space element to construct a space-time network that can efficiently and reasonably plan the expected path under the condition of avoiding conflicts, and used the Dijkstra algorithm to solve its model. In the comparison between the final optimization results and the GA model, its dynamic programming algorithm efficiency is dozens of times higher than that of the ordinary GA algorithm. Li Nan et al. used the aircraft taxiing speed profile that takes into account both taxiing time and taxiing fuel consumption. By applying different weights to taxiing time and fuel consumption, different Pareto optimal speed profiles were obtained to complete the path planning of aircraft considering turning factors on the airport surface. After obtaining preliminary taxi path results through the Dijkstra algorithm, Zhang Zhaoning and others adjusted the pushback time of different aircraft according to the taxiing complexity of the aircraft, and continued to iterate until the threshold of taxiing complexity was met. They further optimized fuel consumption by converting the conflict waiting time of departing aircraft into the waiting time before departure, but the delayed departure time of departing aircraft will affect airport traffic and passenger sentiment.

[0003] Abroad, Brown lee et al. used a rolling time window method including genetic algorithm to establish an aircraft taxiing model, and used the A* algorithm to find the shortest path for aircraft taxiing. They could allocate a reasonable taxiing path for the aircraft under the premise of avoiding aircraft taxiing conflicts, thus reducing the total taxiing time of the aircraft. Landry et al. designed a complex network airport surface model to solve the conflicts that may occur during aircraft taxiing. They could solve the conflict problem and optimize the path to a certain extent when the airport traffic is small and the airport surface path is simple. RAVIZZA et al., this paper took both aircraft fuel and taxiing time as considerations, and optimized the constructed model using the sequence diagram method. Li N et al. used the aircraft taxiing time as the solution function for model construction, listed the constraints with the conflict interval of the aircraft passing the node time, and solved the optimization through genetic algorithm, which alleviated the aviation conflict problem to a certain extent and reduced the aircraft surface taxiing time.

[0004] It is found from domestic and foreign research on taxiing path optimization that most scholars use static planning methods to plan aircraft paths. Although the goals are optimized, the traversal and verification of conflict nodes in static planning is too cumbersome. Studies have found that dynamic planning of space-time network models can greatly improve planning efficiency, and can timely respond to aircraft operating status and airport node status. It can help aircraft avoid tedious conflict detection during taxiing while optimizing fuel targets or time targets.

[0005] Regarding the modeling of taxiway path planning, the spatiotemporal network modeling method can clearly represent the operating status and availability of aircraft and airport path nodes. In terms of shortest path optimization, dynamic programming can effectively reduce the complexity of calculations and solve conflicts on taxiways by dynamically adjusting the aircraft's pre-selected path at each moment. Summary of the invention

[0006] In view of this, the present invention aims to propose a 4D trajectory planning method for aircraft taxiing based on spatiotemporal network.

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

[0008] A 4D trajectory planning method for aircraft taxiing based on a spatiotemporal network comprises the following steps:

[0009] Step 1: Construct an aircraft taxiing model;

[0010] Step 2: Establish the objective function and constraints for the model constructed in step 1;

[0011] Step 3: Use the dynamic programming method to solve the model and obtain the planned 4D trajectory of the aircraft taxiing on the ground.

[0012] The present invention also provides a 4D trajectory planning device for aircraft taxiing based on a spatiotemporal network, comprising:

[0013] A model building device for building an aircraft taxiing model;

[0014] A data processing device, for establishing an objective function and constraint conditions for the model constructed in the model construction device;

[0015] The solving device is used to solve the model using a dynamic programming method to obtain a planned 4D flight path of the aircraft taxiing on the ground.

[0016] The present invention also provides an electronic device, comprising

[0017] at least one processor, and

[0018] at least one memory in communication with the processor, wherein:

[0019] The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute any of the methods described above.

[0020] The present invention also provides a non-volatile computer-readable storage medium, which, when the computer-executable instructions are executed by one or more processors, enables the processors to execute any of the methods described above.

[0021] Compared with the prior art, the 4D trajectory planning method for aircraft taxiing based on a spatiotemporal network described in the present invention has the following advantages:

[0022] (1) The present invention considers both taxiing time and fuel consumption rate to perform 4D trajectory planning for aircraft taxiing, which can effectively reduce the operating costs of airlines;

[0023] (2) The method adopted by the present invention fully considers various conflicts of aircraft taxiing on the ground, and performs 4D flight path of aircraft taxiing on the ground on the premise of avoiding conflicts, which can ensure safe and conflict-free taxiing of aircraft on the ground. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0025] Figure 1 The aircraft turning path diagram of the present invention;

[0026] Figure 2 It is a schematic diagram of the adjacency matrix of the present invention;

[0027] Figure 3 The Xi'an Airport network diagram of the present invention;

[0028] Figure 4 The aircraft intersection conflict determination diagram of the present invention;

[0029] Figure 5 The aircraft head-on conflict determination diagram of the present invention;

[0030] Figure 6 A three-dimensional diagram of the aircraft taxi path of the present invention;

[0031] Figure 7 This is a schematic diagram of aircraft fuel and path according to the present invention. DETAILED DESCRIPTION

[0032] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0033] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and the like are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", and the like may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0034] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood by specific circumstances.

[0035] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0036] The present invention provides a 4D trajectory planning method for aircraft taxiing based on a spatiotemporal network, comprising the following steps:

[0037] Step 1: Construct an aircraft taxiing model;

[0038] Step 2: Establish the objective function and constraints for the model constructed in step 1;

[0039] Step 3: Use the dynamic programming method to solve the model and obtain the planned 4D trajectory of the aircraft taxiing on the ground.

[0040] Specifically, in step 1, before building the aircraft taxiing model, the following issues need to be noted as prerequisites for model building:

[0041] (1) The starting and ending points of each aircraft are known and will not change during the link.

[0042] (2) Aircraft can adjust their routes in a timely manner according to tower instructions.

[0043] (3) The aircraft that has taken off will be deleted from the airport's collection of all aircraft and will no longer participate in scheduling.

[0044] (4) All aircraft will be stopped immediately if a malfunction occurs, and can resume dispatching after repairs are completed. When an aircraft malfunctions, the route it is on will be inaccessible to other aircraft.

[0045] (5) Taxi dispatch only dispatches aircraft to the corresponding apron or runway end.

[0046] (6) Departing aircraft have priority over arriving aircraft.

[0047] (7) Aircraft of different models taxi at a speed of 5 m / s (this speed can also be set to other speed values ​​that meet the requirements), and the fuel consumption rates of different aircraft are different.

[0048] (8) A single taxiway can only support one aircraft at a time.

[0049] (9) Aircraft must meet wake turbulence separation requirements during taxiing.

[0050] Table 1 Minimum taxiing separation standards between different types of aircraft (m)

[0051]

[0052]

[0053] Spatiotemporal network discretization process:

[0054] The time axis is discretized at a distance of 1 second, and the time dimension is added to the two-dimensional airport network diagram. The discretized time axis is represented by T = {t0, t0 + ψ, ..., t0 + n*ψ}, where t represents a time node in the discretized time axis and ψ represents the time interval on the discrete time axis. In the model, ψ is 1 second.

[0055] Construct a set of spatiotemporal network nodes. When the first aircraft starts taxiing, a spatiotemporal point set E is established for each node of the airport. st (E i , t).

[0056] Construct the space-time network travel arc [i, j, t, t+n*ψ] of the aircraft, which means that the aircraft glides from point i to point j within Mψ time, where (i, j)∈E i

[0057] Construct the waiting arc of the aircraft taxiing space-time network, [i,i,t,t+n*ψ], which means that the aircraft stays at point i for N*ψ time, where N is a sufficiently large positive integer, which is set to 10000 in this paper.

[0058] Expand the attributes of the aircraft in three-dimensional space and time, and determine the aircraft status at each moment in real time - M.

[0059] That is, the nodes experienced by the aircraft become space-time arcs, changing from [i, j, t, t+n*ψ] to [M t ,M t+n*ψ ,t,t+n*ψ]. t It is the state attribute of the aircraft at time t, not limited to time and space attributes.

[0060] Specifically, in step 2, the objective function is:

[0061]

[0062]

[0063] t total(k) is the total taxiing time of an aircraft, Fuel total(k) is the fuel consumption of an aircraft during taxiing, and n is the number of aircraft participating in taxiing.

[0064] Aircraft surface taxi path planning means that each aircraft can find a path without conflict and with the least taxi time or the least taxi fuel in the existing path network of the airport.

[0065] Since the aircraft velocity model is a constant velocity, the aircraft fuel flow is a linear expression of the taxi time.

[0066] Fuel total(k) =t total(k) *Flow (3)

[0067] t total(k) =D k / V k (4)

[0068] V k is the speed of aircraft k, D k is the total taxiing distance of the aircraft, and Flow represents the fuel consumption per unit time;

[0069] Since the aircraft speed is constant, the total taxiing time of the aircraft is linearly related to the total taxiing distance of the aircraft. Therefore, without considering the conflict, the optimization objective can be equivalent to the taxiing distance, and the taxiing distance of aircraft k is:

[0070]

[0071] L ij is the taxiing distance of aircraft k on link (i, j), i.e., the length weight of link (i, j), d Δ Corrected length for the sliding distance in a turn.

[0072]

[0073] When the aircraft is at node V i When turning, R r =1, otherwise R r =0,d r For aircraft k at node V i Turning length, 2l r is the time when aircraft k passes through turning point V i When the starting turning point reaches V i , and from node V i The sum of the distances to the end of the turn.

[0074] d r =θ r *π*r / 180° (7)

[0075] l r =r*sin(θ r / 2) (8)

[0076] The turning radius r of each turning point is known, and the coordinates of the two nodes adjacent to the turning point are known. According to the cosine theorem, the turning angle, θ r Size

[0077]

[0078] Where: x w , x g , x u , respectively represent the node V w , V g , V u The horizontal axis, y w ,y g ,y u Respectively represent V w , V g , V u The vertical coordinate of .

[0079] For example, the aircraft passes through V u 、V g 、V w , where V g is the turning point, then the original aviation should take a right-angle route (V r , Vg , V f ) becomes an arc (V r , V f ), the actual distance traveled becomes shorter.

[0080] The right angle route is (V r , V g ) and (V g , V f ) are the sum of the straight-line distances of the two arcs. Since the two arcs are tangent, their lengths are the same, and the total length is recorded as 2l r , and the arc (V r , V f ) is denoted as d r .

[0081] And in In the function, the distance change caused by turning is not taken into account. Therefore, in this function, all routes are straight lines, and the only thing sliding on the turning section is (V r , V g ) and (V g , V f ), rather than the actual arc (V r , V f ).

[0082] And d r =2l r +(d r -2l r ), so the turning correction value of this section is (d r -2l r ).

[0083] Aircraft turning path diagram Figure 1 shown.

[0084] Constraints:

[0085] In the process of dynamic planning, since the aircraft has solved the overall path problem and divided it into multiple interconnected single-stage problems, in the dynamic algorithm, when the aircraft reaches the next point during the path retrieval process, the next path planning is made according to the real-time information of the airport network. Therefore, the constraints of static planning are converted into constraints in dynamic planning as follows:

[0086] The nodes that the aircraft passes through are blocked for a period of time, and the next selected node is pre-determined to be closed, so as to remove the existence of the link in the aircraft's search path, and then search for other links. When there are no other links, choose to wait until the node is unblocked.

[0087] Z ijk *Tik +t ijk <(T j(ban) -t secure )Z ijk (10)

[0088] Z ijk *T ik +t ijk >(T j(ban) +t secure )Z ijk (11)

[0089] T i =T i(ban) (12)

[0090] Z ij *Z ji =0 (13)

[0091]

[0092] Among them, when aircraft k pre-selects to go from node i to node j, Z ijk =1, otherwise Z ijk =0; T ik is the time when aircraft k arrives at node i; t ijk is the time required for aircraft k to taxi from node i to node j; T i(ban) is the blocking time of airport node i, which is determined by the arrival time of the aircraft at the node; t secure For safe separation between aircraft; Z ij represents whether there is an aircraft sliding from point i to point j on link (i, j) at the current moment. ij =1 means that there is an aircraft taxiing from point i to point j at the current moment; t is the current moment, and t=0 when the first aircraft starts taxiing; is the estimated taxi start time for aircraft k;

[0093] When either formula (10) or formula (11) is not true, the aircraft with lower priority needs to wait or change the pre-selected path. When formula (13) is not true, the aircraft with lower priority needs to change the path or wait. When formula (14) is true, it means that aircraft k starts to taxi.

[0094] Specifically, in step 3, the model is solved using a dynamic programming method, that is, the total problem of the state essence is dispersed into sub-problems. The advantage is that in the state transition model, the search complexity can be greatly reduced during the conflict search process. That is, in the dynamic programming of this model, the search range of the node state will be much smaller than the conflict node search range of the static planning, and the algorithm complexity will be greatly reduced. However, it has certain defects, that is, after the time is discretized, the final path selection result is a combination of a single sliding path selection scheme within multiple time steps. The refinement of the time step selection will directly lead to the accuracy of the dynamic programming. If the step is too long, although the operation time will be greatly reduced, it cannot meet the accuracy requirements of the path planning. If the time step is too small, it will affect the speed of the path dynamic planning. Therefore, the time step ψ=1 in the model of this paper can complete the path planning in a relatively short time and ensure that the path planning accuracy is sufficient.

[0095] When looking for the minimum taxiing time for an aircraft on an airport surface, if both the starting node and the ending node are known, it can be transformed into a problem of finding the shortest path. Since the link weights between adjacent nodes in the airport network graph have no negative values, the present invention uses the Dijkstra algorithm to solve it.

[0096] The dynamic programming algorithm process of the present invention is as follows:

[0097] Data initialization, read the aircraft's taxiing start time, start node and end node, and generate the shortest path plan from the current node to the taxiing end point for the aircraft that has already started taxiing through the Dijkstra algorithm.

[0098] Determine whether the current time is greater than or equal to the aircraft's starting taxiing time. If it is less than, the aircraft waits in place. If it is greater than or equal to, the aircraft begins to taxi along the minimum path generated by the Dijkstra algorithm. At each time step, the status of each node in the network is updated, as well as the distance from the aircraft to the node to be reached, to determine whether the aircraft has reached the next node. If not, continue taxiing.

[0099] This aircraft takes the current node as the starting node and the global ending node as the target, and uses the Dijkstra algorithm to solve the next target point expected to be reached by the aircraft.

[0100] By calculating the estimated time of arrival at the next node and determining whether there is a conflict with the status of the next node, if there is no conflict, then determine whether there is a head-on conflict. If there is no conflict and the next target point can be correctly solved, the aircraft enters the path from the current node to the next target point, continues taxiing, and sets a blocking time for this target node.

[0101] If the time to reach the next node conflicts with the state of the next node, conflict resolution is performed by changing the adjacency matrix to disconnect the link between the current node and the originally pre-selected node, and find a new shortest path. If the solution cannot be found, wait in place until the node is open and continue gliding.

[0102] (6) When arriving at the next position node, determine whether the current node is the global destination of the aircraft taxiing. If the current node is the global destination of the aircraft, remove this flight from the optimization list. If it is not the global destination, the aircraft continues to execute the third step. The algorithm ends when all aircraft have completed the scheduling to reach their respective global destinations.

[0103] In order to plan the taxiing path, it is necessary to convert the airport nodes and networks in the selected airport into a link node diagram. The present invention uses the example of Xi'an Airport for application analysis. When it is actually used in different airports, the link node diagram of the corresponding airport should also be drawn to carry out subsequent taxiing path planning research.

[0104] The node set and link set constitute the two-dimensional network graph G = (V, E) of the airport surface airside system.

[0105] V i (X i ,Y i )Node attributes include node number, node horizontal coordinate X, node vertical coordinate Y.

[0106] E i Link attributes include link number, link start node, link end node, and link weight (glide path length).

[0107] In a two-dimensional network diagram, not every node is connected by a link. By constructing a link matrix between points, it is easier to calculate and more intuitively obtain information on whether there is a link between nodes.

[0108] Assuming that the number of elements in {V} in the airport two-dimensional network diagram is n, we can represent it by establishing an n*n matrix C. The row and column elements C in the matrix are ij 0 and 1 represent different meanings, such as Figure 2 The node schematic diagram shown can be represented using an association matrix.

[0109] C ij =1 means there is a link between node i and node j; C ij = 0 means that there is no link connection between node i and node j (i≤m; j≤m). When i=j, C ij=0, indicating that the node itself has no connection.

[0110] According to this logic, the airport nodes and networks selected in the case are transformed into link node diagrams as shown in Figure 3 shown.

[0111] By obtaining the aircraft airport taxiing data of Xi'an Xianyang Airport, flight information was selected for analysis. Table 2 shows the path planning results of six aircraft.

[0112] Table 2 Aircraft taxi paths

[0113]

[0114]

[0115]

[0116] Since there are no conflict points in this dynamic path planning, a conflict simulation analysis is performed

[0117] like Figure 4 As shown, when aircraft B arrives at decision point j, it conducts a path search and the next node of the shortest path is i. However, since point i is the location node that aircraft A is about to arrive at, there is already a restricted time period set by aircraft A at point i. Aircraft B needs to calculate in advance whether the estimated arrival time at point i is within its restricted time.

[0118] If the estimated arrival time of aircraft B at point i is within the closed time period of point i, it is judged as a conflict event, and the element in row j and column i in the adjacency matrix is ​​set to 0, indicating that j is no longer on the way to i. Then the shortest path search is performed again to find a next location point for aircraft B that does not pass through node i and has no conflict.

[0119] If the aircraft is expected to arrive at node i at a time other than its closed time, then determine whether X ij Is it zero? If X ij = 0, it means that there is no aircraft on the link from node i to node j, so aircraft B can select point i as the next target node, and then set a forbidden time period for point i, leading X ji =1, aircraft B taxis to point i.

[0120] After a period of time, a Figure 5 In this case, aircraft A has already arrived at node i, and the next node of the shortest path generated by global path planning with i as the starting node is j. Even if the estimated arrival time of aircraft A at point j does not conflict with the blocking time of point j, X ji=1, indicating that an aircraft is taxiing from point j to point i, resulting in a head-on collision.

[0121] Therefore, point j is removed, and a node path that satisfies the closed time and does not have head-on conflicts is searched for aircraft A again. The next node (v) selected by it sets the closed time, and X is set to iv =1, aircraft A continues to taxi and the conflict is resolved.

[0122] A three-dimensional image of an aircraft taxi path Figure 6 As shown, the time axis Z axis is added to the two-dimensional graph. From the three-dimensional graph, it can be directly observed that the taxiing paths of the aircraft do not overlap on the time coordinate axis, and when the spatial coordinates are the same, the time and space coordinates of different aircraft differ, meeting their safety time interval. The third aircraft MU4405 and the fourth aircraft MU5569 both have taxiing paths from node 101 to node 100 in a relatively close time. During the taxiing process, the two aircraft have a time interval of 68 seconds when they arrive at node 101, which is much larger than the required safety time interval.

[0123] Finally, the conflict-free path dynamic planning was completed for the six aircraft. After the turning distance correction, the fuel consumption and taxiing distance are shown in Table 3 and Figure 7 shown.

[0124] Table 3 Aircraft fuel and path

[0125]

[0126] The total fuel consumption of the six aircraft is 564.29 kg. The total path planned for the six aircraft is 9115.66 m. Different aircraft have different fuel consumption rates. Figure 7 It can be seen that although the fuel consumption rates of different aircraft during taxiing are different, their final fuel consumption shows a trend that the greater the taxiing distance, the more fuel is consumed.

[0127] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A 4D trajectory planning method for aircraft taxiing based on a spatiotemporal network, characterized by: The steps include: Step 1: Construct an aircraft taxiing model; include: The time axis is discretized at a distance of 1 second, and the time dimension is added to the two-dimensional airport network diagram. The discretized time axis is represented by T = {t0, t0 + ψ, ..., t0 + n*ψ}, where t represents a time node in the discretized time axis, and ψ represents the time interval on the discrete time axis. In the model, ψ is 1 second. Construct a set of spatiotemporal network nodes. When the first aircraft starts taxiing, a spatiotemporal point set E is established for each node of the airport. st (E i , t); Construct the space-time network travel arc [i, j, t, t+n*ψ] of the aircraft, which means that the aircraft glides from point i to point j within Mψ time, where (i, j)∈E i ; Construct the waiting arc of the aircraft taxiing space-time network, [i,i,t,t+n*ψ], which means that the aircraft stayed at point i for N*ψ time, where N is a positive integer; Expand the attributes of aircraft in three-dimensional space and time, and determine the aircraft status at each moment in real time - M; That is, the nodes experienced by the aircraft become space-time arcs, changing from [i, j, t, t+n*ψ] to [M t ,M t+n*ψ ,t,t+n*ψ];M t is the state attribute of the aircraft at time t, not limited to time and space attributes; Step 2: For the model constructed in step 1, establish the objective function and constraints; the established objective function is t total(k) is the total taxiing time of an aircraft, Fuel total(k) is the fuel consumption of an aircraft during taxiing, and n is the number of aircraft participating in taxiing; Aircraft surface taxi path planning, that is, in the existing path network of the airport, each aircraft can find a path without conflict and with the least taxi time or the least taxi fuel; Since the aircraft velocity model is a constant velocity, the aircraft fuel flow rate is a linear expression of the taxiing time; Fuel total(k) =t total(k) *Flow t total(k) =D k / V k V k is the speed of aircraft k, D k is the total taxiing distance of the aircraft, and Flow represents the fuel consumption per unit time; Since the aircraft speed is constant, the total taxiing time of the aircraft is linearly related to the total taxiing distance of the aircraft. Without considering conflicts, the optimization objective can be equivalent to the taxiing distance, and the taxiing distance of aircraft k is: L ij is the taxiing distance of aircraft k on link (i, j), i.e., the length weight of link (i, j), d Δ Corrected length for the sliding distance in the turn; When the aircraft is at node V i When turning, R r =1, otherwise R r =0,d r For aircraft k at node V i Turning length, 2l r is the time when aircraft k passes through turning point V i When the starting turning point reaches V i , and from node V i The sum of the distances to the end of the turn; d r =θ r *π*r / 180° l r =r*sin(θ r / 2) The turning radius r of each turning point is known, and the coordinates of the two nodes adjacent to the turning point are known. According to the cosine theorem, θ can be determined. r The size of; the constraints are: Block the nodes that the aircraft passes through for a certain period of time, and pre-judge whether the next selected node is blocked, so as to remove the existence of the link in the aircraft's search path, and then search for other links. When there are no other links, wait until the node is unblocked; Z ijk *T ik +t ijk <(T j(ban) -t secure )Z ijk (1) Z ijk *T ik +t ijk >(T j(ban) +t secure )Z ijk (2) T i =T i(ban) (3) WITH ij *WITH ji =0(4) Among them, when aircraft k pre-selects to go from node i to node j, Z ijk =1, otherwise Z ijk =0; T ik is the time when aircraft k arrives at node i; t ijk is the time required for aircraft k to taxi from node i to node j; T i(ban) is the blocking time of airport node i, which is determined by the arrival time of the aircraft at the node; t secure For safe separation between aircraft; Z ij represents whether there is an aircraft sliding from point i to point j on link (i, j) at the current moment. ij =1 means that there is an aircraft taxiing from point i to point j at the current moment; t is the current moment, and t=0 when the first aircraft starts taxiing; is the estimated taxi start time for aircraft k; When either formula (1) or formula (2) is not true, the aircraft with lower priority needs to wait or change the pre-selected path. When formula (4) is not true, the aircraft with lower priority needs to change the path or wait. When formula (5) is true, it means that aircraft k starts to taxi. Step 3: Use the dynamic programming method to solve the model and obtain the planned 4D trajectory of the aircraft taxiing on the ground.

2. The method for 4D trajectory planning of aircraft taxiing based on a spatiotemporal network according to claim 1, characterized in that: In step 3, the specific steps of solving the model using the dynamic programming method are as follows: (1) Data initialization: reading the taxi start time, start node, and end node of the aircraft, and generating the shortest path from the current node to the taxi end point for the aircraft that has already started taxiing using the Dijkstra algorithm; (2) Determine whether the current time is greater than or equal to the aircraft's initial taxiing time. If it is less than the current time, the aircraft will wait in place. If it is greater than or equal to the current time, the aircraft will start taxiing according to the minimum path generated by the Dijkstra algorithm. At each time step, the status of each node in the network and the distance between the aircraft and the node to be reached will be updated to determine whether the aircraft has reached the next node. If not, the aircraft will continue taxiing. (3) The aircraft uses the current node as the starting node and the global ending node as the target, and uses the Dijkstra algorithm to solve the next target point that the aircraft is expected to reach; (4) Calculate the estimated time of arrival at the next node and determine whether there is a conflict with the state of the next node. If there is no conflict, determine whether there is a head-on conflict. If there is no conflict and the next target point can be correctly solved, the aircraft enters the path from the current node to the next target point, continues taxiing, and sets a blocking time for this target node; (5) If the time to reach the next node conflicts with the state of the next node, the conflict is resolved by changing the adjacency matrix to disconnect the link between the current node and the originally pre-selected node and find a new shortest path. If the solution cannot be found, the system chooses to wait in place until the node is open and continues to glide. (6) When arriving at the next position node, determine whether the current node is the global destination of the aircraft taxiing. If the current node is the global destination of the aircraft, remove this flight from the optimization list. If it is not the global destination, the aircraft continues to execute the third step until all aircraft have completed the scheduling to reach their respective global destinations and the algorithm ends.

3. A 4D trajectory planning device for aircraft taxiing based on a spatiotemporal network, characterized in that: include A model building device for building an aircraft taxiing model; comprising: The time axis is discretized at a distance of 1 second, and the time dimension is added to the two-dimensional airport network diagram. The discretized time axis is represented by T = {t0, t0 + ψ, ..., t0 + n*ψ}, where t represents a time node in the discretized time axis, and ψ represents the time interval on the discrete time axis. In the model, ψ is 1 second. Construct a set of spatiotemporal network nodes. When the first aircraft starts taxiing, a spatiotemporal point set E is established for each node of the airport. st (E i , t); Construct the space-time network travel arc [i, j, t, t+n*ψ] of the aircraft, which means that the aircraft glides from point i to point j within Mψ time, where (i, j)∈E i ; Construct the waiting arc of the aircraft taxiing space-time network, [i,i,t,t+n*ψ], which means that the aircraft stayed at point i for N*ψ time, where N is a positive integer; Expand the attributes of aircraft in three-dimensional space and time, and determine the aircraft status at each moment in real time - M; That is, the nodes experienced by the aircraft become space-time arcs, changing from [i, j, t, t+n*ψ] to [M t ,M t+n*ψ ,t,t+n*ψ];M t is the state attribute of the aircraft at time t, not limited to time and space attributes; The data processing device is used to establish an objective function and constraint conditions for the model constructed in the model construction device; the established objective function is t total(k) is the total taxiing time of an aircraft, Fuel total(k) is the fuel consumption of an aircraft during taxiing, and n is the number of aircraft participating in taxiing; Aircraft surface taxi path planning, that is, in the existing path network of the airport, each aircraft can find a path without conflict and with the least taxi time or the least taxi fuel; Since the aircraft velocity model is a constant velocity, the aircraft fuel flow rate is a linear expression of the taxiing time; Fuel total(k) =t total(k) *Flow t total(k) =D k / V k V k is the speed of aircraft k, D k is the total taxiing distance of the aircraft, and Flow represents the fuel consumption per unit time; Since the aircraft speed is constant, the total taxiing time of the aircraft is linearly related to the total taxiing distance of the aircraft. Without considering conflicts, the optimization objective can be equivalent to the taxiing distance, and the taxiing distance of aircraft k is: L ij is the taxiing distance of aircraft k on link (i, j), i.e., the length weight of link (i, j), d Δ Corrected length for the sliding distance in the turn; When the aircraft is at node V i When turning, R r =1, otherwise R r =0,d r For aircraft k at node V i Turning length, 2l r is the time when aircraft k passes through turning point V i When the starting turning point reaches V i , and from node V i The sum of the distances to the end of the turn; d r =θ r *π*r / 180° l r =r*sin(θ r / 2) The turning radius r of each turning point is known, and the coordinates of the two nodes adjacent to the turning point are known. According to the cosine theorem, θ can be determined. r The size of; the constraints are: Block the nodes that the aircraft passes through for a certain period of time, and pre-judge whether the next selected node is blocked, so as to remove the existence of the link in the aircraft's search path, and then search for other links. When there are no other links, wait until the node is unblocked; Z ijk *T ik +t ijk <(T j(ban) -t secure )Z ijk (1) Z ijk *T ik +t ijk >(T j(ban) +t secure )Z ijk (2) T i =T i(ban) (3) WITH ij *WITH ji =0(4) Among them, when aircraft k pre-selects to go from node i to node j, Z ijk =1, otherwise Z ijk =0; T ik is the time when aircraft k arrives at node i; t ijk is the time required for aircraft k to taxi from node i to node j; T i(ban) is the blocking time of airport node i, which is determined by the arrival time of the aircraft at the node; t secure For safe separation between aircraft; Z ij represents whether there is an aircraft sliding from point i to point j on link (i, j) at the current moment. ij =1 means that there is an aircraft taxiing from point i to point j at the current moment; t is the current moment, and t=0 when the first aircraft starts taxiing; is the estimated taxi start time for aircraft k; When either formula (1) or formula (2) is not true, the aircraft with lower priority needs to wait or change the pre-selected path. When formula (4) is not true, the aircraft with lower priority needs to change the path or wait. When formula (5) is true, it means that aircraft k starts to taxi. The solving device is used to solve the model using a dynamic programming method to obtain a planned 4D flight path of the aircraft taxiing on the ground.

4. An electronic device, characterized in that: include at least one processor, and at least one memory in communication with the processor, wherein: The memory stores program instructions executable by the processor, and the processor can execute the method according to any one of claims 1-2 by calling the program instructions.

5. A non-volatile computer-readable storage medium, when the computer-executable instructions are executed by one or more processors, causes the processors to perform the method according to any one of claims 1 to 2.

Images