An Optimization Method for the Mixed Operation of Human and Unmanned Airport Aircraft Tractors

Abstract

The present invention relates to an optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors, including: setting the operation trajectory of the manned tractor to be fixed and planning the operation trajectory of the unmanned tractor; using the A* algorithm to generate the global path planning of the unmanned tractor; performing local path planning for the unmanned tractor, considering three situations of maintaining a following state, performing lane-changing behavior, and competing for priority at intersections, and establishing a feedback mechanism for the relatively complex lane-changing behavior and competing-for-priority behavior; considering to avoid vehicle-aircraft conflict accidents on the airport runway, a decision-making judgment model before entering the runway area is proposed; the present invention directly avoids the influence of static obstacles on the vehicle operation environment; enables the mixed flow of unmanned and manned vehicles to reach the operation area efficiently and quickly without conflict, reduces the operation delay of the tractor within the flight time window, and ensures the safe operation of airport ground service vehicles.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle scheduling, and in particular to an optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors. Background Art

[0002] Aircraft tractors can push aircraft away from airport boarding bridges or aprons, push aircraft onto runways, and push aircraft away from maintenance sites, and can also serve as emergency towing vehicles in special cases. At present, most aircraft tractors complete towing operations under manual guidance. During the vehicle operation process, there is a possibility of human error and extremely high requirements for operators.

[0003] Currently, China is accelerating the development of vehicle-road collaborative systems and implementing unmanned driving technology at airports. With the development of vehicle-road collaborative technology and the construction of intelligent airports, unmanned tractors use advanced sensors and navigation technologies to monitor the surrounding environment in real time, effectively avoid human contact, solve the problem of human error, improve the safety of apron operations, and at the same time reduce airport operating costs. Therefore, it is a new trend for airports to replace manned ground service vehicles with unmanned vehicle fleets in the future. And in the next period of time, there will inevitably be scenarios of an airport traffic environment with mixed operation of unmanned and manned ground service vehicles. To ensure the efficient operation and conflict resolution of unmanned and manned ground service vehicles, it is crucial to consider the scheduling optimization method in the mixed operation environment of airport tractors.

[0004] The related existing technologies are: Khatib proposed an artificial potential field method, which converts the influence of target points and obstacles on the unmanned vehicle into an artificial potential field. The target point is located at a low potential energy and the obstacle is located at a high potential energy, that is, the target point provides gravity to the vehicle and the obstacle provides repulsion to the vehicle. The movement direction of the unmanned vehicle is consistent with the direction of the combined force of gravity and repulsion, but in a complex potential field environment, it is easy to fall into a local minimum point and the target is unreachable; Xiong Xiaoyong, considering the vehicle contour constraints and kinematic constraints in an unstructured environment, proposed a trajectory planning method based on an improved A* algorithm. In view of the lack of decision-making feedback mechanism and the lane change time and space optimization problems in a structured environment, a driving decision method based on lane line parameters was established. Policy feedback mechanism; Wang Hongchang summarized the multiple factors that affect the lane change of the self-driving semi-trailer tractor into a single influencing factor of relative acceleration, and obtained the lane change necessity index based on the relative acceleration index of the current lane and the front vehicle in the target lane by fuzzy reasoning; the lane change feasibility index was obtained by fuzzy reasoning based on this lane change necessity index and the relative acceleration index of the rear vehicle in the target lane, as the basis for whether to implement the lane change; finally, the effectiveness of the decision model was verified by using MATLAB fuzzy logic design application; Xiao Xiangjie regarded the conflict point as a resource and proposed the collision avoidance conditions of manned and unmanned vehicles at the intersection, clarifying that the arrival time of two vehicles with potential conflict points should be staggered and the minimum safe headway should be guaranteed. The above existing technologies take into account the trajectory planning of unmanned vehicles and the conflict resolution of unsignalized intersections, and provide a reference for guiding airport ground service vehicles to safely and quickly reach the apron for operations in mixed vehicle scenarios, reducing time costs and improving airport ground safety.

[0005] The above existing airport shuttle bus dispatching technology has the following defects:

[0006] (1) Most current studies focus on trajectory planning for unmanned vehicles in multiple scenarios, without considering the interference and randomness of manned vehicles on driving in mixed traffic scenarios;

[0007] (2) Some studies have explored the coordinated optimization scheduling of airport ground service vehicles and the number optimization scheduling of single-type vehicles, but have not considered the trajectory planning of vehicles in mixed traffic scenarios of unmanned vehicles and manned vehicles at airports;

[0008] (3) In scenarios where manned and unmanned vehicles coexist, most studies only consider the lane-changing behavior of unmanned vehicles, but lack research on vehicles as following vehicles and the lane-changing behavior of vehicles in areas where lane changes are allowed.

[0009] (4) Failure to properly guide the towing vehicle across the runway may result in a collision between the towing vehicle and the aircraft on the runway, leading to a dangerous accident. Summary of the invention

[0010] Therefore, the technical problem to be solved by the present invention is to overcome the interference and randomness of manned vehicles to driving in the prior art without considering the mixed driving scenario of driverless vehicles and manned vehicles; the trajectory planning of vehicles in the mixed driving scenario of driverless vehicles and manned vehicles at the airport is not considered; the research on vehicles as following vehicles and the research on the lane-changing behavior of vehicles in the permitted lane-changing area are insufficient; the tractor is not reasonably guided to cross the runway, which may cause the tractor and the aircraft on the runway to collide with each other and result in dangerous accidents.

[0011] To solve the above technical problems, the present invention provides an optimization method for mixed driving scheduling of manned and unmanned airport aircraft tractors, including:

[0012] Classify and number tractors of different models; according to the flight information, the aircraft model corresponding to the flight to be served can be known, so as to determine the tractor models that can be called for the flight to be served; select a tractor from the information of the tractors that can be called for the flight to be served to serve the flight;

[0013] Locate the tractor in real time within the specified area, and establish a communication connection with the tractor within this area by using the central controller; the tractors that can establish a communication connection are driverless tractors, and the tractors that cannot establish a communication connection are manned tractors;

[0014] Set the operation trajectory of the manned tractor to be fixed; use the A * algorithm to perform global path planning on all currently serving driverless tractors at the same time, and determine the driving routes of all currently serving driverless tractors;

[0015] Analyze all currently serving driverless tractors respectively; according to the surrounding traffic environment of the current driverless tractor, consider avoiding dynamic obstacles, and perform local path planning on the current driverless tractor to construct a vehicle conflict decision, including:

[0016] When considering vehicle following, when in the vehicle following scenario on normal roads, according to the vehicle following theory, add an artificial oscillation term to determine the safe following distance that needs to be maintained between vehicles; when in the scenario of signal-free intersections on airport roads, consider the diversion and confluence conflict situations to determine the safe following distance that needs to be maintained between vehicles;

[0017] When considering vehicle lane changes, the road at the airport intersection is divided into a driving area, a buffer area, and a restricted area to discuss lane change situations. It is stipulated that vehicles must successfully change lanes before the restricted area, and lane changes are not allowed within the restricted area. Considering the total number of lane changes, the total lane change delay, and driving safety factors, a lane change cost function is established. At the same time, considering the number of lanes crossed by the vehicle during lane change, the lane change time constraint, and the continuous lane change time constraint, and solving to obtain an executable optimal lane change decision;

[0018] When considering vehicle cutting in, in the scenario of an unsignalized intersection on the airport road, at this time, the area where the probability of conflict between two-way vehicle flows at the intersection is greater than the threshold is set as the conflict area. Starting from the moment the tractor enters the restricted area, it sends a passing demand to the central controller. The central controller calculates the time required for different vehicles in the two-way vehicle flows to travel from the boundary of the restricted area to the boundary of the conflict area at this time, and determines whether the time periods corresponding to different vehicles overlap. If they do not overlap, they drive normally; if they overlap, the central controller issues a cutting-in instruction to the vehicle closer to the conflict area;

[0019] The central controller issues a towing operation instruction to the driverless tractor, and the driverless tractor executes the corresponding instruction according to the planned path.

[0020] In an embodiment of the present invention, the tractors of different models are classified and numbered; according to the flight information, the aircraft model corresponding to the flight to be served can be known, so as to determine the tractor models that can be called for the flight to be served; selecting a tractor from the tractors that can be called for the flight to be served to serve the flight includes:

[0021] The tractor models can be divided into small tractors, medium tractors, and large tractors. They are classified and numbered according to the tractor models, that is, Q si 、Q mi 、Q li , i = 1, 2, 3... where s represents a small tractor, and Q si represents the set of corresponding small tractors; m represents a medium tractor, and Q mi represents the set of corresponding medium tractors; l represents a large tractor, and Q li represents the set of corresponding large tractors;

[0022] Tractors of different models can tow different maximum weights, and aircraft also have different models. Therefore, consider using an appropriate model of tractor to tow the corresponding model of aircraft; then according to the flight information, the aircraft model corresponding to the flight to be served can be known, so as to determine the tractor models that can be called for the flight to be served;

[0023] According to the tractor information, the historical parking duration of the aircraft, and the flight departure time, select a tractor from the tractors that can be called for the flight to be served to serve the flight;

[0024] The information of the tractor includes the tractor status, the fuel or power of the tractor, and the actual distance from the aircraft apron; the sum of the historical average operation times of other service vehicles; the tractor status includes an idle status and an operation status, where the tractors in the idle status include the tractors that have completed the last operation and are on the way back, and the tractors waiting for dispatch orders in the parking lot.

[0025] In an embodiment of the present invention, the tractor is positioned in real time within a specified area, and a communication connection is established between the central controller and the tractors within this area; the tractors that can establish a communication connection are driverless tractors, and the tractors that cannot establish a communication connection are manned tractors, including:

[0026] In the area from the tractor parking lot to the aircraft apron, a wireless sensor network is used to position the vehicles in real time; for area β from the service vehicle parking lot to the aircraft apron, the central controller establishes a communication connection with the vehicles within the area, where the tractors that can establish a communication connection are driverless tractors, and the tractors that cannot establish a communication connection are manned tractors; and according to the communication connection results, the driverless tractors and the manned tractors are numbered again, such as AV i 、RV i , i = 1, 2, 3...; Therefore, in the central controller, combined with the digital map, the position information of the two types of vehicles in the driving state can be obtained in real time, and then scheduling and trajectory tracking can be carried out.

[0027] In an embodiment of the present invention, the operation trajectory of the manned tractor is set to be fixed; the A * algorithm is used to perform global path planning on all the currently serving driverless tractors at the same time to determine the driving routes of all the currently serving driverless tractors, including:

[0028] Set the operation trajectory of the manned tractor to be fixed: For the manned tractor, it travels along the existing driving route. At the same time, an in-vehicle voice package is equipped for the manned vehicle to guide and correct the driving style of the driver to ensure that it travels in accordance with the specifications; where the existing driving route refers to the driving route of the existing manned tractors in the actual situation of the airport; in order to reduce costs, the driving route of the manned tractor remains unchanged during the driving process;

[0029] Use the A *The algorithm simultaneously performs global path planning on all the self-driving tractors in service, and determines the driving routes of all the self-driving tractors in service currently: divides the area β from the service vehicle parking lot to the aircraft apron into multiple grid cells of the same size, and assumes that during the path planning process, the positions and sizes of static obstacles are known and stationary; and assumes that for the planned vehicle driving path, each step occupies the entire grid; when the grid is occupied, it is considered that there is an obstacle in the current grid; when the grid is not occupied, it is considered that this grid is a free space without obstacles; uses the A * algorithm, and obtains the best path from the current point to the end point according to the lowest cost of different grids, that is, the minimum evaluation function value. Among them, since the idle tractors may be on the way back, the current point is not necessarily the starting point at this time;

[0030] A * algorithm's evaluation function is

[0031] f(n) = g(n) + h(n)

[0032] where, g(n) = (X n -X s ) 2 +(Y n -Y s ) 2 ; h(n) = (X t -X n ) 2 +(Y t -Y n ) 2 ; f(n) is the evaluation function of the current point; g(n) is the past cost function, used to evaluate the cost from the starting point to the current point; h(n) is the current cost function, used to evaluate the cost from the current point to the target node; at this time, the Euclidean distance is used for calculation; (X s ,Y s ) is the coordinate of the starting point P s , (X n ,Y n ) is the coordinate of the current point P n , (X t ,Y t ) is the coordinate of the target point P t .

[0033] In an embodiment of the present invention, when vehicle following, in the scenario of normal road vehicle following, according to the vehicle following theory, an artificial oscillation term is added to determine the safe following distance that needs to be maintained between vehicles; in the scenario of signal-free intersections on airport roads, considering the diversion and confluence conflicts, the safe following distance that needs to be maintained between vehicles is determined, including:

[0034] When in the normal road vehicle following scenario, to ensure the driving safety of vehicles, a certain safety distance must be maintained between vehicles when they are driving on the road. According to the following theory, the basic formula for vehicle following on normal road sections can be known. Considering the situation of mixed operation of manned and unmanned vehicles in airport tractors, an artificial oscillation term is added to the basic formula. Then the safe following distance that needs to be maintained between vehicles can be obtained as

[0035]

[0036] Among them, d0 represents the headway between two vehicles at time t; d1 represents the travel distance of the leading vehicle within the time t + T; d2 represents the travel distance of the current vehicle within the time t + T, where T = (v1 - v2) / a; v1 represents the speed of the leading vehicle; v2 represents the speed of the current vehicle; t represents the reaction time of the current vehicle; a represents the deceleration of the current vehicle; represents the average speed of the manned vehicle during driving, represents the average time of the manned vehicle during driving;

[0037] When in the scenario of an unsignalized intersection on the airport road, the vehicle maintains a following state, and situations such as diverging conflicts and merging conflicts will occur during the driving process. To ensure the driving safety of airport ground service vehicles, in the case of divergence and merging, the safe following distance that needs to be maintained between vehicles is

[0038]

[0039] Among them, x min represents the minimum safe distance between the following two vehicles, represents the arc length when the vehicle turns and leaves the lane. It can be known from vehicle dynamics theory that

[0040]

[0041] Among them, r represents the turning radius of the vehicle, and h represents the lane width.

[0042] In an embodiment of the present invention, when considering vehicle lane change, the road at the airport intersection is divided into a driving area, a buffer area, and a restricted area to discuss the lane change situation. It is stipulated that the vehicle must successfully change lanes before the restricted area, and lane change is not allowed within the restricted area; considering the total number of lane changes, the total lane change delay, and the driving safety factors, a lane change cost function is established; at the same time, considering the number of lanes crossed by the vehicle during lane change, the lane change time constraint, and the continuous lane change time constraint, and solving to obtain an executable optimal lane change decision, including:

[0043] Considering the driving safety of ground service vehicles, the roads near intersections at the airport are divided into different areas, namely the free driving area, the buffer zone, and the restricted area. It is stipulated that the vehicle must successfully change lanes before the restricted area, and lane changes are not allowed within the restricted area. The following is a discussion of the lane change situation:

[0044] If there is no vehicle in front of the current driving lane of the vehicle, or the distance from the vehicle in front is much greater than the minimum safety distance d safe (d safe = x), it indicates that the current driving lane of the vehicle is idle, and the vehicle remains in the current lane.

[0045] If there is a vehicle driving in the front of the current driving lane of the vehicle at a short distance, and there is no vehicle in front of the adjacent lane, the driverless vehicle can freely change lanes to the adjacent idle lane.

[0046] If there are vehicles in both the front and adjacent lanes, first judge the size relationship between the distance d between the driving vehicle and the vehicle in front of the current lane and d safe , and then judge the size relationship between the distance between the driving vehicle and the vehicle in front and behind in the adjacent lane and the corresponding minimum safety distance, that is, d _a and d safe_a , d _b and d safe_b ; when d ≥ d safe , the vehicle still remains in the current lane; when d ≤ d afe and d _a ≥ d safe_a , d _b ≥ d safe_b , d ≤ d _a , the vehicle must execute the lane change command to avoid collisions; at this time, if both the adjacent left and right lanes meet this condition, judge the size of the lane change cost function and execute the optimal lane change decision; otherwise, the vehicle cannot execute the lane change command, and the current vehicle reduces its speed and continues to maintain a following state.

[0047] If the vehicle has a steering requirement in the restricted area, it must execute the lane change decision.

[0048] Considering the total number of lane changes, the total lane change delay, and the driving safety factors, different weights are assigned to them, and a lane change cost function is established. The minimum value is obtained, which is expressed as

[0049]

[0050] Among them, α1, α2, α3 represent different weight coefficients, which can take different values in different situations, and α1 + α2 + α3 = 1; N i represents the total number of lane changes of the i-th vehicle; t i represents the lane change time of the i-th vehicle, and L irepresents the lane-changing distance of the \(i\)-th vehicle; according to the airport high-precision map and GPS positioning technology, \(t\) i -L i / V represents the lane-changing delay time of the \(i\)-th vehicle; \(y\) i0 represents the ordinate of the lane-changing start point of the \(i\)-th vehicle; \(y\) i1 represents the ordinate of the vehicle in the front on the current lane at the start of lane-changing; \(y\) i2 represents the ordinate of the lane-changing end point of the \(i\)-th vehicle; Therefore, the distance \(d\) between the vehicle and the vehicle in the front on the current lane is \(d = y\) i1 -y i0 -L;

[0051] To ensure the stability and safety of vehicle driving, the following constraints are imposed on the lane-changing cost function:

[0052] (1) It is stipulated that when a vehicle performs a lane-changing behavior, it can only change lanes to an adjacent lane in sequence, that is, \(l\) i ≤1, where \(l\) i represents the number of lanes that the \(i\)-th vehicle is allowed to cross at one time

[0053] (2) Lane-changing time constraint; A vehicle cannot make a lane-changing decision arbitrarily, so it is stipulated that the vehicle must complete the lane change within the allowed maximum lane-changing length, that is, the lane-changing time of the vehicle must be less than the specified maximum allowed lane-changing time, that is

[0054]

[0055] where \(T\) max represents the specified maximum allowed lane-changing time, \(L\) max represents the allowed maximum lane-changing length; \(V\) represents the current vehicle speed;

[0056] (3) Consecutive lane-changing time constraint; To meet the vehicle driving requirements, the vehicle is allowed to change lanes continuously for multiple times, and it is stipulated that the lane-changing time of the vehicle satisfies

[0057]

[0058] where represents the time point when the \(i\)-th vehicle finishes lane-changing, represents the time point when the \(i\)-th vehicle finishes the next lane-changing, \(T\) min represents the allowed minimum consecutive lane-changing time;

[0059] According to the constraint conditions, find the minimum value of the lane-changing cost function.

[0060] In an embodiment of the present invention, when considering vehicle cutting in, in the scenario of an unsignalized intersection on an airport road, the area where the probability of conflict between two-way vehicle flows at the intersection is greater than a threshold is set as a conflict area. Starting from the moment when the tractor enters the restricted area, a passing demand is sent to the central controller. The central controller calculates the time required for different vehicles in the two-way vehicle flows to travel from the boundary of the restricted area to the boundary of the conflict area at this time, and determines whether the time periods corresponding to different vehicles overlap. If they do not overlap, the vehicles travel normally; if they overlap, the central controller issues a cutting-in instruction to the vehicle closer to the conflict area, including:

[0061] The area where the probability of conflict between two-way vehicle flows at the intersection is greater than a threshold is set as a conflict area; when the lane change of the towing vehicle is completed and it enters the restricted area, starting from that moment, a passing demand is sent to the central controller. The central controller determines whether the time period Δt from the ordinate of the boundary point of the restricted area to the ordinate of the boundary point of the conflict area plus the length of one vehicle body of the conflict area will overlap by calculating the time required for different vehicles in the two-way vehicle flows to travel from the boundary of the restricted area to the conflict area at this time, that is, to determine whether Δt m =(t m ,t' m ) overlaps with Δt n =(t n ,t' n );

[0062] where t' m =t m +(w m +L) / V, t' n =t n +(w n +L) / V; t m represents the moment when the m-th vehicle arrives at the boundary of the restricted area; t' m represents the moment when the i-th vehicle arrives at the boundary of the conflict area; t n represents the moment when the n-th vehicle arrives at the boundary of the restricted area; t' n represents the moment when the n-th vehicle arrives at the boundary of the conflict area; w m represents the ordinate difference between the boundary point of the restricted area and the boundary point of the conflict area of the m-th vehicle; w n represents the ordinate difference between the boundary point of the restricted area and the boundary point of the conflict area of the n-th vehicle; the positions of the m-th vehicle and the n-th vehicle are in different directions at the intersection. If Δt does not overlap, the two-way vehicle flows travel normally; if Δt overlaps, the central controller issues an acceleration cutting-in instruction to the driverless vehicle closer to the conflict point, that is, issues an acceleration cutting-in instruction to the driverless vehicle corresponding to min{w m ,w n}; the driverless vehicle accelerates to the maximum speed limit V max at the maximum acceleration a maxAfter driving at a constant speed through the restricted section, the driverless vehicle can pass through the intersection first.

[0063] In an embodiment of the present invention, for relatively complex lane-changing behaviors and cutting-in behaviors, considering sudden changes in traffic conditions, a feedback mechanism is established, including:

[0064] During the execution of the decision-making instruction, considering that the traffic environment suddenly changes and poses a threat to vehicle safety, a decision feedback mechanism is established, and a 0-1 variable f is set to facilitate timely interruption of the instruction;

[0065] Among them, when f = 0, it means that the decision is being executed and other instructions are not executed. At this time, the current instruction is in an interruptible state; when f = 1, it means that the decision has been completed or aborted, and new instructions can be continued to be executed. At this time, the current instruction ends and cannot be interrupted and fed back;

[0066] If the decision is in a following state, the execution state is set to completed at this time, that is, f = 1;

[0067] If the decision-making instruction is a lane-changing instruction at this time, combined with the high-precision map and GPS, judge the change in the distance Δl between the vehicle and the lane line, and the moment t when the distance between the vehicle and the lane line is 0 h ; According to Δl and t h Judge whether the vehicle has completed lane-changing; before the distance is getting smaller but still not reaching 0, that is, before the moment t h A lane-changing interruption instruction can be issued before the moment, otherwise the instruction cannot be interrupted;

[0068] If the decision-making instruction is an accelerating cutting-in instruction at this time, combined with the high-precision map and GPS, when the vehicle is in the restricted area section, obtain the driving time t within any Δt time period a ; If t a is less than the vehicle's constant-speed driving time, it indicates that the vehicle is accelerating. At this time, a cutting-in interruption instruction can be issued, that is, f = 0 can be set, and then continue to judge the next cycle instruction, otherwise the instruction cannot be interrupted.

[0069] In an embodiment of the present invention, to avoid vehicle-machine conflict accidents in the airport runway, a binary state variable is set, and a decision-making judgment model before entering the runway area is established, so as to judge the state of the tractor crossing the aircraft runway until the aircraft safely reaches the apron for towing operations, including:

[0070] Before the towing vehicle reaches the airport runway, all vehicles receive a suspension waiting instruction. The state of the vehicle before entering the runway is represented by a binary state variable g. Then, when the vehicle receives the suspension instruction and suspends waiting, g = 0; when the vehicle continues to pass, g = 1;

[0071] If the central controller does not receive the instruction that the aircraft passes the runway, it issues a passing instruction to the towing vehicle, and at this time, g = 1; if it receives the instruction that the aircraft is passing the runway, the vehicle continues to pause and wait, that is, g = 0. Until the aircraft leaves the runway, it issues a passing instruction to the towing vehicle, and the vehicle can continue to pass, and at this time, g = 1;

[0072] If it receives the instruction that the aircraft is about to pass the runway, according to the time t1 = la / V for the towing vehicle to pass the runway and the time t2 when the aircraft enters the runway, it further judges whether the towing vehicle can pass the runway first, where la represents the longitudinal distance for the towing vehicle to cross the runway, and V represents the driving speed of the towing vehicle;

[0073] If t0 + t1 < t2, where t0 represents the time for the towing vehicle to reach the pause area, it means that the towing vehicle can safely pass the runway before the aircraft enters the runway. To improve the safety and efficiency of the operation, a passing instruction can be issued at this time, that is, g = 1, otherwise g = 0; Finally, the towing vehicle reaches the corresponding aircraft apron, and the task ends.

[0074] The present invention also provides an optimized dispatching device for mixed operation of manned and unmanned airport aircraft towing vehicles, including:

[0075] A memory for storing a computer program;

[0076] A processor for implementing the steps of an optimized dispatching method for mixed operation of manned and unmanned airport aircraft towing vehicles as described in any one of claims 1 to 9 when executing the computer program.

[0077] The above technical solution of the present invention has the following advantages compared with the prior art:

[0078] (1) The optimized dispatching method for mixed operation of manned and unmanned airport aircraft towing vehicles according to the present invention fully considers the airport ground environment where unmanned towing vehicles and manned towing vehicles are mixed in the airport, which is closer to the reality. By means of the vehicle-road coordination system and autonomous driving technology, it weakens the influence of human factors in airport ground service operations, improves the safety, efficiency and digital nature of airport ground service operations, ensures airport ground safety, and lays a foundation for the construction of a digital intelligent airport;

[0079] (2) The optimized dispatching method for mixed operation of manned and unmanned airport aircraft towing vehicles according to the present invention takes into account the process of the towing operation of the aircraft towing vehicle, considers selecting the towing vehicle with the best current state for each step of the process until the towing vehicle safely and quickly reaches the apron, and proposes a specific optimized dispatching scheme for manned and unmanned airport towing vehicles. Thereby, it improves the operation efficiency of the towing vehicle, reduces its operation cost, reduces fuel or power consumption, improves economy, and improves the safety of airport ground services, providing a reference basis for the construction of a smart airport and the mixed operation service of towing vehicles;

[0080] (3) The method for optimizing the mixed operation scheduling of manned and unmanned airport aircraft tractors according to the present invention comprehensively considers the path planning problem of unmanned tractors; uses grid networking combined with the traditional A* algorithm to generate the global path of unmanned vehicles, ensuring driving safety; considering the influence of dynamic obstacles such as pedestrians and other vehicles during vehicle driving, a dynamic obstacle avoidance strategy is formulated, that is, a local path is planned, which is divided into three possible obstacle avoidance behaviors: when maintaining a following state, the following characteristics of normal road surfaces and the diversion and confluence at intersections are considered, as well as the corresponding safe following distance; before executing a lane change command, the necessity of lane change is determined according to the actual distance and the safe distance between vehicles, and a lane change cost function and constraints on lane change safety are established, so that the lane change cost function is minimized while the lane change task is completed; before executing an intersection cutting-in command, by determining the longitudinal distance between vehicles in two directions at the intersection and the conflict area, it is further determined whether the cutting-in command can be executed, reducing behaviors with greater risks; in addition, the present invention establishes an instruction feedback mechanism, which can interrupt the lane change and cutting-in commands in case of emergency, thus ensuring the stability and safety of vehicle driving and reducing traffic flow oscillation;

[0081] (4) The method for optimizing the mixed operation scheduling of manned and unmanned airport aircraft tractors according to the present invention considers that when the aircraft tractor enters the runway, there may be a vehicle-aircraft conflict with the aircraft taxiing on the runway, and a decision-making method for the tractor to cross the aircraft runway is proposed; the state of the tractor crossing the aircraft runway is determined according to the value of the binary variable until the aircraft safely reaches the parking apron for towing operation, guiding the safe and efficient towing operation of unmanned and manned tractors and improving the airport ground operation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] In order to make the content of the present invention easier to be clearly understood, the following further details the present invention according to the specific embodiments of the present invention and in combination with the drawings, where

[0083] Figure 1 is the flowchart of the method for optimizing the mixed operation scheduling of manned and unmanned airport aircraft tractors provided by the present invention;

[0084] Figure 2 is the schematic diagram of vehicle lane change provided by the present invention;

[0085] Figure 3 is the schematic diagram of the global driving route generated by the A* algorithm provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0086] The following further illustrates the present invention in combination with the drawings and specific embodiments, so that those skilled in the art can better understand the present invention and be able to implement it, but the embodiments cited are not used as a limitation to the present invention.

[0087] The preset conditions for the path planning of unmanned and manned tractor-trailers at the airport include: assuming that flight information is known and the dispatcher starts dispatching ground service vehicles in advance; assuming that tractors of the same model have the same configuration, service time, and initial driving speed; assuming that only one tractor serves one flight and no service will be provided when vehicle resources are insufficient; assuming that after vehicle service is completed, the vehicle can continue to choose to return to the parking lot or continue to provide service; assuming that a tractor can serve a flight only when the tractor model matches the aircraft model of the flight;

[0088] Embodiment 1

[0089] Referring to Figure 1 as shown Figure 1 is a flowchart of any unmanned mixed operation scheduling optimization method for airport aircraft tractors; the specific operation steps are as follows:

[0090] Step 1: Classification of tractor models;

[0091] The models of tractors can be divided into small tractors, medium tractors, and large tractors. They are classified and numbered according to the tractor models, that is, Q si , Q mi , Q li , i = 1, 2, 3... where s represents small tractors, and Q si represents the set of corresponding small tractors; m represents medium tractors, and Q mi represents the set of corresponding medium tractors; l represents large tractors, and Q li represents the set of corresponding large tractors;

[0092] Tractors of different models can tow different maximum weights, and aircraft also have different models. Therefore, consider using tractors of appropriate models to tow aircraft of corresponding models; then, according to the flight information, the aircraft model corresponding to the flight to be served can be known, and thus the category of tractor models that can be dispatched for the flight to be served can be determined;

[0093] Step 2: Determine the tractor;

[0094] According to the information of the tractor, the historical parking duration of the aircraft, and the flight departure time, select a tractor from the tractors that can be called for the flight to be served to serve the flight; the information of the tractor includes the tractor status, the fuel or power of the tractor, the actual distance from the aircraft apron, and the sum of the historical average operation times of other service vehicles; the tractor status includes the idle state and the operation state. Among them, tractors in the idle state include tractors that completed the previous operation and are on the way back, and tractors waiting for dispatch orders in the parking lot;

[0095] Step 3: Vehicle driving type identification;

[0096] In the area from the tractor parking lot to the aircraft apron, a wireless sensor network is used to locate the vehicle in real time; for area β from the service vehicle parking lot to the aircraft apron, the central controller establishes a communication connection with the vehicles in the area. Among them, the vehicles that can establish a communication connection are driverless tractors, and the vehicles that cannot establish a communication connection are manned tractors; and according to the communication connection results, the driverless tractors and manned tractors are numbered again, such as AV i 、RV i , i = 1, 2, 3...; Therefore, in the central controller, combined with the digital map, the position information of the two types of vehicles in motion can be obtained in real time, and then scheduling and trajectory tracking can be carried out;

[0097] Step 4: Vehicle path generation;

[0098] For manned tractors, they drive according to the existing driving route. At the same time, an in-vehicle voice package is equipped for manned vehicles to guide and correct the driving style of the driver to ensure that they drive according to the specifications; among them, the existing driving route refers to the driving route of the existing manned tractors in the actual situation of the airport; in order to reduce fuel or power consumption and improve economy, the route of the manned tractor remains unchanged during driving;

[0099] For the driving route of driverless tractors, Algorithm A * is used for global path planning: that is, area β from the service vehicle parking lot to the aircraft apron is divided into multiple grid cells of the same size, and it is assumed that during the path planning process, the position and size of static obstacles are known and stationary; and it is assumed that the driving path of the planned vehicle occupies the entire grid in each step; when the grid is occupied, it is considered that there is an obstacle in the current grid; when the grid is not occupied, it is considered that this grid is a free space without obstacles; using Algorithm A * , according to the lowest cost of different grids, that is, the minimum evaluation function value, the best path from the current point to the end point is obtained. Among them, since the tractor in the idle state may be on the way back, the current point is not necessarily the starting point at this time;

[0100] A * The evaluation function of the algorithm is

[0101] f(n) = g(n) + h(n)

[0102] where g(n) = (X n - X s ) 2 + (Y n - Y s ) 2 ; h(n) = (X t - Xn ) 2 +(Y t -Y n ) 2 ; f(n) is the evaluation function of the current point; g(n) is the past cost function, used to evaluate the cost from the starting point to the current point; h(n) is the current cost function, used to evaluate the cost from the current point to the target node; at this time, the Euclidean distance is used for calculation; (X s ,Y s ) is the coordinate of the starting point P s , (X n ,Y n ) is the coordinate of the current point P n , (X t ,Y t ) is the coordinate of the target point P t ;

[0103] Step 5: Construction of vehicle conflict decision-making;

[0104] Analyze all the currently serving driverless tractors respectively; when different vehicles performing towing tasks are at the same time node, if the driving trajectory of the current vehicle conflicts with or coincides with that of other vehicles, the point where the conflict or trajectory coincidence occurs is taken as the conflict node at this time; then, based on the running speed of the current vehicle, the distance between the target vehicle and the conflict node, the length of the target vehicle, and the sensors of the driverless vehicle itself, determine whether the driverless vehicle needs to re-plan the route, so as to perform local path planning for the target driverless tractor and construct a vehicle conflict decision-making, including:

[0105] (1) Vehicle following

[0106] 1) When in the scenario of vehicle following on normal roads, to ensure the driving safety of vehicles, when vehicles are driving on the road, a certain safety distance should be maintained between vehicles. According to the following theory, the basic formula for vehicle following on normal road sections can be known. Considering the situation of mixed operation of manned and driverless vehicles for airport tractors, an artificial oscillation term is added to the basic formula Then the required safe following distance between vehicles can be obtained as

[0107]

[0108] where, d0 represents the headway between two vehicles at time t; d1 represents the traveling distance of the leading vehicle within time t + T; d2 represents the traveling distance of the current vehicle within time t + T, where T = (v1 - v2) / a; v1 represents the speed of the leading vehicle; v2 represents the speed of the current vehicle; t represents the reaction time of the current vehicle; a represents the deceleration of the current vehicle; represents the average speed of the manned vehicle during driving, Represents the average driving time of a manually driven vehicle;

[0109] 2) When in the scenario of an intersection without signals on the airport road, the vehicle maintains a following state, and there will be situations of diversion conflicts and merging conflicts during the vehicle's driving. To ensure the driving safety of airport ground service vehicles, in the case of diversion and merging, the safe following distance that needs to be maintained between vehicles is

[0110]

[0111] where x min represents the minimum safe distance between the following vehicles, represents the arc length when the vehicle turns and leaves the lane. It can be known from vehicle dynamics theory that

[0112]

[0113] where r represents the turning radius of the vehicle and h represents the lane width;

[0114] (2) Vehicle lane change

[0115] Considering the driving safety of ground service vehicles, the road near the intersection of the airport is divided into different areas, namely the free driving area, the buffer area, and the restricted area. It is stipulated that the vehicle must successfully change lanes before the restricted area, and lane change is not allowed within the restricted area; discuss the lane change situation:

[0116] If there is no vehicle driving in front of the current driving lane of the vehicle, or the distance from the vehicle in front is much greater than the minimum safe distance d safe (d safe = x), it indicates that the current driving lane of the vehicle is idle, and the vehicle maintains driving in the current lane;

[0117] If there is a vehicle driving in the front of the current driving lane of the vehicle at a short distance, and there is no vehicle driving in front of the adjacent lane, the driverless vehicle can freely change lanes to the adjacent idle lane;

[0118] If there are vehicles in both the front and adjacent lanes, first judge the size relationship between the distance d between the driving vehicle and the vehicle in front of the current lane and d safe , and secondly judge the size relationship between the distances between the driving vehicle and the vehicle in front and behind in the adjacent lane and the corresponding minimum safe distances, that is, d _a and d safe_a , d _b and d safe_b ; when d ≥ d safe , the vehicle still maintains driving in the current lane; when d ≤ d safe and d _a ≥ d safe_a , d _b ≥ dsafe_b When d ≤ d _a the vehicle must execute the lane - changing instruction to avoid collisions. At this time, if both the adjacent left and right lanes meet this condition, then the size of the lane - changing cost function is judged, and the optimal lane - changing decision is executed; otherwise, the vehicle cannot execute the lane - changing instruction, and the current vehicle reduces its speed and continues to maintain a following state;

[0119] When the vehicle has a steering demand in the restricted area, it must execute the lane - changing decision;

[0120] Considering the total number of lane - changes, the total lane - changing delay, and the driving safety factors, different weights are assigned to them, a lane - changing cost function is established, and the minimum value is obtained, which is expressed as

[0121]

[0122] where α1, α2, α3 represent different weight coefficients, and different values can be taken in different situations, and α1 + α2 + α3 = 1; N i represents the total number of lane - changes of the i - th vehicle; t i represents the lane - changing time of the i - th vehicle, L i represents the lane - changing distance of the i - th vehicle; According to the airport high - precision map and GPS positioning technology, t i -L i / V represents the lane - changing delay time of the i - th vehicle, y i0 represents the ordinate of the starting point of the i - th vehicle's lane - change, y i1 represents the ordinate of the vehicle in the front of the current lane at the start of the lane - change, y i2 represents the ordinate of the end point of the i - th vehicle's lane - change; Therefore, the distance d between the vehicle and the vehicle in the front of the current lane is d = y i1 -y i0 -L;

[0123] To ensure the stability and safety of vehicle driving, the following constraints are imposed on the lane - changing cost function:

[0124] 1) It is stipulated that when the vehicle performs a lane - changing behavior, it can only change lanes to the adjacent lane successively, that is, l i ≤1, where l i represents the number of lanes that the i - th vehicle is allowed to cross at one time

[0125] 2) Lane - changing time constraint; The vehicle cannot execute the lane - changing decision arbitrarily, so it is stipulated that the vehicle must complete the lane - change within the allowed maximum lane - changing length, that is, the lane - changing time of the vehicle must be less than the specified maximum allowed lane - changing time, that is

[0126]

[0127] where T maxRepresents the maximum lane-changing time allowed by regulations, L max Represents the maximum lane-changing length allowed; V represents the current vehicle driving speed;

[0128] 3) Continuous lane-changing time constraint; To meet the vehicle driving requirements, the vehicle is allowed to change lanes continuously multiple times, and it is stipulated that the vehicle lane-changing time satisfies

[0129]

[0130] Among them, Represents the time point when the i-th vehicle finishes lane-changing, Represents the time point when the i-th vehicle finishes the next lane-changing, T min Represents the minimum continuous lane-changing time allowed;

[0131] According to the constraint conditions, find the minimum value of the lane-changing cost function; Refer to Figure 2

[0132] (3) Vehicle cutting in

[0133] When considering vehicle cutting in, in the scenario of an unsignalized intersection on the airport road, the area where the probability of conflict between two-way vehicle flows at the intersection is greater than the threshold is set as the conflict area; When the towing vehicle finishes lane-changing and starts to enter the restricted area, it sends a passing demand to the central controller. The central controller calculates the time required for different vehicles in the two-way vehicle flow to travel from the boundary of the restricted area to the conflict area at this time, and judges whether the time period Δt from the ordinate of the boundary point of the restricted area to the ordinate of the boundary point of the conflict area plus the length of one vehicle body will overlap, that is, judge whether Δt m =(t m ,t' m ) overlaps with Δt n =(t n ,t' n );

[0134] Among them, t' m =t m +(w m +L) / V, t' n =t n +(w n +L) / V; t m Represents the moment when the m-th vehicle arrives at the boundary of the restricted area; t' m Represents the moment when the m-th vehicle arrives at the boundary of the conflict area; t n Represents the moment when the n-th vehicle arrives at the boundary of the restricted area; t' n Represents the moment when the n-th vehicle arrives at the boundary of the conflict area; X m Represents the ordinate difference between the boundary point of the restricted area of the m-th vehicle and the boundary point of the conflict area; X nDenote the vertical coordinate difference between the boundary point of the restricted area of the nth vehicle and the boundary point of the conflict area; the positions of the mth vehicle and the nth vehicle are in different directions at the intersection. If Δt does not overlap, the two-way traffic flows normally; if Δt overlaps, the central controller issues an acceleration and cut-in instruction to the driverless vehicle closer to the conflict point, that is, to the driverless vehicle corresponding to min{w m , w n}, and the driverless vehicle accelerates to the maximum acceleration a max and then travels at a constant speed through the restricted section. At this time, the driverless vehicle can pass through the intersection first. max

[0135] (4) Feedback mechanism

[0136] During the execution of the decision-making instruction, considering that the traffic environment suddenly changes and threatens the safety of the vehicle, a decision-making feedback mechanism is established, and a 0-1 variable f is set to facilitate interrupting the instruction in a timely manner;

[0137] Among them, when f = 0, it means that the decision-making is being executed and other instructions are not executed. At this time, the current instruction is in an interruptible state; when f = 1, it means that the decision-making has been completed or aborted, and new instructions can be continued to be executed. At this time, the current instruction ends and cannot be interrupted and feedback;

[0138] Then when the decision-making is in a following state, the execution state is set to completed at this time, that is, f = 1;

[0139] If the decision-making instruction is a lane-changing instruction at this time, combined with the high-precision map and GPS, the change in the size of the distance Δl between the vehicle and the lane line and the moment t when the distance between the vehicle and the lane line is 0 are obtained h ; According to Δl and t h to judge whether the vehicle lane change is completed; before the distance is getting smaller but still not reaching 0, that is, before the moment of t h , a lane-changing interruption instruction can be issued, otherwise the instruction cannot be interrupted;

[0140] If the decision-making instruction is an acceleration and cut-in instruction at this time, combined with the high-precision map and GPS, when the vehicle is in the restricted area section, the driving time t within any Δt time period is obtained a ; If t a is less than the vehicle's constant-speed driving time, it means that the vehicle is accelerating. At this time, a cut-in interruption instruction can be issued, that is, f = 0 can be set, and then the next cycle instruction can be continued to be judged, otherwise the instruction cannot be interrupted;

[0141] Step 6: The central controller issues a towing vehicle operation instruction to the driverless towing vehicle, and the driverless towing vehicle executes the corresponding instruction according to the planned path; the manned towing vehicle can be informed to avoid obstacles by means of a walkie-talkie or a ground commander, etc.;

[0142] Step 7: Before the towing vehicle arrives at the airport runway, all vehicles receive a suspension and waiting instruction. The state of the vehicle before entering the runway is represented by a binary state variable g. When the vehicle receives the suspension instruction and pauses to wait, g = 0; when the vehicle continues to move forward, g = 1.

[0143] If the central controller does not receive the instruction that the aircraft has passed through the runway, it issues a passing instruction to the towing vehicle. At this time, g = 1; if it receives the instruction that the aircraft is passing through the runway, the vehicle continues to pause and wait, that is, g = 0. Until the aircraft leaves the runway, a passing instruction is issued to the towing vehicle, and the vehicle can continue to move forward. At this time, g = 1.

[0144] If it receives the instruction that the aircraft is about to pass through the runway, according to the time t1 = la / V for the towing vehicle to pass through the runway and the time t2 when the aircraft enters the runway, it is further determined whether the towing vehicle can pass through the runway first, where la represents the longitudinal distance for the towing vehicle to cross the runway, and V represents the driving speed of the towing vehicle.

[0145] If t0 + t1 < t2, where t0 represents the time for the towing vehicle to reach the suspension area, it means that the towing vehicle can safely pass through the runway before the aircraft enters the runway. To improve the safety and efficiency of the operation, a passing instruction can be issued at this time, that is, g = 1, otherwise g = 0; finally, the towing vehicle reaches the corresponding aircraft apron, and the task ends.

[0146] Embodiment 2

[0147] For the global driving route of the driverless towing vehicle, the present invention uses the A* algorithm for global path planning: The area β is divided into multiple grid cells of the same size. The grid cell size is selected as 1.0. Each step of the vehicle movement occupies the entire grid, and it is assumed that during the path planning process, the positions and sizes of static obstacles are known and stationary.

[0148] The set starting point coordinates (X s , Y s ) are (-3, -5), and the end point coordinates (X t , Y t ) are (30, 30). The ranges of the boundaries and obstacles are y = -10, x ∈ (-10, 30); y = 60, x ∈ (-10, 60); x = -10, y ∈ (-10, 60); x = 40, y ∈ (20, 60); x = 60, y ∈ (-10, 40).

[0149] Calculate the evaluation function value f(n) = g(n) + h(n), and finally obtain the optimal path from the starting point to the end point; where g(n) = (X n - X s ) 2 + (Y n - Ys ) 2 ; h(n) = (X t - X n ) 2 + (Y t - Y n ) 2 ; The specific visualization result is as Figure 3 shown.

[0150] Example 3

[0151] The driverless tractor enters the parking lot to the aircraft parking apron area. Its own motion state is collected by GPS positioning technology and wireless sensor network technology and is recognized as a driverless vehicle by the system, and the trajectory optimization process is started for it;

[0152] The driverless tractor detects the surrounding traffic environment status information through a high-precision map and its own radar detection equipment, and sends the vehicle and its traffic information in front of and behind the corresponding lane to the cloud and transmits it to the in-vehicle device, so that the traffic status information of the vehicles adjacent to the front and rear of each lane can be known, such as speed and position;

[0153] When the driverless vehicle knows that the distance d of the vehicle in front in the current lane is small while the distance d _a of the vehicle in front in the adjacent lane is large, and the safety distance constraint for lane change can be satisfied, a lane change demand is generated; combined with the definition of each parameter in the lane change cost function minC i in, the weights α1, α2, α3 in the function are valued according to different objective functions, and combined with the position and speed information of the vehicles adjacent to the front and rear of the target lane, the most suitable lane change timing is selected for lane change on the premise of minimizing the lane change cost and satisfying the lane change safety constraint;

[0154] After the lane change is completed, the current vehicle still needs to maintain a following safety distance x from the vehicles in front of and behind in the target lane, that is

[0155] Example 4

[0156] The specific embodiment of the present invention also provides an airport aircraft tractor manned and unmanned mixed operation scheduling optimization device, including: a memory for storing a computer program; a processor for implementing the steps of any of the above-mentioned airport aircraft tractor manned and unmanned mixed operation scheduling optimization methods when executing the computer program.

[0157] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code.

[0158] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one or more of the flows Figure 1 or a combination of multiple flows and / or blocks

[0159] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device that implements the functions specified in Figure 1 one or more of the flows Figure 1 or a combination of multiple flows and / or blocks

[0160] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one or more of the flows Figure 1 or a combination of multiple flows and / or blocks

[0161] Obviously, the above embodiments are merely examples given for clear illustration and are not limitations on the implementation manners. For those of ordinary skill in the art, other different forms of changes or variations can be made based on the above description. It is not necessary and impossible to exhaustively list all implementation manners here. And the obvious changes or variations derived therefrom are still within the protection scope of the present invention.

Claims

1. An optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors, characterized in that, Including: Classifying and numbering tractors of different vehicle models; Determining the available tractor models and information according to the aircraft model of the flight to be served, and selecting a tractor to serve the flight; Real-time positioning of tractors within a specified area, and establishing a communication connection with the tractors in this area using a central controller; Tractors that can establish a communication connection are driverless tractors, and those that cannot establish a communication connection are manned tractors; The operation trajectory of the manned tractor is fixed; the A * algorithm is used to perform global path planning on all unmanned tractors in service simultaneously; Performing local path planning for the current driverless tractor and constructing a vehicle conflict decision: Determine the safe following distance between vehicles in the normal road vehicle following scenario d0 is the headway between the two vehicles at time t; d1 and d2 are the travel distances of the leading vehicle and the current vehicle within the time t+T respectively, where T=(v1-v2) / a; v1 is the speed of the leading vehicle; v2, t, and a are the speed, reaction time, and deceleration of the current vehicle respectively; They are the average speed and average time of the vehicle driven manually respectively; is the artificial oscillation term; In the scenario of an unsignalized intersection on the airport road, considering the merging and diverging conflicts, determine the safe following distance between vehicles x min is the minimum safe distance between the following vehicles; is the arc length when the vehicle turns and leaves the lane; r is the turning radius of the vehicle; h is the lane width; Dividing the airport intersection road into a driving area, a buffer area, and a restricted area. Vehicles must successfully change lanes before the restricted area, and lane changes are not allowed within the restricted area. Discuss the lane change situation: If there is no vehicle in front of the current driving lane of the vehicle, or the distance from the vehicle in front is much greater than the minimum safe distance d safe = x, the vehicle maintains the current lane; If there is a vehicle in the front of the current driving lane of the vehicle at a short distance, and there is no vehicle in the front of the adjacent lane, the driverless vehicle freely changes lanes to the adjacent idle lane; If there are vehicles in both the front and adjacent lanes, when d≥d safe , the vehicle maintains the current lane; when d≤d safe and d _a ≥d safe_a , d _b ≥d safe_b , d≤d _a , the vehicle executes the lane change command. If both adjacent lanes meet this condition, the lane change cost function is judged, and the optimal lane change decision is executed; otherwise, the current vehicle reduces its speed and maintains a following distance; d is the distance between the driving vehicle and the vehicle in front in the current lane, d _a , d _b , d safe_a , d safe_b are the distances between the driving vehicle and the vehicles in front and behind in the adjacent lanes and the corresponding minimum safety distances respectively. If the vehicle has a steering requirement in the restricted area, it must execute the lane change decision; Considering factors such as the total number of lane changes, total delay, and driving safety, establish a lane change cost function; Considering the number of lanes crossed by the vehicle during lane change, lane change time constraint, and continuous lane change time constraint, obtain the optimal executable lane change decision; In the scenario of an airport road without signal intersections, set the area where the probability of conflict between two-way vehicle flows at the intersection is greater than the threshold as the conflict area. When the tractor enters the restricted area, it sends a passing demand to the central controller. The central controller calculates the time required for different vehicles in the two-way vehicle flows to travel from the boundary of the restricted area to the boundary of the conflict area at this time. If the time periods corresponding to different vehicles do not overlap, they drive normally; If they overlap, the central controller issues a cutting-in instruction to the vehicle closer to the conflict area.

2. An optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors according to claim 1, characterized in that, Classifying and numbering tractors of different vehicle models; According to the flight information, the aircraft model corresponding to the flight to be served can be known, so as to determine the available tractor models for the flight to be served; Selecting a tractor from the available tractors for the flight to be served according to the information of the available tractors for the flight to be served, including: The vehicle models of tractors can be divided into small tractors, medium tractors, and large tractors. Classification numbers are assigned according to the vehicle models of tractors, namely Q si 、Q mi 、Q li , i = 1, 2, 3...; where s represents small tractors, and Q si represents the set corresponding to small tractors; m represents medium tractors, and Q mi represents the set corresponding to medium tractors; l represents large tractors, and Q li represents the set corresponding to large tractors; Tractors of different vehicle models can tow different maximum weights, and aircraft also have different models. Therefore, consider using an appropriate model of tractor to tow the corresponding model of aircraft; According to the flight information, the aircraft model corresponding to the flight to be served can be known, so as to determine the available tractor models for the flight to be served; Determining and selecting a tractor from the available tractors for the flight to be served according to the information of the tractor, the historical parking time of the aircraft, and the flight departure time; The information of the tractor includes the tractor status, the fuel or power of the tractor, the actual distance from the aircraft apron, and the sum of the historical average operation times of other service vehicles; The tractor status includes an idle status and an operation status. Among them, the tractors in the idle status include the tractors that have completed the previous operation and are on the way back, and the tractors waiting for dispatch orders in the parking lot.

3. An optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors according to claim 2, characterized in that, The real-time positioning of tractors within a specified area and establishing a communication connection with the tractors in this area using a central controller; Tractors that can establish a communication connection are driverless tractors, and those that cannot establish a communication connection are manned tractors include: In the area from the tractor parking lot to the aircraft apron, a wireless sensor network is used to locate vehicles in real time. For area β from the service vehicle parking lot to the aircraft apron, the central controller establishes communication connections with the vehicles in the area. Among them, the unmanned tractors can establish communication connections, while the manned tractors cannot. And according to the communication connection results, the unmanned tractors and the manned tractors are numbered again. Therefore, in the central controller, combined with the digital map, the position information of the two types of vehicles in driving states can be obtained in real time, and then scheduling and trajectory tracking can be carried out.

4. An optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors according to claim 3, characterized in that, Set the operation trajectory of the manned tractor to be fixed; use Algorithm A * to simultaneously perform global path planning on all unmanned tractors that are currently in service, and determine the driving routes of all unmanned tractors that are currently in service, including: Set the operation trajectory of the manned tractor to be fixed: For the manned tractor, it travels according to the existing driving route. At the same time, an in-vehicle voice package is equipped for the manned vehicle to guide and correct the driver's driving style to ensure that it travels according to the specifications. Among them, the existing driving route refers to the driving route of the existing manned tractors in the actual situation of the airport. To reduce costs, the driving route of the manned tractor remains unchanged during driving. Adopt A * The algorithm simultaneously performs global path planning on all the self-driving tractors in service to determine the driving routes of all the self-driving tractors in service at present: divide the area β from the service vehicle parking lot to the aircraft apron into multiple grid cells of the same size, and assume that during the path planning process, the positions and sizes of the static obstacles are known and stationary; and assume that the driving path of the planned vehicle occupies the entire grid at each step; when the grid is occupied, it is considered that there is an obstacle in the current grid; when the grid is not occupied, it is considered that this grid is a free space without obstacles; use A * The algorithm obtains the best path from the current point to the end point according to the lowest cost of different grids, that is, the minimum evaluation function value. Among them, since the idle tractors may be on the way back, the current point is not necessarily the starting point at this time; A * The evaluation function of the algorithm is as follows: f(n) = g(n) + h(n); where g(n) = (X n - X s ) 2 + (Y n - Y s ) 2 ; h(n) = (X t - X n ) 2 + (Y t - Y n ) 2 ; f(n) is the evaluation function of the current point; g(n) is the past cost function, used to evaluate the cost from the starting point to the current point; h(n) is the current cost function, used to evaluate the cost from the current point to the target node; at this time, the Euclidean distance is used for calculation; (X s , Y s ) are the coordinates of the starting point P s , (X n , Y n ) are the coordinates of the current point P n , and (X t , Y t ) are the coordinates of the target point P t .

5. An optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors according to claim 1, characterized in that Considering the total number of lane changes, the total delay of lane changes, and the driving safety factor, establish a lane-changing cost function. At the same time, consider the number of lanes crossed by the vehicle during lane change, the lane-changing time constraint, and the continuous lane-changing time constraint, and solve to obtain an executable optimal lane-changing decision, including: Considering the total number of lane changes, the total delay of lane changes, and the driving safety factor, assign different weights to them, establish a lane-changing cost function, and find the minimum value, which is expressed as: Among them, α1, α2, and α3 represent different weight coefficients, which can take different values in different cases, and α1 + α2 + α3 = 1; N i represents the total number of lane changes of the i-th vehicle; t i represents the lane-changing time of the i-th vehicle, L i represents the lane-changing distance of the i-th vehicle; According to the airport high-precision map and GPS positioning technology, t i -L i / V represents the lane-changing delay time of the i-th vehicle; y i0 represents the ordinate of the starting point of the i-th vehicle's lane change; y i1 represents the ordinate of the vehicle in front in the current lane at the start of the lane change; y i2 represents the ordinate of the end point of the i-th vehicle's lane change; Therefore, the distance d between the vehicle and the vehicle in front in the current lane is d = y i1 -y i0 -L; To ensure the stability and safety of vehicle driving, the following constraints are imposed on the lane-changing cost function: (1) It is stipulated that when a vehicle changes lanes, it can only change to an adjacent lane sequentially during lane changing, that is, l i ≤ 1, where l i represents the number of lanes that the i-th vehicle is allowed to cross at one time (2) Lane-changing time constraint; Vehicles cannot execute lane-changing decisions arbitrarily. It is stipulated that the vehicle must complete the lane change within the allowed maximum lane-changing length, that is, the lane-changing time of the vehicle must be less than the specified maximum allowed lane-changing time, that is: Among them, T max represents the maximum lane-changing time allowed by regulations, L max represents the maximum lane-changing length allowed; V represents the current vehicle driving speed; (3) Continuous lane-changing time constraint; To meet the driving requirements of the vehicle, the vehicle is allowed to change lanes continuously for multiple times. It is stipulated that the lane-changing time of the vehicle satisfies: wherein, represents the time point when the lane change of the i-th vehicle ends, represents the time point when the next lane change of the i-th vehicle ends, and T min represents the minimum allowable continuous lane change time; According to the constraint conditions, find the minimum value of the lane-changing cost function.

6. An optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors according to claim 1, characterized in that When considering vehicle cutting in, in the scenario of an unsignalized intersection on the airport road, at this time, the area where the probability of conflict between two-way vehicle flows at the intersection is greater than the threshold is set as the conflict area. Starting from the moment when the tractor enters the restricted area, it sends a traffic demand to the central controller. The central controller calculates the time required for different vehicles in the two-way vehicle flows to travel from the boundary of the restricted area to the boundary of the conflict area at this time, and judges whether the corresponding time periods of different vehicles overlap. If they do not overlap, they drive normally; If they overlap, the central controller issues a cutting-in instruction to the vehicle closer to the conflict area, including: Set the area where the probability of conflict between two-way traffic at an intersection is greater than the threshold as the conflict area; after the towing vehicle finishes changing lanes and starts entering the restricted area, send a traffic demand to the central controller. The central controller calculates the time required for different vehicles in the two-way traffic to travel from the boundary of the restricted area to the conflict area at this time, and determines whether the time period Δt from the ordinate of the boundary point of the restricted area to the ordinate of the boundary point of the conflict area plus the width of one vehicle body will overlap, that is, determine whether Δt m =(t m ,t′ m ) overlaps with Δt n =(t n ,t′ n ). where t' m = t m + (w m + L) / V, t' n = t n + (w n + L) / V; t m represents the moment when the m-th vehicle arrives at the boundary of the restricted area; t' m represents the moment when the m-th vehicle arrives at the boundary of the conflict area; t n represents the moment when the n-th vehicle arrives at the boundary of the restricted area; t' n represents the moment when the n-th vehicle arrives at the boundary of the conflict area; X m represents the vertical coordinate difference between the boundary point of the restricted area of the m-th vehicle and the boundary point of the conflict area; X n represents the vertical coordinate difference between the boundary point of the restricted area of the n-th vehicle and the boundary point of the conflict area; The positions of the m-th vehicle and the n-th vehicle are in different directions at the intersection. If Δt does not overlap, the two-way traffic flows normally; if Δt overlaps, the central controller issues an acceleration and preemption instruction to the driverless vehicle closer to the conflict point, that is, to the driverless vehicle corresponding to min{w m , w n}; The driverless vehicle accelerates to the maximum speed limit V max at the maximum acceleration a max and then travels at a constant speed through the restricted section, and at this time the driverless vehicle can preemptively pass through the intersection.

7. An optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors according to claim 1, characterized in that, For more complex lane-changing behaviors and cutting-in behaviors, considering sudden changes in traffic conditions, establish a feedback mechanism including: During the execution of the decision instruction, considering that the traffic environment suddenly changes and poses a threat to vehicle safety, establish a decision feedback mechanism, set a 0-1 variable f to facilitate interrupting the instruction in time; Among them, when f = 0, it means that the decision-making is being executed and no other instructions are executed. At this time, the current instruction is in an interruptible state; when f = 1, it means that the decision-making has been completed or aborted, and new instructions can be continued to be executed. At this time, the current instruction ends and cannot be interrupted and fed back. When the decision-making is in a following state, the execution state is set to completed at this time, that is, f = 1. If the decision instruction at this time is a lane change instruction, combined with the high-precision map and GPS, the change in the size of the distance Δl between the vehicle and the lane line is obtained, as well as the moment t when the distance between the vehicle and the lane line is 0. h ; According to Δl and t h Determine whether the vehicle lane change is completed; before the distance is getting smaller but has not yet changed to 0, that is, when it is before the moment t h A lane change interruption instruction can be issued, otherwise the instruction cannot be interrupted; If the decision instruction at this time is an acceleration and cut-in instruction, and it is determined that the vehicle is in a restricted area section by combining the high-precision map and GPS, the driving time t within any Δt time period is obtained a ; if t a is less than the vehicle's uniform driving time, it indicates that the vehicle is accelerating. At this time, a cut-in interruption instruction can be issued, that is, f can be set to 0, and then the next cycle instruction can be continuously judged. Otherwise, the instruction cannot be interrupted.

8. An optimization method for mixed operation scheduling of manned and unmanned airport aircraft tractors according to claim 1, characterized in that To avoid vehicle-aircraft conflict accidents on the airport runway, a binary state variable is set up, and a decision-making judgment model before entering the runway area is established to determine the state of the tractor crossing the aircraft runway until the aircraft safely reaches the parking apron for towing operations, including: Before the towing vehicle arrives at the airport runway, all vehicles receive a suspension waiting instruction. The state of the vehicle before entering the runway is represented by a binary state variable g. When the vehicle receives the suspension instruction and suspends waiting, g = 0; when the vehicle continues to pass, g = 1. If the central controller does not receive the instruction that the aircraft has passed through the runway, it issues a passing instruction to the towing vehicle. At this time, g = 1; if it receives the instruction that the aircraft is passing through the runway, the vehicle continues to suspend waiting, that is, g = 0, until after the aircraft leaves the runway, a passing instruction is issued to the towing vehicle, and the vehicle can continue to pass. At this time, g = 1. If it receives the instruction that the aircraft is about to pass through the runway, according to the time t1 = la / V for the tractor to pass through the runway and the time t2 for the aircraft to enter the runway, it is further determined whether the tractor can pass through the runway first, where la represents the longitudinal distance for the tractor to cross the runway and V represents the driving speed of the tractor. If t0 + t1 < t2, where t0 represents the time for the tractor to reach the suspension area, it means that the tractor can safely pass through the runway before the aircraft enters the runway. To improve the safety and efficiency of the operation, a passing instruction can be issued at this time, that is, g = 1, otherwise g = 0; finally, the tractor reaches the corresponding aircraft parking apron, and the task ends.

9. An optimization device for the mixed operation of manned and unmanned airport aircraft tractors, characterized in that, Including: A memory for storing a computer program; A processor for implementing the steps of a method for optimizing the mixed operation of manned and unmanned airport aircraft tractors as described in any one of claims 1 to 8 when executing the computer program.

Images