CN109147396A - The distribution method and device of airport aircraft gate - Google Patents
The distribution method and device of airport aircraft gate Download PDFInfo
- Publication number
- CN109147396A CN109147396A CN201810969491.9A CN201810969491A CN109147396A CN 109147396 A CN109147396 A CN 109147396A CN 201810969491 A CN201810969491 A CN 201810969491A CN 109147396 A CN109147396 A CN 109147396A
- Authority
- CN
- China
- Prior art keywords
- flight
- probability
- aircraft gate
- airport
- sample
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G5/00—Traffic control systems for aircraft, e.g. air-traffic control [ATC]
- G08G5/0043—Traffic management of multiple aircrafts from the ground
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G5/00—Traffic control systems for aircraft, e.g. air-traffic control [ATC]
- G08G5/0073—Surveillance aids
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Traffic Control Systems (AREA)
Abstract
The embodiment of the present invention provides the distribution method and device of a kind of airport aircraft gate, this method comprises: according to the status information of flight and airport environment to be allocated, constructing environment state matrix；By ambient condition Input matrix to probability assignments model, the probability graph of probability assignments model output is obtained, probability graph, which is used to indicate, distributes flight to the probability of each aircraft gate；It is that flight distributes aircraft gate according to probability graph.The embodiment of the present invention is constructed airport environment state matrix, is promoted the accuracy of modeling, increase the safety of break indices by the status information according to flight and airport environment to be allocated；When the status information of flight and airport environment to be allocated changes, it is only necessary to rebuild airport environment state matrix, can obtain corresponding break indices scheme, be easily managed；Directly by ambient condition Input matrix to probability assignments model, flight is distributed using probability assignments model in real time, save operation time, improves Gate Position Scheduling problem efficiency.
Description
Technical field
The present embodiments relate to the distribution methods and device in Airport Operation optimization field more particularly to airport aircraft gate.
Background technique
In recent years, air transportation is quickly grown.However, the development of air transportation also brought to civil aviaton's operation management it is huge
Challenge, airport are continuously increased as the air traffic pressure that the starting point of flow carries, and operational efficiency is affected, and becoming causes
Delayed one of the main reasons.During Airport Operation, break indices are one of core resources, are station departure flights
The terminal of starting point and flight of marching into the arena, the allocation result of aircraft gate directly affect the programs of personnel and goods and materials.Therefore, it shuts down
The optimization distribution of position is ensureing airport security, plays highly important effect in efficient operation.
In the actual motion of airport break indices, the distribution of aircraft gate is generally divided into predistribution and dynamic allocation two
Stage.
In pre-allocation stage, the algorithm of the prior art is according to known flight planning, seat in the plane resource, operation constraint condition etc.
Seat in the plane predistribution is carried out to flight.However, the distribution method of the prior art in optimization problem modeling, is needed to all kinds of practical fortune
Row constraint is abstracted, and since constraint condition in the prior art is difficult to cover all operations constraint in practice comprehensively, is caused
There is very big security risk in the allocation plan of the prior art.In addition, when the modification of Gate Position Scheduling administrative provisions, the prior art
Airport distribution method needs to model the constraint condition of optimization problem again, needs the intervention again of algorithm development personnel, is not easy
In final-period management.
In the stage of dynamic allocation, then need for flight number variation issue caused by the abnormal conditions such as flight delay, it is right
Break indices scheme is adjusted in time.Since emergency case is more in Airport Operation, flight number change the case where it is frequent
Occur, therefore it is big but higher to the robustness and requirement of real-time of the algorithm of the prior art to dynamically distribute problem randomness.However
The distribution method of the prior art is based substantially on the predistribution of flight, so calculating, the time is long, and difficulty or ease meet the real-time of dynamic allocation
Property require.
For security risk existing in the prior art, difficulty or ease management and calculate time long problem, it would be highly desirable to propose a kind of new
Technical solution overcomes the above problem.
Summary of the invention
The embodiment of the present invention provides a kind of airport shutdown for overcoming the above problem or at least being partially solved the above problem
The distribution method and device of position.
In a first aspect, the embodiment of the present invention provides a kind of distribution method of airport aircraft gate, comprising: according to boat to be allocated
The status information of class and airport environment, constructing environment state matrix；By ambient condition Input matrix to probability assignments model, obtain
The probability graph of probability assignments model output, probability graph, which is used to indicate, distributes flight to the probability of each aircraft gate；According to probability
Figure is flight distribution aircraft gate.
Second aspect, the embodiment of the present invention provide a kind of distributor of airport aircraft gate, comprising: modeling module is used for
According to the status information of flight and airport environment to be allocated, constructing environment state matrix；Processing module is used for ambient condition
Input matrix obtains the probability graph of probability assignments model output, probability graph, which is used to indicate, distributes flight to probability assignments model
To the probability of each aircraft gate；Distribution module, for being that flight distributes aircraft gate according to probability graph.
The third aspect, the embodiment of the present invention provides a kind of electronic equipment, including memory, processor and is stored in memory
Computer program that is upper and can running on a processor, processor realize the airport aircraft gate that first aspect provides when executing program
Distribution method the step of.
Fourth aspect, the embodiment of the present invention provide a kind of non-transient computer readable storage medium, are stored thereon with calculating
Machine program realizes the step of the distribution method for the airport aircraft gate that first aspect provides when the computer program is executed by processor
Suddenly.
The embodiment of the present invention constructs airport environment state by the status information according to flight and airport environment to be allocated
Matrix improves the accuracy of modeling, increases the safety of break indices；When the shape of flight and airport environment to be allocated
When state information changes, it is only necessary to airport environment state matrix is rebuild, can obtain corresponding break indices scheme,
It is easily managed；Directly by ambient condition Input matrix to probability assignments model, flight is carried out using probability assignments model real-time
Distribution, saves operation time, improves Gate Position Scheduling problem efficiency.
Detailed description of the invention
Fig. 1 is the flow diagram of the distribution method of airport aircraft gate provided in an embodiment of the present invention；
Fig. 2 is the tactful network model training flow diagram that the specific embodiment of the invention provides；
Fig. 3 is the plot simulation flow schematic diagram that the specific embodiment of the invention provides；
Fig. 4 is the structural schematic diagram of the distributor of airport aircraft gate provided in an embodiment of the present invention；
Fig. 5 is the entity structure schematic diagram of a kind of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
With reference to the accompanying drawings and examples, specific embodiments of the present invention will be described in further detail.Implement below
Example is not intended to limit the scope of the invention for illustrating the present invention.
The distribution of airport aircraft gate is the core business of airport operation scheduling.Unreasonable airport break indices scheme meeting
Cause flight delay and it is crowded, reduce passenger satisfaction degree, influence the normal operation of relevant departments, or even will cause the hair of accident
It is raw.Therefore, it is necessary to reasonably be distributed airport aircraft gate, with guarantee passenger can easily up and down flight, pick up one's luggage,
Favourable turn and entering and leaving port, while rationally utilizing the aircraft gate on airport, it is ensured that airport ground items operation is gone on smoothly.
Fig. 1 is the flow diagram of the distribution method of airport aircraft gate provided in an embodiment of the present invention.As shown in Figure 1, should
Method includes:
Step 101, according to the status information of flight and airport environment to be allocated, constructing environment state matrix；
Step 102, by ambient condition Input matrix to probability assignments model, the probability of probability assignments model output is obtained
Figure, probability graph, which is used to indicate, distributes flight to the probability of each aircraft gate；
It step 103, is that flight distributes aircraft gate according to probability graph.
Specifically, step 101, the state of airport environment includes airport aircraft gate resource occupation state, break indices rule
Then, the estimated time and flight of flight into aircraft gate go out the estimated time etc. of aircraft gate.It is practical according to flight aircraft gate to be allocated
Flight corresponding to listed flight to be allocated is into stopping in occupied state, current break indices rule and flight schedule
The estimated time and flight of seat in the plane go out the estimated time of aircraft gate, construct ambient condition matrix corresponding to flight to be allocated.
Step 102, probability assignments model is constructed, such as by being trained to neural network, obtains probability assignments model.
The probability assignments model is used for the ambient condition matrix according to flight, provides the allocation plan of flight aircraft gate.By the ring of flight
Border state matrix, input probability distribution model, probability assignments model will export the corresponding probability graph of the flight.The institute of flight is right
The probability graph answered is recorded the flight and is distributed to the probability of each aircraft gate on airport.For example, it is assumed that there are 100 aircraft gates on airport,
The probability graph of flight A is recorded flight A and is distributed to 100 corresponding probability of 100 aircraft gates.
Step 103, show that flight is distributed to the probability graph of each aircraft gate in airport, system can be according to the probability graph
Corresponding aircraft gate is selected, actual aircraft gate as flight.For example, describing flight A distribution in the probability graph of flight A
To the probability of each aircraft gate, the maximum aircraft gate of the select probabilities such as system is as the actual aircraft gate flight A.
The embodiment of the present invention constructs airport environment state by the status information according to flight and airport environment to be allocated
Matrix improves the accuracy of modeling, increases the safety of break indices；When the shape of flight and airport environment to be allocated
When state information changes, it is only necessary to airport environment state matrix is rebuild, can obtain corresponding break indices scheme,
It is easily managed；Directly by ambient condition Input matrix to probability assignments model, flight is carried out using probability assignments model real-time
Distribution, saves operation time, improves Gate Position Scheduling problem efficiency.
On the basis of the above embodiments, as a kind of optional embodiment, the status information of airport environment includes: to shut down
Position resource occupation status information, break indices rule, flight into aircraft gate estimated time and flight go out the estimated of aircraft gate
Moment；Correspondingly, according to the status information of flight and airport environment to be allocated, constructing environment state matrix, comprising: according to stopping
Seat in the plane resource occupation state constructs shutdown status matrix；And according to break indices rule, construct the constrained vector of flight；
Go out the estimated time building of aircraft gate according to the estimated time and flight of shutdown status matrix, constrained vector, flight into aircraft gate
Flight status matrix；According to shutdown status matrix and flight status matrix, constructing environment state matrix.
Specifically, current time t is as the 0th time step, then from the 0th time step to m-th time step, i.e., currently
Moment t is used to indicate that the observable airport aircraft gate resource of t moment is accounted for t+M × time Δt aircraft gate resource occupation state
Use state.Wherein, Δ t indicates time step, i.e., preset duration, such as is within 5 minutes a time step, and M indicates of time step
Number.For example, current time is 10:00, T is five minutes, M 10, then moment t is indicated to t+M × time Δt from the moment
10:00 to moment 10:50.According to aircraft gate resource occupation state, to establish corresponding aircraft gate state matrix.For example, airport
The number of aircraft gate is N, then the aircraft gate state matrix of moment t, is exactly the matrix A of (M+1) × Nt。
Wherein, aij,tIt indicates to observe j-th of aircraft gate in the t+i occupied situation of time step in t-th time step,
If j-th of aircraft gate is in the t+i time step, occupied aij,t=1, otherwise aij,t=0.
According to break indices rule, the rule constraint vector of flight is constructed.For example, the rule constraint vector of flight v is
Pv。
Pv=(pv1,pv2,...,pvN) (2)
Wherein pvnIndicate whether flight v can stop into aircraft gate n, if flight v can stop into aircraft gate n, pvn=1, otherwise
pvn=0.
According to the Flight Information of flight v in flight schedule, the t for expecting to reach be constantly after t of flight v is obtained1When a
Spacer step, i.e. moment t+t1× Δ t, then occupancy matrix be
According to the rule constraint vector P of flight vvWith occupancy matrixShow that flight v can use shutdown bit vector Cvt。
Assuming that the estimated time of flight into aircraft gate is the t after t1A time step, the estimated time that flight goes out aircraft gate is t
T afterwards2A time step, then the flight status matrix B of flight vvt。
Wherein, δm,t=Cvt,m∈[t1,t2],If flight v can be distributed to aircraft gate n, and flight v
The downtime on the n of aircraft gate be after t t1A time step, until t after t2A time step, then bmn,t=1, m ∈
[t1,t2],
According to the shutdown status matrix A of flight vtWith state matrix Bvt, construct the ambient condition matrix s of flight vt,
The embodiment of the present invention by according to aircraft gate resource occupation state, break indices rule and flight into aircraft gate and
The scheduled time of aircraft gate out constructs airport environment state matrix, can be established corresponding according to different break indices principles
Constrained vector, improve the accuracy of modeling, increase the safety of break indices.
On the basis of the above embodiments, as a kind of optional embodiment, by ambient condition Input matrix to probability point
Before model, comprising: construction strategy network model；Based on tactful network model, carry out plot emulation, each plot emulation from
0th time step starts, and successively plus 1 time step, imitates the flight at each emulation moment in flight schedule sample
Very, corresponding state sample of each emulation moment is obtained；Adoption status sample game omits network model and is updated, and updates and completes
Afterwards, probability assignments model is obtained.
Specifically, the structural parameters of Provisioning Policy network model, the number of plies, every layer of neuron number including neural network
Deng, call Theano deep learning library or other can be realized the library of neural network, obtain tactful network model, and use random number
The coefficients θ such as weight, biasing to the strategy network are initialized.Concrete type of the embodiment of the present invention to tactful network model
With no restriction, such as tactful network model can be full Connection Neural Network.Since the 0th time step, successively plus at one
Spacer step, until maximum time step number T, wherein the specific duration of time step can self-setting, for example, time step can be set to 5
Minute, it also can be set 1 minute etc., time step is smaller, observes required for the aircraft gate resource occupation state of same time length
Matrix dimensionality it is bigger.The flight at each of flight schedule sample emulation moment is emulated, is obtained each imitative
State sample corresponding to the true moment.By certain method, will emulate state sample obtained to tactful network model into
Row training constantly updates tactful net coefficients θ, finally obtains probability assignments model.
Fig. 2 is the tactful network model training flow diagram that the specific embodiment of the invention provides.As shown in Figure 2, comprising:
Step 201: initializing tactful network model training parameter, which includes flight schedule sample number J, training
Iteration wheel number I, each flight schedule sample parallel artificial plot quantity K in the training of every wheel, each plot emulation are maximum
Time step number T；
Step 202: initialization aircraft gate information, for example, the quantity of aircraft gate and position, allow to stop the aircraft into aircraft gate
The essential informations such as type；
Step 203: break indices rule is read, for example, each aircraft gate allows to stop into which airline flights etc.
Other more complicated rules；
Step 204: according to initialization aircraft gate information and break indices principle, initializing the ambient condition matrix of flight
StSize；
Step 205: reading flight schedule sample；
Step 206: judging whether the number of flight schedule sample is more than or equal to J；If flight number table sample is known in judgement
This number is more than or equal to J, thens follow the steps 209；If judgement knows that the number for reading flight schedule sample is less than J, hold
Row step 207；
Step 207: a flight schedule is randomly choosed from flight schedule sample；
Step 208: it is new to be superimposed a random fluctuation moment generation one for each of selected flight schedule flight
Flight schedule, and the new flight schedule is added in flight schedule sample, and execute step 206；
Step 209: being every in flight schedule sample according to Flight Information and aircraft gate information and allocation rule information
A flight all create-rule constrained vectors；
Step 210: the structural parameters of Provisioning Policy network model, call Theano deep learning library or other can be realized
The library of neural network obtains tactful network model, and is carried out just with random number to coefficients θ such as weight, the biasings of the strategy network
Beginningization；
Step 211: initializing tactful network training cyclic variable, i=1, j=1, k=1；
Step 212: starting the tactful network model training of the i-th round；
Step 213: j-th of flight schedule sample of selection is trained；
Step 214: according to selected flight schedule sample, k plot emulation is carried out, when each plot is emulated from the 0th
Spacer step starts, and successively plus 1 time step, obtains corresponding state sample, and wherein the mode of plot emulation can be imitated successively
It very, can also be with parallel artificial；
Step 215: adoption status sample game omits net coefficients θ and is updated；
Step 216: judging whether j is less than J, if judgement knows that j is less than J, j=j+1, execute step 213, if judgement obtains
Know that j not less than J, thens follow the steps 217；
Step 217: judging whether i is less than I, if judgement knows that i is less than I, i=i+1, execute step 212, if judgement obtains
Know that j not less than J, thens follow the steps 218；
Step 218: storage updates the summary net coefficients θ after the completion of all round strategy network trainings, obtains probability assignments
Model.
The embodiment of the present invention obtains corresponding state sample, and benefit by carrying out plot emulation to flight schedule sample
Tactful network model is trained with state sample, it is final to obtain probability assignments model.
On the basis of the above embodiments, as a kind of optional embodiment, state sample, comprising: ambient condition matrix
Sample, action sequence sample and sample is awarded immediately；Correspondingly, based on tactful network model, since the 0th time step, according to
Secondary plus 1 time step, emulates the flight at each emulation moment in flight schedule sample, obtains each emulation moment
Corresponding state sample, comprising: since the 0th time step, successively plus 1 time step, building each emulation moment are corresponding
The ambient condition matrix samples of flight；If judgement knows that the corresponding true moment at emulation moment, there are flights, by ambient condition
Matrix samples are input to tactful network model, and output obtains corresponding sample probability figure, stopped according to sample probability figure to flight
Seat in the plane is allocated, and obtains allocation result；And according to allocation result, the action sequence sample of emulation moment corresponding flight is obtained
Sample is awarded immediately；If judgement knows that the corresponding true moment at emulation moment, there is no flights, and it is corresponding will to emulate the moment
The action sequence sample of flight and immediately award sample be set as 0.
Specifically, the number of flight schedule sample is J, and the number of plot emulation is K, then to J flight schedule
K plot emulation is carried out respectively, obtains corresponding state sample.Wherein each plot emulation of each flight schedule sample,
Since the 0th time step, successively plus 1 time step, until maximum time step number T.For example, to j-th of flight schedule
The emulation of k-th of plot, since the 0th time step, successively plus 1 time step, until maximum time step number T, each time
Step all corresponds to an emulation moment.Fig. 3 is the three unities simulation flow schematic diagram that the specific embodiment of the invention provides.Such as Fig. 3
Shown, in the i-th round strategy network model training, carrying out k-th of plot emulation to j-th of flight schedule includes:
Step 301: setting t=0, and construct ambient condition matrix S corresponding to the 0th time0 k, specifically, when according to the 0th
The aircraft gate resource occupation state of spacer step establishes corresponding shutdown status matrix A according to formula (1)0 k, with full 0 data initialization
The flight status matrix B of 0th time step0 k, and the ambient condition matrix according to corresponding to the 0th time step of formula (6) building
S0 k；
Step 302: starting the emulation of t-th of time step；
Step 303: the judgement true moment corresponding with the emulation moment corresponding to t-th of time step whether there is flight
It reaches, if judgement is known there are flight arrival, thens follow the steps 304, if judgement is known there is no flight, then follow the steps 308；
Step 304: building St k, by St kIn input policing network model, one in j-th of flight schedule sample is obtained
A flight v1Probability graph；
Step 305: increasing safety filtering and obtain filtered probability graph, specifically, when being constructed t-th according to formula (4)
The available shutdown bit vector C of spacer stept k, according to Ct kThe not available aircraft gate in wherein aircraft gate is found out, will be stopped accordingly in probability graph
Probability corresponding to seat in the plane is set as 0, obtains corresponding probability graph；
Step 306: flight v1 being allocated according to filtered flight probability graph, aircraft gate is obtained according to allocation result
Action sequence at kR is awarded immediatelyt k, execute step 309；
Step 308: setting at k=0, rt k=0；
Step 309: storage St k、at kAnd rt k；
Step 310: judge whether t is greater than T, wherein T is preset maximum time step number, if judgement is known, t is not more than T,
311 are thened follow the steps, if judgement is known, t is greater than T, then terminates；
Step 311: aircraft gate resource occupation state A is updated according to the allocation result of flight aircraft gatet k；
Step 312: the information of next different flight v2 is read in from flight schedule；
Step 314 is based on updated aircraft gate resource occupation state At k, construct the corresponding ambient condition matrix of flight v2
St k, execute step 302.
The embodiment of the present invention obtains corresponding state sample by carrying out plot emulation to flight schedule sample, for strategy
Network model training provides sample.
On the basis of the above embodiments, as a kind of optional embodiment, adoption status sample game omits network model
It is trained, comprising: network model is omited by gradient descent method adoption status sample game and is trained.
Specifically, state sample is corresponding with the emulation of the plot of flight schedule sample, and each flight schedule passes through one
Secondary plot emulates to obtain a corresponding state sample, will obtain K corresponding state samples by K plot emulation.Together
The one wheel plot of J flight schedule sample of reason emulates corresponding J × K state sample, carries out the feelings of I trained iteration wheel number
Section emulation will obtain J × K × I corresponding state samples.
So, in the tactful network training of a round, for a flight schedule in J flight schedule sample
The three unities emulation is carried out, T+1 S will be obtainedt、atAnd rt, t ∈ [0, T].By following formula (7), it is imitative to obtain the plot
The V for the discount sum awarded immediately corresponding to from t-th of time step to the T time step in veryt, substitute into formula (8), to strategy
Net coefficients θ is updated.J × K update is successively carried out, the tactful net coefficients θ that last time updates is retained, is obtained corresponding
Probability assignments model, and as the initial value of net coefficients tactful in next one strategy network training.
Wherein, α is the step-length for updating coefficient θ,It is gradient direction；γ is the discount awarded immediately
Rate, rtIt is the award immediately in t moment；VtBe t moment and later the discount awarded immediately at all moment and, θ is tactful network
Coefficient.
The embodiment of the present invention is omited network model by gradient descent method adoption status sample game and is trained, and obtains general
Rate distribution model.
On the basis of the above embodiments, as a kind of optional embodiment, in the aircraft gate according to probability graph to flight
Before being allocated, comprising: go out to stop according to the estimated time and flight of shutdown status matrix, constrained vector, flight into aircraft gate
The estimated time of seat in the plane constructs the available shutdown bit vector of flight；According to shutdown bit vector can be used, not available aircraft gate is existed
Corresponding probability is set as 0 in probability graph.
Specifically, according to the shutdown status matrix A of flight vt, estimated time (t t after of the flight into aircraft gate1When a
At the time of corresponding to spacer step) and flight go out the estimated time (t after t of aircraft gate2At the time of corresponding to a time step), according to
Formula (3), obtain flight into the aircraft gate moment occupancy matrixWith flight into the occupancy matrix at aircraft gate moment
According to the constrained vector P of flight vvWith occupancy matrixWithAccording to formula (4), show that flight v can use aircraft gate
Vector Cvt.It wherein can be with shutting down in bit vector, if cvt,n=0, then it represents that n-th of aircraft gate is unavailable.Aircraft gate can be used
The allocation probability of aircraft gate corresponding to 0 in vector is set as 0.
The embodiment of the present invention by constructing the available shutdown bit vector of flight according to shutdown status matrix and constrained vector,
And it will can be set as 0 with the allocation probability for shutting down aircraft gate corresponding to 0 in bit vector, guarantee that flight will not be distributed to not available
Aircraft gate improves the safety of airport break indices.
On the basis of the above embodiments, as a kind of optional embodiment, according to probability graph to the aircraft gate of flight into
Row distribution, comprising: according to probability graph, break indices corresponding to maximum probability are to flight in select probability figure；Alternatively, according to
Probability graph, using break indices corresponding to probability corresponding in roulette method choice probability graph to flight.
Specifically, after the probability graph for obtaining flight, system is carried out according to practical aircraft gate of the probability graph to flight to be allocated
Distribution.Break indices corresponding to most probable value are to flight in Systematic selection probability graph.For example, in the probability graph of flight most
Greatest corresponds to the aircraft gate n, practical aircraft gate of the aircraft gate Systematic selection n as flight.Alternatively, waiting to obtain flight corresponding
After probability graph, system removes aircraft gate corresponding to probability value 0 automatically, is carried out using the method for roulette to remaining probability value
Selection, using aircraft gate corresponding to the probability value selected as the practical aircraft gate of flight.Or probability is first greater than one by system
Definite value, such as 50%, it is first screened, then selected from the probability value after screening with the method that wheel disc figure is gambled, will be selected
Practical aircraft gate of the corresponding aircraft gate of probability value as flight.For selecting the practical aircraft gate of flight according to probability graph
Mode is not particularly limited herein.
The embodiment of the present invention is by directly selecting reality of the corresponding aircraft gate of most probable value in probability graph as flight
Reality of the corresponding aircraft gate of corresponding probability value as flight in the method choice probability graph of aircraft gate or use roulette
Border aircraft gate, the method for salary distribution for the aircraft gate on airport provide multiple choices.
Fig. 4 is the structural schematic diagram of the distributor of airport aircraft gate provided in an embodiment of the present invention.As shown in figure 4, should
Device includes: modeling module 401, processing module 402 and distribution module 403.Wherein, modeling module 401 is used for according to be allocated
Flight and airport environment status information, constructing environment state matrix；Processing module 402 is used for ambient condition Input matrix
To probability assignments model, the probability graph of probability assignments model output is obtained, probability graph, which is used to indicate to distribute flight to each, to stop
The probability of seat in the plane；Distribution module 403 is used to be that flight distributes aircraft gate according to probability graph.
Specifically, the airport state of flight include airport aircraft gate resource occupation state, break indices rule, flight into
The estimated time and flight of aircraft gate go out the estimated time etc. of aircraft gate.Modeling module 401 is real according to flight aircraft gate to be allocated
In border occupied state, current break indices rule and flight schedule flight corresponding to listed flight to be allocated into
The estimated time and flight of aircraft gate go out the estimated time of aircraft gate, construct ambient condition matrix corresponding to flight to be allocated.
Probability assignments model is constructed, such as by being trained to neural network, obtains probability assignments model.The probability point
It is used for the ambient condition matrix according to flight with model, the allocation plan of flight aircraft gate is provided.Processing module 402 is by flight
Ambient condition matrix, input probability distribution model, probability assignments model will export the corresponding probability graph of the flight.The institute of flight
Corresponding probability graph is recorded the flight and is distributed to the probability of each aircraft gate on airport.For example, it is assumed that there are 100 shutdown on airport
Position, the probability graph of flight A are recorded flight A and are distributed to 100 corresponding probability of 100 aircraft gates.
Show that flight is distributed to the probability graph of each aircraft gate in airport, distribution module 403 can be according to the probability graph
Corresponding aircraft gate is selected, actual aircraft gate as flight.For example, describing flight A distribution in the probability graph of flight A
To the probability of each aircraft gate, the maximum aircraft gate of staff's select probability is as the actual aircraft gate flight A.
The embodiment of the present invention is by modeling module 401 according to the status information of flight and airport environment to be allocated, building
Airport environment state matrix improves the accuracy of modeling, increases the safety of break indices；When flight to be allocated and
When the status information of airport environment changes, modeling module 401 only needs to rebuild airport environment state matrix, can obtain
Corresponding break indices scheme out, is easily managed；Processing module 402 is directly by ambient condition Input matrix to probability assignments mould
Type distributes flight in real time using probability assignments model, saves operation time, improves Gate Position Scheduling problem efficiency.
On the basis of the above embodiments, as a kind of optional embodiment, the status information of airport environment includes: to shut down
Position resource occupation status information, break indices rule, flight into aircraft gate estimated time and flight go out the estimated of aircraft gate
Moment；Correspondingly, according to the status information of flight and airport environment to be allocated, constructing environment state matrix, comprising: according to stopping
Seat in the plane resource occupation state constructs shutdown status matrix；And according to break indices rule, construct the constrained vector of flight；
Go out the estimated time building of aircraft gate according to the estimated time and flight of shutdown status matrix, constrained vector, flight into aircraft gate
Flight status matrix；According to shutdown status matrix and flight status matrix, constructing environment state matrix.
On the basis of the above embodiments, as a kind of optional embodiment, by ambient condition Input matrix to probability point
Before model, comprising: construction strategy network model；Based on tactful network model, since the 0th time step, successively plus 1
Time step carries out plot emulation to the flight at each emulation moment in flight schedule sample, obtains each emulation moment pair
The state sample answered；Tactful network model is trained by using state sample, after the completion of training, obtains probability assignments mould
Type.
On the basis of the above embodiments, as a kind of optional embodiment, state sample, comprising: ambient condition matrix
Sample, action sequence sample and sample is awarded immediately；Correspondingly, based on tactful network model, since the 0th time step, according to
Secondary plus 1 time step, emulates the flight at each emulation moment in flight schedule sample, obtains each emulation moment
Corresponding state sample, comprising: since the 0th time step, successively plus 1 time step, building each emulation moment are corresponding
The ambient condition matrix samples of flight；If judgement knows that the corresponding true moment at emulation moment, there are flights, by ambient condition
Matrix samples are input to tactful network model, and output obtains corresponding sample probability figure, stopped according to sample probability figure to flight
Seat in the plane is allocated, and obtains allocation result；And according to allocation result, the action sequence sample of emulation moment corresponding flight is obtained
Sample is awarded immediately；If judgement knows that the corresponding true moment at emulation moment, there is no flights, and it is corresponding will to emulate the moment
The action sequence sample of flight and immediately award sample be set as 0.
On the basis of the above embodiments, as a kind of optional embodiment, by using state sample to tactful network
Model is trained, comprising: is omited network model by gradient descent method adoption status sample game and is trained.
On the basis of the above embodiments, as a kind of optional embodiment, in the aircraft gate according to probability graph to flight
Before being allocated, comprising: go out to stop according to the estimated time and flight of shutdown status matrix, constrained vector, flight into aircraft gate
The estimated time of seat in the plane constructs the available shutdown bit vector of flight；According to shutdown bit vector can be used, not available aircraft gate is existed
Corresponding probability is set as 0 in probability graph.
On the basis of the above embodiments, as a kind of optional embodiment, according to probability graph to the aircraft gate of flight into
Row distribution, comprising: according to probability graph, break indices corresponding to maximum probability are to flight in select probability figure；Alternatively, according to
Probability graph, using break indices corresponding to probability corresponding in roulette method choice probability graph to flight.
Fig. 5 is the entity structure schematic diagram of a kind of electronic equipment provided in an embodiment of the present invention.As shown in figure 5, the electronics
Equipment may include: processor (processor) 510, communication interface (Communications Interface) 520, storage
Device (memory) 530 and bus 540, wherein processor 510, communication interface 520, memory 530 complete phase by bus 540
Communication between mutually.Communication interface 540 can be used for the transmission of the information between the distributor and electronic equipment of airport aircraft gate.Place
Reason device 510 can call the logical order in memory 530, to execute following method: according to flight to be allocated and airport ring
The status information in border, constructing environment state matrix；By ambient condition Input matrix to probability assignments model, probability assignments mould is obtained
The probability graph of type output, probability graph, which is used to indicate, distributes flight to the probability of each aircraft gate；It is flight point according to probability graph
With aircraft gate.
In addition, the logical order in above-mentioned memory 530 can be realized by way of SFU software functional unit and conduct
Independent product when selling or using, can store in a computer readable storage medium.Based on this understanding, originally
Substantially the part of the part that contributes to existing technology or the technical solution can be in other words for the technical solution of invention
The form of software product embodies, which is stored in a storage medium, including some instructions to
So that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation of the present invention
The all or part of the steps of example the method.And storage medium above-mentioned include: USB flash disk, mobile hard disk, read-only memory (ROM,
Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. it is various
It can store the medium of program code.
The embodiment of the present invention provides a kind of non-transient computer readable storage medium, the non-transient computer readable storage medium
Matter stores computer instruction, which makes computer execute the distribution side of airport aircraft gate provided by above-described embodiment
Method, for example, according to the status information of flight and airport environment to be allocated, constructing environment state matrix；By ambient condition
Input matrix obtains the probability graph of probability assignments model output, probability graph, which is used to indicate, distributes flight to probability assignments model
To the probability of each aircraft gate；It is that flight distributes aircraft gate according to probability graph.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It realizes by means of software and necessary general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be embodied in the form of software products in other words, should
Computer software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
Method described in certain parts of example or embodiment.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of distribution method of airport aircraft gate characterized by comprising
According to the status information of flight and airport environment to be allocated, constructing environment state matrix；
By the ambient condition Input matrix to probability assignments model, the probability graph of the probability assignments model output, institute are obtained
It states probability graph and is used to indicate and distribute the flight to the probability of each aircraft gate；
It is that the flight distributes aircraft gate according to the probability graph.
2. the distribution method of airport aircraft gate according to claim 1, which is characterized in that the state of the airport environment is believed
Breath include: aircraft gate resource occupation status information, break indices rule, flight into aircraft gate estimated time and flight go out stop
The estimated time of seat in the plane；
Correspondingly, the status information according to flight and airport environment to be allocated, constructing environment state matrix, comprising:
According to the aircraft gate resource occupation state, shutdown status matrix is constructed；And according to the break indices rule, structure
Build the constrained vector of flight；
Gone out according to the shutdown status matrix, the constrained vector, the flight into the estimated time of aircraft gate and the flight
The estimated time of aircraft gate constructs flight status matrix；
According to the shutdown status matrix and the flight status matrix, the ambient condition matrix is constructed.
3. the distribution method of airport aircraft gate according to claim 1, which is characterized in that the ambient condition matrix is defeated
Enter to before probability assignments model, comprising:
Construction strategy network model；
Based on the tactful network model, plot emulation is carried out, each plot emulation is since the 0th time step, successively plus 1
Time step emulates the flight at each emulation moment in flight schedule sample, obtains each emulation moment pair
The state sample answered；
The tactful network model is updated using the state sample, after the completion of update, obtains the probability assignments mould
Type.
4. the distribution method of airport aircraft gate according to claim 3, which is characterized in that the state sample, comprising: ring
Border state matrix sample, action sequence sample and sample is awarded immediately；
Correspondingly, based on the tactful network model, since the 0th time step, successively plus 1 time step, to flight number
The flight at each emulation moment in table sample sheet is emulated, and corresponding state sample of each emulation moment is obtained, comprising:
Since the 0th time step, successively plus 1 time step, the environment shape of corresponding flight of each emulation moment is constructed
State matrix samples；
If judgement knows that the corresponding true moment at the emulation moment, there are flights, and the ambient condition matrix samples are inputted
To the tactful network model, output obtains corresponding sample probability figure, according to the sample probability figure to the aircraft gate of flight
It is allocated, obtains allocation result；And according to the allocation result, the action sequence of the emulation moment corresponding flight is obtained
Sample and sample is awarded immediately；
If judgement knows that the corresponding true moment at the emulation moment, there is no flights, by the emulation moment corresponding flight
Action sequence sample and immediately award sample be set as 0.
5. the distribution method of airport aircraft gate according to claim 3, which is characterized in that described to use the state sample
The tactful network model is updated, comprising: by gradient descent method using the state sample to the policy network
Network model is updated.
6. the distribution method of airport aircraft gate according to claim 2, which is characterized in that described according to the probability graph
Before being allocated to the aircraft gate of flight, comprising:
Gone out according to the shutdown status matrix, the constrained vector, the flight into the estimated time of aircraft gate and the flight
The estimated time of aircraft gate constructs the available shutdown bit vector of the flight；
According to described with bit vector is shut down, by not available aircraft gate, corresponding probability is set as 0 in the probability graph.
7. the distribution method of airport aircraft gate according to claim 1, which is characterized in that described according to the probability graph pair
The aircraft gate of flight is allocated, comprising:
According to the probability graph, select in the probability graph break indices corresponding to maximum probability to the flight；Alternatively,
According to the probability graph, given using break indices corresponding to probability corresponding in probability graph described in roulette method choice
The flight.
8. a kind of distributor of airport aircraft gate characterized by comprising
Modeling module, for the status information according to flight and airport environment to be allocated, constructing environment state matrix；
Processing module, for it is defeated to obtain the probability assignments model by the ambient condition Input matrix to probability assignments model
Probability graph out, the probability graph, which is used to indicate, distributes the flight to the probability of each aircraft gate；
Distribution module, for being that the flight distributes aircraft gate according to the probability graph.
9. a kind of electronic equipment including memory, processor and stores the calculating that can be run on a memory and on a processor
Machine program, which is characterized in that realize that the airport as described in any one of claim 1 to 7 is stopped when the processor executes described program
The step of distribution method of seat in the plane.
10. a kind of non-transient computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer
It is realized when program is executed by processor as described in any one of claim 1 to 7 the step of the distribution method of airport aircraft gate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810969491.9A CN109147396B (en) | 2018-08-23 | 2018-08-23 | Method and device for allocating airport parking positions |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810969491.9A CN109147396B (en) | 2018-08-23 | 2018-08-23 | Method and device for allocating airport parking positions |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109147396A true CN109147396A (en) | 2019-01-04 |
| CN109147396B CN109147396B (en) | 2020-10-30 |
Family
ID=64827657
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810969491.9A Active CN109147396B (en) | 2018-08-23 | 2018-08-23 | Method and device for allocating airport parking positions |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109147396B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109872064A (en) * | 2019-02-12 | 2019-06-11 | 民航成都信息技术有限公司 | A kind of Airport Resources distribution method, device, equipment and medium |
| CN109948844A (en) * | 2019-03-15 | 2019-06-28 | 民航成都信息技术有限公司 | A kind of optimization method, device, equipment and the medium of break indices robustness |
| CN110288857A (en) * | 2019-06-26 | 2019-09-27 | 中国民航大学 | It is a kind of based on when Spatial Dimension decompose airport aircraft gate fast dispatch method |
| CN110443448A (en) * | 2019-07-01 | 2019-11-12 | 华中科技大学 | A kind of aircraft seat in the plane classification prediction technique and system based on two-way LSTM |
| CN111554127A (en) * | 2019-02-12 | 2020-08-18 | 通用电气公司 | Aircraft stand recovery optimization |
| CN111627255A (en) * | 2019-02-27 | 2020-09-04 | 阿里巴巴集团控股有限公司 | Information processing method, device and system |
| TWI724755B (en) * | 2020-01-17 | 2021-04-11 | 桃園國際機場股份有限公司 | Flight operation and transportation system |
| CN112785097A (en) * | 2019-11-04 | 2021-05-11 | 顺丰科技有限公司 | Parking space allocation method and device, storage medium and computer equipment |
| CN113095543A (en) * | 2021-03-01 | 2021-07-09 | 北京工业大学 | Distribution method and system for airport parking space and taxiways |
| CN116993137A (en) * | 2023-09-28 | 2023-11-03 | 民航成都信息技术有限公司 | Method and device for determining stand, electronic equipment and medium |
| CN117933490A (en) * | 2024-03-14 | 2024-04-26 | 中国民航大学 | Airport scene dragging scheduling optimization method, electronic equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104361769A (en) * | 2014-10-27 | 2015-02-18 | 广州市中南民航空管通信网络科技有限公司 | Flight data processing method and scene monitoring front-end processor |
| CN104751681A (en) * | 2015-03-09 | 2015-07-01 | 西安理工大学 | Statistical learning model based gate position allocation method |
| CN106447215A (en) * | 2016-10-12 | 2017-02-22 | 合肥飞友网络科技有限公司 | Method for pre-distributing aircraft berths in airport |
| CN107085976A (en) * | 2017-04-21 | 2017-08-22 | 民航成都信息技术有限公司 | The time-bounded dynamic constrained method in airliner station level ground aircraft gate |
| CN107578159A (en) * | 2017-08-29 | 2018-01-12 | 飞友科技有限公司 | It is a kind of to adapt to the abnormal seat in the plane auto-allocation method of flight itineraries |
-
2018
- 2018-08-23 CN CN201810969491.9A patent/CN109147396B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104361769A (en) * | 2014-10-27 | 2015-02-18 | 广州市中南民航空管通信网络科技有限公司 | Flight data processing method and scene monitoring front-end processor |
| CN104751681A (en) * | 2015-03-09 | 2015-07-01 | 西安理工大学 | Statistical learning model based gate position allocation method |
| CN106447215A (en) * | 2016-10-12 | 2017-02-22 | 合肥飞友网络科技有限公司 | Method for pre-distributing aircraft berths in airport |
| CN107085976A (en) * | 2017-04-21 | 2017-08-22 | 民航成都信息技术有限公司 | The time-bounded dynamic constrained method in airliner station level ground aircraft gate |
| CN107578159A (en) * | 2017-08-29 | 2018-01-12 | 飞友科技有限公司 | It is a kind of to adapt to the abnormal seat in the plane auto-allocation method of flight itineraries |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109872064A (en) * | 2019-02-12 | 2019-06-11 | 民航成都信息技术有限公司 | A kind of Airport Resources distribution method, device, equipment and medium |
| CN111554127A (en) * | 2019-02-12 | 2020-08-18 | 通用电气公司 | Aircraft stand recovery optimization |
| CN111627255A (en) * | 2019-02-27 | 2020-09-04 | 阿里巴巴集团控股有限公司 | Information processing method, device and system |
| CN109948844A (en) * | 2019-03-15 | 2019-06-28 | 民航成都信息技术有限公司 | A kind of optimization method, device, equipment and the medium of break indices robustness |
| CN110288857A (en) * | 2019-06-26 | 2019-09-27 | 中国民航大学 | It is a kind of based on when Spatial Dimension decompose airport aircraft gate fast dispatch method |
| CN110443448B (en) * | 2019-07-01 | 2022-03-29 | 华中科技大学 | Bidirectional LSTM-based airplane position classification prediction method and system |
| CN110443448A (en) * | 2019-07-01 | 2019-11-12 | 华中科技大学 | A kind of aircraft seat in the plane classification prediction technique and system based on two-way LSTM |
| CN112785097A (en) * | 2019-11-04 | 2021-05-11 | 顺丰科技有限公司 | Parking space allocation method and device, storage medium and computer equipment |
| TWI724755B (en) * | 2020-01-17 | 2021-04-11 | 桃園國際機場股份有限公司 | Flight operation and transportation system |
| CN113095543A (en) * | 2021-03-01 | 2021-07-09 | 北京工业大学 | Distribution method and system for airport parking space and taxiways |
| CN113095543B (en) * | 2021-03-01 | 2024-01-12 | 北京工业大学 | Distribution method and system for airport stand and taxiway |
| CN116993137A (en) * | 2023-09-28 | 2023-11-03 | 民航成都信息技术有限公司 | Method and device for determining stand, electronic equipment and medium |
| CN116993137B (en) * | 2023-09-28 | 2023-12-05 | 民航成都信息技术有限公司 | Method and device for determining stand, electronic equipment and medium |
| CN117933490A (en) * | 2024-03-14 | 2024-04-26 | 中国民航大学 | Airport scene dragging scheduling optimization method, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109147396B (en) | 2020-10-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109147396A (en) | The distribution method and device of airport aircraft gate | |
| CN109621422A (en) | Electronics chess and card decision model training method and device, strategy-generating method and device | |
| US8498952B2 (en) | Systems and methods for box fitness termination of a job of an evolutionary software program | |
| CN107423442A (en) | Method and system, storage medium and computer equipment are recommended in application based on user's portrait behavioural analysis | |
| Avni et al. | Run-time optimization for learned controllers through quantitative games | |
| Choi et al. | Artificial neural network models for airport capacity prediction | |
| CN103473111A (en) | Brain-imitation calculation virtualization method and brain-imitation calculation virtualization system | |
| KR101503542B1 (en) | Human behavioral modeling and simulation framework | |
| WO2008106663A2 (en) | Method and apparatus for human behavior modeling in adaptive training | |
| CN109983480A (en) | Use cluster loss training neural network | |
| US10922092B2 (en) | Administrator-monitored reinforcement-learning-based application manager | |
| CN112948412A (en) | Flight inventory updating method, system, electronic equipment and storage medium | |
| CN106779240A (en) | The Forecasting Methodology and system of civil aviaton's market macroscopic view index | |
| Kumar et al. | Graph learning based decision support for multi-aircraft take-off and landing at urban air mobility vertiports | |
| CN102521202B (en) | Automatic discovery method of complex system oriented MAXQ task graph structure | |
| CN103838964A (en) | Social relationship network generation method and device based on artificial transportation system | |
| CN106529835A (en) | Daily traffic capacity determining method and device of terminal area | |
| US20080301631A1 (en) | Human transparency paradigm | |
| Jarrar et al. | Formal approach to model complex adaptive computing systems | |
| Tien et al. | Deep Reinforcement Learning Applied to Airport Surface Movement Planning | |
| de Freitas Cunha et al. | On the impact of MDP design for reinforcement learning agents in resource management | |
| CN112085520A (en) | Flight space control method based on artificial intelligence deep reinforcement learning | |
| WO2024031986A1 (en) | Model management method and related device | |
| KrisshnaKumar et al. | Graph Learning Based Decision Support for Multi-Aircraft Take-Off and Landing at Urban Air Mobility Vertiports | |
| US20230237352A1 (en) | Systems and methods for end-to-end multi-agent reinforcement learning on a graphics processing unit |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |