CN105023068A - Rule mining based flight arrival and departure cooperative scheduling method - Google Patents

Abstract

The present invention discloses a rule mining based flight arrival and departure cooperative scheduling method and belongs to the technical field of flight scheduling. The method comprises the steps of: firstly, analyzing historical operating data of a target airport, performing mapping modeling by applying an artificial neural network, and extracting out operation modes and rules of the airport under weather conditions in different seasons; and then establishing a flight arrival and departure cooperative scheduling model, designing a dynamic optimization algorithm by combination with a cellular automaton, performing optimized sequencing on arrival and departure flights, and inputting an updated scheduling scheme according to real-time data. The method takes the characteristics and interacting factors of the arrival and departure processes of the flights into comprehensive consideration; and meanwhile, with reference to different operation modes, obtained by performing rule mining on the historical operating data of the target airport, of the target airport under different conditions, a flight arrival and departure scheduling model that is more comprehensive and closer to actual scheduling is established, so that the arrival and departure of the flights can be ensured to more efficient.

Description

A kind of flight of rule-based excavation enters to leave the theatre coordinated dispatching method

Technical field

The invention belongs to flight dispatching technical field, the flight being specifically related to a kind of rule-based excavation enters to leave the theatre coordinated dispatching method.

Background technology

Rapidly, CAAC has become world's second largest air transport system in Chinese Aviation Transportation industry development in recent years.Along with the continuous expansion of air transportation scale, airport and terminal area operational efficiency has lowly become the bottleneck of restriction aviation development, causes serious airliner delay and blocks up.Therefore, be necessary to carry out Optimized Operation to the station departure flight stream that enters of termination environment, reduce airliner delay, improve Airport Operation efficiency.Flight enters to leave the theatre cooperative scheduling (Aircraft Landing and Departure CollaborativeScheduling; Hereinafter referred to as ALDCS) be intended to by change into station departure flight land, take off order and be distribute suitable landing, the departure time into station departure flight, solve the brought conflict because termination environment blocks up, improve Airport Operation efficiency as far as possible, reduce landing airliner delay.

Marching into the arena of flight is mainly regarded as two processes be separated independently carry out with being left the theatre by traditional flight dispatching method, the flight of do not consider when marching into the arena (leaving the theatre) scheduling to leave the theatre (marching into the arena).Single (leaving the theatre) scheduling of marching into the arena is to minimize the flight overall delay time at stop for target, ensure schedule flight safely for constraint is optimized flight landing (taking off) order and time, under the prerequisite ensureing constraint, obtain entirety incur loss through delay minimum scheduling scheme.

But also there is certain problem in classic method: in the actual motion of termination environment; entering the use of station departure flight to termination environment resource (as spatial domain, runway, aircraft gate etc.) is restriction mutually, only considers that single scheduling of marching into the arena or leave the theatre is difficult to realize global optimum.

The flight cooperative scheduling that enters to leave the theatre is a complicated engineering optimization, and in actual schedule, have many difficult points: constraint is many, flight of marching into the arena, station departure flight will meet security constraint separately, march into the arena simultaneously and will meet security constraint equally between station departure flight; Uncertain large, as in Various Seasonal, different weather situation, different operational modes is taked on airport accordingly, and operational mode Sum fanction is not what determine; Real-time is high, and flight cooperative scheduling problem of entering to leave the theatre is an optimization problems, needs obtain scheduling result within a short period of time.

Summary of the invention

The object of the invention is to solve problems of the prior art, by the analysis to target airport history data, consider the factor mutually restricted into station departure flight, obtain target air station flight under different operational mode to enter to leave the theatre operation rule, set up flight on this basis to enter to leave the theatre cooperative scheduling model, and design a kind of dynamic optimization algorithm in conjunction with cellular automaton, arrange suitable to take off for entering station departure flight, landing times, simultaneously requirement of real time.

To achieve these goals, the flight that the invention provides a kind of rule-based excavation enters to leave the theatre coordinated dispatching method, comprising:

Step 1, target airport history data (comprise season, weather, aerodrome capacity, enter station departure flight ratio etc.) to be analyzed, using artificial neural networks carries out mapping, modeling, extracts the operational mode on airport under Various Seasonal, weather condition and rule;

Step 2, set up flight and to enter to leave the theatre cooperative scheduling model, take into full account into influencing each other and restricting and Airport Operation pattern between station departure flight;

Step 3, in conjunction with cellular automaton, designing a kind of dynamic optimization algorithm, being optimized sequence to entering station departure flight, and scheduling scheme can be upgraded according to real time data input.

The flight of rule-based excavation provided by the invention enters to leave the theatre coordinated dispatching method, compared with prior art, achieves following technique effect:

(1) consider flight to march into the arena process and the feature of process and the mutual restraining factors of leaving the theatre, simultaneously with reference to target airport history data being carried out to the different operational mode in target airport under different condition that regulation excavation obtains, establish and to enter to leave the theatre scheduling model close to the flight that scheduling is actual more comprehensively, more, can guarantee that flight enters to leave the theatre and run more efficiently.

(2) cellular automaton can automatically upgrade under certain rule, has good dynamic perfromance.Utilize this feature, improve cellular automaton and be used for simulating the process of entering to leave the theatre of flight, and formulate flight targetedly and to enter to leave the theatre rule, ensure that good real-time and accuracy, be more applicable for actual flight and enter to leave the theatre scheduling.

Accompanying drawing explanation

Fig. 1 is that the flight that the present invention is based on rule digging enters to leave the theatre algorithm flow chart in coordinated dispatching method embodiment;

Fig. 2 is target airport of the present invention historical data rule digging block flow diagram;

Fig. 3 is data mining neural network structure figure of the present invention;

Fig. 4 is that the flight that the present invention is based on cellular automaton enters to leave the theatre scheduler module process flow diagram;

Fig. 5 is Least-cost strategic process figure of the present invention.

Embodiment

Also by reference to the accompanying drawings the present invention is described in further detail below by specific embodiment.

The flight of a kind of rule-based excavation of the present invention enters to leave the theatre coordinated dispatching method, and flow process as shown in Figure 1, comprises following step:

Step 1, target airport history data (comprise season, weather, aerodrome capacity, enter station departure flight ratio etc.) to be analyzed, using artificial neural networks carries out mapping, modeling, extracts the operational mode on airport under the condition such as Various Seasonal, weather and rule.

As shown in Figure 2, specifically comprise the steps:

Step 101, set up artificial neural network, comprise input layer, hidden layer, an output layer, structural drawing as shown in Figure 3, input layer is input as Airport Operation environment vector Env [W (t), D (t), A (t), R (t)], wherein element represents meteorology, flight takeoff demand, flight landing demand, landing runway respectively; Output layer exports as entering to leave the theatre landing scheduling result vector Flow [F _a(t), F _d(t), Δ T (t)], wherein element represents the rear unit interval landing sortie of scheduling, the sortie that takes off, minimum safety interval time respectively; The input/output relation of described artificial neural network is expressed as:

$\left\{\begin{array}{c}{u}_1\left(i\right)={\Σ}_{j=1}^4{w}_1(i,j)a_1\left(j\right)+{\mathrm{\θ}}_1\left(i\right)\\ a_2\left(i\right)=f\left(u_1\right(i\left)\right)\\ u_2\left(k\right)={\Σ}_{i=1}^4{w}_2(k,i)a_2\left(i\right)+{\mathrm{\θ}}_2\left(k\right)\\ Out\left(k\right)=f\left(u_2\right(k\left)\right)\end{array}\right.---\left(1\right)$

A in formula ₁j () is an input layer jth input vector, four components of corresponding Airport Operation environment vector Env [W (t), D (t), A (t), R (t)], j=1,2,3,4, a ₂i () is hidden layer input vector, i=1,2,3,4;

W ₁(i, j) is weight vectors, represents from an input layer jth vectorial weight vectorial to hidden layer i-th, in like manner,

W ₂(k, i) represents the weight vectors of a vector from hidden layer i-th vector to output layer kth, k=1,2,3;

θ ₁(i), θ ₂k () is respectively input layer and hidden layer neuron threshold value;

U ₁(i), u ₂k () represents the output of input layer, hidden layer respectively;

Out (k) is that output layer exports, and correspondence is entered to leave the theatre the predicted value of landing scheduling result vector;

Getting excitation function is $f\left(x\right)=\frac{1}{1+e^{-x}};$

Step 102, initializes weights vector w ₁(i, j), w ₂(k, i) and neuron threshold value θ ₁(i), θ ₂(k), and output Out (k) calculating artificial neural network output layer with first group of learning data;

Step 103, oppositely revises weight vectors w according to formula (2) ₁(i, j), w ₂(k, i) and neuron threshold value θ ₁(i), θ ₂k the value of (), gets η=0.1;

$\left\{\begin{array}{c}{\mathrm{\δ}}_2^{p+1}\left(k\right)=\left({\mathrm{Flow}}^p\right(k)-{\mathrm{Out}}^p(k\left)\right)f^{\′}\left({u_2}^p\right(k\left)\right)\\ {\mathrm{\δ}}_1^{p+1}\left(k\right)=f^{\′}\left(u_2^p\right(k\left)\right){\mathrm{\δ}}_2^{p+1}\left(k\right)w_2^{p+1}(k,j)\\ w_2^{p+1}(k,i)=w_2^p(k,i)+{\mathrm{\η\δ}}_2^{p+1}{a}_2^{p+1}\left(i\right)\\ w_1^{p+1}(i,j)=w_1^p(i,j)+{\mathrm{\η\δ}}_1^{p+1}{a}_1^{p+1}\left(j\right)\end{array}\right.---\left(2\right)$

Wherein, p is iterations;

Step 104, training of human artificial neural networks.

Calculate artificial neural network output layer according to formula (3) and export Out (k) and the actual variance Δ exporting Flow (k), if variance Δ is less than setting threshold value (described setting threshold value can value be 0.00001), artificial neural network can be regarded as trained, now weight vectors w ₁(i, j), w ₂(k, i) and neuron threshold value θ ₁(i), θ ₂k the value of () is artificial neural network end value, input new Airport Operation environment vector Env [W (t), D (t), A (t), R (t)] the landing scheduling result vector Flow [F that enters to leave the theatre accordingly can be obtained _a(t), F _d(t), Δ T (t)]; Otherwise input new training data, p=p+1, get back to step 103;

$\mathrm{\Δ}={\mathrm{\Σ}}_{k=1}^3{\left(Flow\right(k)-Out(k\left)\right)}^2---\left(3\right)$

Step 2, set up flight and to enter to leave the theatre cooperative scheduling model, take into full account into influencing each other and restricting and Airport Operation pattern between station departure flight;

In the present invention, problem is described as: known target airfield runway configuration, real time data (comprising flight number, time, position, speed, weather etc.) and the flight planning on station departure flight and target airport is entered in given a period of time, suitable landing time is distributed for every frame enters station departure flight, ensureing, under the prerequisite that airport security is run, to minimize flight landing and incur loss through delay.Be described by the cooperative scheduling model that enters to leave the theatre to the flight in the present invention below, this flight enters to leave the theatre the foundation of cooperative scheduling model based on following hypothesis:

1. landing scheduling clears the runway and lands flight arrival runway for terminating with the flight that takes off, and can not clash, only need ensure that landing flight does not clash at runway place after the flight that takes off clears the runway with the flight that do not land of airflight;

2. taxiway, hardstand and airport service are ideal situation, and namely landing flight can arrive runway in the stipulated time, there is not the problem that flight is incured loss through delay because of hardstand deficiency or taxiway conflict;

Flight enters to leave the theatre cooperative scheduling model to minimize landing airliner delay cost for objective function, each flight planned time (comprising plan landing times and Proposed Departure time), early than or be later than the increase that planned time all can cause flight cost, the unit interval delay cost of flight landing is relevant with the type, airline, cabin factor etc. of flight, land in advance (taking off) according to aircraft and postpone landing (taking off) two kinds of situation definition and incur loss through delay cost function, so definable flight i lands in t or incurs loss through delay cost when taking off be:

$c_i\left(t\right)=\left\{\begin{array}{c}{c}_{mi}^{\′}(t-T_i)\\ c_{ni}^{\′}(T_i-t)\end{array}\right.---\left(4\right)$

C in formula _it () is the delay cost of flight i in t landing, c ' _mifor flight i unit delay cost, c ' _nifor flight i unit shifts to an earlier date cost, T _ifor flight i plans the landing time;

The expense of flight and station departure flight of marching into the arena is different, and flight of marching into the arena is incured loss through delay cost due to the factor such as oil consumption, flight safety and is greater than station departure flight, and the present invention introduces and marches into the arena and station departure flight delay factor α _aand α _d, adjust into station departure flight ratio by adjustment delay factor, airliner delay cost can be calculated more accurately.Ultimate aim for scheduling problem is that tardiness cost minimizes, and sets up objective function to be:

min Cost(t)＝α _a∑ _t∈T∑ _i∈Nc _i(t)+α _d∑ _t∈T∑ _i∈Nc _i(t) (5)

Wherein Cost (t) is landing flight total delay cost, and T is for entering to leave the theatre scheduling T.T. interval, and N enters station departure flight sum in scheduling time.

Only to consider to march into the arena the delay cost of flight or station departure flight relative to existing single flight model of marching into the arena or leave the theatre, the present invention will enter station departure flight delay cost and unite, the flight cooperative scheduling model that enters to leave the theatre is converted into by the single optimum solution solving march into the arena flight or station departure flight the globally optimal solution solved into station departure flight, thus makes the flight cooperative scheduling model that enters to leave the theatre more accord with flight and to enter to leave the theatre actual schedule problem.

Flight constraint condition in cooperative scheduling model of entering to leave the theatre has three, is respectively:

(1) landing time windows constraints: with wherein with for the take-off and landing time of flight i, with be respectively effective departure time scope and the landing times scope of flight i;

(2) minimum safety interval constraint: $\left\{\begin{array}{c}{t}_i^d-t_j^d\≥{\mathrm{\Δt}}^d\\ t_i^a-t_j^a\≥{\mathrm{\Δt}}^a\\ t_i^d-t_j^a\≥{\mathrm{\Δt}}^m\end{array}\right.,$ Wherein Δ t ^d, Δ t ^a, Δ t ^mbe respectively the minimum landing personal distance time between the minimum personal distance time of taking off between adjacent two flight i and j, adjacent two flight i and j, minimum take-off and landing personal distance time between flight i and flight j, namely the corresponding minimum safety interval time to be not less than, to ensure flight safety the interval time of adjacent two frame flights;

(3) landing capacity-constrained: N ^a(t)≤F _a(t) and N ^d(t)≤F _d(t), i.e. the Field Activity time actual flight sortie N that takes off ^dt () is no more than present mode of operation lower unit interval runway departure capacity F _d(t), and Field Activity time actual landing flight sortie N ^at () is no more than present mode of operation lower unit interval Runway Landing Capacity F _a(t).

To sum up, in the present invention, the flight cooperative scheduling model representation that enters to leave the theatre is:

Objective function:

min Cost(t)＝α _a∑ _t∈T∑ _i∈Nc _i(t)+α _d∑ _t∈T∑ _i∈Nc _i(t) (6)

Constraint condition:

$t_i^d\∈\[\underset{\‾}{T_i^d},\stackrel{\‾}{T_i^a}\],i\∈N$

$t_i^a\∈\[\underset{\‾}{T_i^a},\stackrel{\‾}{T_i^a}\],i\∈N$

$t_i^d-t_j^d\≥{\mathrm{\Δt}}^d,i,j\∈N$

$t_i^a-t_j^a\≥{\mathrm{\Δt}}^a,i,j\∈N$

$t_i^d-t_j^a\≥{\mathrm{\Δt}}^m,i,j\∈N$

N ^a(t)≤F _A(t)

N ^d(t)≤F _D(t)

Step 3, in conjunction with cellular automaton, designing a kind of dynamic optimization algorithm, being optimized sequence to entering station departure flight, and scheduling scheme can be upgraded according to real time data input.

After obtaining target airport operational mode and respective rule, in conjunction with cellular Automation Model, the present invention proposes a kind of dynamic optimization algorithm based on cellular automaton.In dynamic optimization algorithm, each is erected and fall flight and regard as a cellular, for each cellular distributes some attributes (in advance cost, delay cost etc.), and the corresponding principle of optimality is set for the renewal of cellular attribute, under the principle of optimality, the attribute of each time step to cellular upgrades, dynamic similation enters the landing flow process of station departure flight, and be finally optimized scheduling result.

Cellular attribute is as follows:

T _ithe plan landing time of flight i, i=1 ..., P;

T _ithe expectation landing time of flight i, i=1 ..., P;

S _ijthe minimum safety interval time (>=0) that flight i and flight j takes off, land, flight i took off, lands before flight j, i=1 ..., P; J=1 ..., P; I ≠ j;

G _iflight i is at plan landing time T _ithe unit before take off, landed shifts to an earlier date cost (>=0), i=1 ..., P;

H _iflight i is at plan landing time T _ithe unit delay cost (>=0) of take off afterwards, landing, i=1 ..., P;

Fig. 4 is dynamic optimization algorithm process flow diagram, specifically comprises the steps:

Step 301, station departure flight data are entered in initialization, calculate every frame flight additional cost value according to formula (7);

In formula (7), t _ifor flight estimates the landing time, T _ifor the flight planning landing time, often erect and fall flight and all can calculate an additional cost value (flight delays landing cost) or (flight shifts to an earlier date landing cost);

Step 302, from the flight of most getting up early (falling), the additional cost value of more adjacent flight successively, adjusted according to the expectation landing time of Least-cost strategy to flight;

Step 303, each time step judges whether that new landing flight adds flight queue, and upgrades each cellular attribute, if having, flight is come flight queue tail, gets back to step 301; If no, then directly get back to step 301;

Step 304, calculate flight according to formula (8) and estimate landing time variations quadratic sum σ, as σ≤α, (α is judgment threshold, generally get 0.0001), namely can be considered that cellular automaton reaches steady state (SS), now the expectation landing time of flight is optimum solution, otherwise returns step 301.

$\mathrm{\σ}={\mathrm{\Σ}}_{i=0}^P{(t_i^g-t_i^{g-1})}^2---\left(8\right)$

In formula, g is cellular evolutionary generation;

Described step 302, process flow diagram as shown in Figure 5, specifically comprises the steps:

Step 3021, if flight meets minimum safety interval constraint at planned time landing (taking off), then adjusts the flight scheduled time, makes flight land (taking off) at planned time;

Step 3022, if flight can not meet minimum safety interval constraint at planned time landing (taking off), the then additional cost of more adjacent flight, higher the adjusting to planned time of cost, lower the remaining unchanged of cost, should note adjusting range, the interval after ensureing adjustment between two frame flights meets minimum safety interval constraint simultaneously;

Last it is noted that above embodiment is only in order to illustrate technical scheme of the present invention, be not intended to limit; Although with reference to previous embodiment to invention has been detailed description, those of ordinary skill in the art is to be understood that: it still can be modified to the technical scheme described in previous embodiment, or carries out equivalent replacement to wherein portion of techniques feature; And these amendments or replacement, do not make the essence of appropriate technical solution depart from the spirit and scope of embodiment of the present invention technical scheme.

Claims (4)

1. the flight of rule-based excavation enters to leave the theatre a coordinated dispatching method, it is characterized in that:

Step 1, analyze target airport history data, using artificial neural networks carries out mapping, modeling, extracts operational mode and the rule on airport under Various Seasonal, weather condition; Described history data comprises season, weather, aerodrome capacity, enters station departure flight ratio data;

Step 2, set up flight and to enter to leave the theatre cooperative scheduling model;

Step 3, in conjunction with cellular automaton, designing a kind of dynamic optimization algorithm, being optimized sequence to entering station departure flight, and scheduling scheme can be upgraded according to real time data input.

2. the flight of a kind of rule-based excavation according to claim 1 enters to leave the theatre coordinated dispatching method, it is characterized in that: described artificial neural network is set up as follows:

Step 101, set up artificial neural network, comprise input layer, hidden layer, an output layer, input layer is input as Airport Operation environment vector Env [W (t), D (t), A (t), R (t)], wherein element represents meteorology, flight takeoff demand, flight landing demand, landing runway respectively; Output layer exports as entering to leave the theatre landing scheduling result vector Flow [F _a(t), F _d(t), Δ T (t)], wherein element represents the rear unit interval landing sortie of scheduling, the sortie that takes off, minimum safety interval time respectively; The input/output relation of described artificial neural network is expressed as:

$\left\{\begin{array}{c}{u}_1\left(i\right)={\Σ}_{j=1}^4{w}_1(i,j)a_1\left(j\right)+{\mathrm{\θ}}_1\left(i\right)\\ a_2\left(i\right)=f\left(u_1\right(i\left)\right)\\ u_2\left(k\right)={\Σ}_{i=1}^4{w}_2(k,i)a_2\left(i\right)+{\mathrm{\θ}}_2\left(k\right)\\ Out\left(k\right)=f\left(u_2\right(k\left)\right)\end{array}\right.---\left(1\right)$

A in formula ₁j () is an input layer jth input vector, four components of corresponding Airport Operation environment vector Env [W (t), D (t), A (t), R (t)], j=1,2,3,4, a ₂i () is hidden layer input vector, i=1,2,3,4;

W ₁(i, j) is weight vectors, represents from an input layer jth vectorial weight vectorial to hidden layer i-th, in like manner,

W ₂(k, i) represents the weight vectors of a vector from hidden layer i-th vector to output layer kth, k=1,2,3;

θ ₁(i), θ ₂k () is respectively input layer and hidden layer neuron threshold value;

U ₁(i), u ₂k () represents the output of input layer, hidden layer respectively;

Out (k) is that output layer exports, and correspondence is entered to leave the theatre the predicted value of landing scheduling result vector;

Getting excitation function is $f\left(x\right)=\frac{1}{1+e^{-x}};$

Step 102, initializes weights vector w ₁(i, j), w ₂(k, i) and neuron threshold value θ ₁(i), θ ₂(k), and output Out (k) calculating artificial neural network output layer with first group of learning data;

Step 103, oppositely revises weight vectors w according to formula (2) ₁(i, j), w ₂(k, i) and neuron threshold value θ ₁(i), θ ₂k the value of (), gets η=0.1;

$\left\{\begin{array}{c}{\mathrm{\δ}}_2^{p+1}\left(k\right)=\left({\mathrm{Flow}}^p\right(k)-{\mathrm{Out}}^p(k\left)\right)f^{\′}\left({u_2}^p\right(k\left)\right)\\ {\mathrm{\δ}}_1^{p+1}\left(k\right)=f^{\′}\left(u_2^p\right(k\left)\right){\mathrm{\δ}}_2^{p+1}\left(k\right)w_2^{p+1}(k,j)\\ w_2^{p+1}(k,i)=w_2^p(k,i)+{\mathrm{\η\δ}}_2^{p+1}{a}_2^{p+1}\left(i\right)\\ w_1^{p+1}(i,j)=w_1^p(i,j)+{\mathrm{\η\δ}}_1^{p+1}{a}_1^{p+1}\left(j\right)\end{array}\right.---\left(2\right)$

Wherein, p is iterations;

Step 104, calculate artificial neural network output layer according to formula (3) and export Out (k) and the actual variance Δ exporting Flow (k), if variance Δ is less than setting threshold value, artificial neural network can be regarded as and trained, now weight vectors w ₁(i, j), w ₂(k, i) and neuron threshold value θ ₁(i), θ ₂k the value of () is artificial neural network end value, input new Airport Operation environment vector Env [W (t), D (t), A (t), R (t)], the landing scheduling result vector Flow [F that enters to leave the theatre accordingly can be obtained _a(t), F _d(t), Δ T (t)]; Otherwise input new training data, p=p+1, get back to step 103;

$\mathrm{\Δ}={\mathrm{\Σ}}_{k=1}^3{\left(Flow\right(k)-Out(k\left)\right)}^2---\left(3\right).$

3. the flight of a kind of rule-based excavation according to claim 1 enters to leave the theatre coordinated dispatching method, it is characterized in that: the described flight cooperative scheduling model representation that enters to leave the theatre is:

Objective function:

Min Cost (t)=α _a∑ _{t ∈ T}∑ _{i ∈ N}c _i(t)+α _d∑ _{t ∈ T}∑ _{i ∈ N}c _i(t) (6) constraint condition:

$t_i^d\∈\[\underset{\‾}{T_i^d},\stackrel{\‾}{T_i^d}\],i\∈N$

$t_i^a\∈\[\underset{\‾}{T_i^a},\stackrel{\‾}{T_i^a}\],i\∈N$

$t_i^d-t_j^d\≥{\mathrm{\Δt}}^d,i,j\∈N$

$t_i^a-t_j^a\≥{\mathrm{\Δt}}^a,i,j\∈N$

$t_i^d-t_j^a\≥{\mathrm{\Δt}}^m,i,j\∈N$

N ^a(t)≤F _A(t)

N ^d(t)≤F _D(t)

Wherein, Cost (t) is landing flight total delay cost, and T is for entering to leave the theatre scheduling T.T. interval, and N enters station departure flight sum, α in scheduling time _aand α _dbe respectively and march into the arena and station departure flight delay factor, c _it delay cost that () is flight i, with for the take-off and landing time of flight i, with be respectively effective departure time scope and the landing times scope of flight i; Δ t ^d, Δ t ^a, Δ t ^mbe respectively the minimum landing personal distance time between the minimum personal distance time of taking off between adjacent two flight i and j, adjacent two flight i and j, minimum take-off and landing personal distance time between flight i and flight j, namely will be not less than the corresponding minimum safety interval time interval time of adjacent two frame flights;

N ^dt () is taken off for the Field Activity time is actual flight sortie, F _dt () is present mode of operation lower unit interval runway departure capacity,

N ^at () is Field Activity time actual landing flight sortie, F _at () is present mode of operation lower unit interval Runway Landing Capacity.

4. the flight of a kind of rule-based excavation according to claim 1 enters to leave the theatre coordinated dispatching method, it is characterized in that: described dynamic optimization algorithm specifically comprises the steps:

Step 301, station departure flight data are entered in initialization, calculate every frame flight additional cost value according to formula (7);

In formula (7), t _ifor flight estimates the landing time, T _ifor the flight planning landing time, often erect and fall flight and all can calculate an additional cost value or

Step 302, from the flight taking off the earliest or land, the additional cost value of more adjacent flight successively, adjusted according to the expectation landing time of Least-cost strategy to flight;

Step 303, each time step judges whether that new landing flight adds flight queue, and upgrades each cellular attribute, if having, flight is come flight queue tail, gets back to step 301; If no, then directly get back to step 301;

Step 304, calculate flight according to formula (8) and estimate landing time variations quadratic sum σ, as σ≤α, namely can be considered that cellular automaton reaches steady state (SS), now the expectation landing time of flight is optimum solution, otherwise returns step 301;

$\mathrm{\σ}={\mathrm{\Σ}}_{i=0}^P{(t_i^g-t_i^{g-1})}^2---\left(8\right)$

In formula, g is cellular evolutionary generation, and α is judgment threshold.