CN106169260A - A kind of air traffic control method - Google Patents

Abstract

The present invention relates to a kind of air traffic control method implemented by air traffic control system, described air traffic control system includes data communication module, monitors data fusion module, Airborne Terminal module, control terminal module, wherein monitor that data fusion module monitors the fusion of data and automatic dependent surveillance data for realizing air traffic control radar, provide real-time flight path information for control terminal module；Control terminal module includes the fly generation of front Lothrus apterus 4D flight path, short-term 4D flight path generation in-flight, real-time flight conflict monitoring and alarm, these four submodules of Solving Flight Conflicts 4D track optimization；The air traffic control method of said system, rely on control terminal module, process flight plan data and utilize HMM to generate 4D flight path, it is achieved the analysis of the traffic conflict that spatial domain traffic is potential, and using Model Predictive Control Theory method to provide optimal solution off-square case.The present invention can effectively prevent flight collision, improves the safety of air traffic.

Description

A kind of air traffic control method

The application is Application No.: 201510008164.3, and invention and created name is " a kind of air traffic control system Method of control ", filing date: the divisional application of the application for a patent for invention on January 7th, 2015.

Technical field

The present invention relates to a kind of air traffic control system and method, particularly relate to a kind of based on the operation of 4D flight path aerial Traffic control system and method.

Background technology

Along with fast-developing the becoming increasingly conspicuous with spatial domain resource-constrained contradiction of World Airways transport service, aloft traffic flow is close The complicated spatial domain of collection, the air traffic control mode still using flight plan to combine interval allotment gradually demonstrates that it falls behind Property, it is in particular in: (1) flight plan does not configure accurate blank pipe interval for airborne vehicle, easily causes traffic flow tactics pipe Crowded in reason, reduces spatial domain safety；(2) reckoning to flight profile, mission profile of the air traffic control automation system centered by flight plan With Trajectory Prediction low precision, cause conflict dissolution ability；(3) job of air traffic control still lays particular emphasis on the single aviation of holding Personal distance between device, is difficult to rise to traffic flow is carried out strategic Management.

4D flight path is with room and time form, in a certain airborne vehicle flight path each point locus (longitude, latitude and Highly) describing with the accurate of time, operation based on flight path refers to use " controlling the time of advent " on the way point of 4D flight path, I.e. control the airborne vehicle " time window " by specific way point.In high density spatial domain operation based on 4D flight path (Trajectory based Operation) as one of basic operating mechanism, be following to big flow, high density, closely-spaced Under the conditions of spatial domain implement a kind of effective means of management, can significantly decrease the uncertainty of airborne vehicle flight path, improve spatial domain Safety with Airport Resources and utilization rate.

The air traffic method of operation run based on flight path needs to carry out single aircraft flight path on strategic level Calculating and optimize, the traffic flow constituting many airborne vehicles is implemented collaborative and adjusts；By revising traffic flow on pre-tactical level In the flight path of indivedual airborne vehicles to solve congestion problems, and ensure the operational efficiency of all airborne vehicles in this traffic flow；And in war In art aspect, scheme is freed in prediction conflict and optimization, and from fixing manual type, airborne vehicle headway management is changed into consideration aviation The factors such as device performance, regulation rule and environment are in interior variable Separation control mode, therefore towards the operation of 4D flight path to sky Middle traffic control proposes new requirement.

Summary of the invention

The technical problem to be solved in the present invention is to be to overcome the deficiencies in the prior art, it is provided that a kind of based on the operation of 4D flight path Air traffic control method, can effectively prevent flight collision, improve the safety of air traffic.

The technical scheme realizing the object of the invention is to provide a kind of air traffic control method by air traffic control system Implementing, described air traffic control system includes Airborne Terminal module, data communication module, supervision data fusion module and pipe Terminal module processed；Monitor that data fusion module monitors the fusion of data and automatic dependent surveillance data for realizing air traffic control radar, Real-time flight path information is provided for control terminal module；

Described control terminal module includes following submodule:

Lothrus apterus 4D flight path generation module before flight, according to flight plan and the forecast data of world area forecast system, Set up airborne vehicle kinetic model, then set up flight path conflict according to flight collision Coupling point and allocate theoretical model in advance, generate boat Pocket Lothrus apterus 4D flight path；

Short-term 4D flight path generation module in-flight, according to the real-time flight path information monitoring that data fusion module provides, utilizes HMM, thus it is speculated that the airborne vehicle 4D track in following certain time window；

Real-time flight conflict monitoring and alarm module, for set up from airborne vehicle continuously dynamically to discrete conflict logic Observer, is the conflict situation that discrete observation value is expressed by the continuous dynamic mapping of Air Traffic System；When system is likely disobeyed During anti-air traffic control rules, the Hybrid dynamics behavior implementing monitoring to air traffic hybrid system, for controller provide and Time warning information；

Solving Flight Conflicts 4D track optimization module, is ensureing that system meets aircraft performance and regulation rule constraints Under, by selecting different object functions of freeing, use Model Predictive Control Theory method, calculate airborne vehicle conflict Resolution 4D boat Mark；And by data communication module, airborne vehicle conflict Resolution 4D flight path is sent to Airborne Terminal module and performs；

Described air traffic control method includes following several step:

Before step A, flight, Lothrus apterus 4D flight path generation module is according to flight plan and the forecast of world area forecast system Data, set up airborne vehicle kinetic model, and foundation flight collision Coupling point is set up flight path conflict and allocated theoretical model in advance, generates Airborne vehicle Lothrus apterus 4D flight path；

Air traffic control radar is monitored that data merge with automatic dependent surveillance data by step B, supervision data fusion module, raw Become airborne vehicle real-time flight path information and be supplied to control terminal module；The flight path of short-term 4D in-flight in control terminal module generates Module speculates the airborne vehicle 4D track in following certain time window according to airborne vehicle real-time flight path information and history flight path information；Institute State the tool according to the airborne vehicle 4D track in airborne vehicle real-time flight path information and the following certain time window of history flight path information supposition Body implementation process is as follows:

Step B6, to airborne vehicle track data pretreatment, according to acquired airborne vehicle original discrete two-dimensional position sequence x =[x₁,x₂,…,x_n] and y=[y₁,y₂,…,y_n], use first-order difference method carry out processing to it obtain new airborne vehicle from Dissipate position sequence Δ x=[Δ x₁,Δx₂,…,Δx_n-1] and Δ y=[Δ y₁,Δy₂,…,Δy_n-1], wherein Δ x_b=x_b+1- x_b,Δy_b=y_b+1-y_b(b=1,2 ..., n-1)；

Step B7, to airborne vehicle track data cluster, to process after new airborne vehicle discrete two-dimensional position sequence Δ x and Δ Y, by setting cluster number M', uses genetic algorithm for clustering to cluster it respectively；

Step B8, to cluster after airborne vehicle track data utilize HMM to carry out parameter training, by will Airborne vehicle running orbit data Δ x and Δ y after process is considered as the aobvious observation of hidden Markov models, by setting hidden state Number N ' and parameter update period ζ ', according to T' nearest position detection value and use B-W algorithm roll acquisition up-to-date hidden horse Er Kefu model parameter λ '；

Step B9, foundation HMM parameter, use Viterbi algorithm to obtain corresponding to current time observation Hidden state q；

Step B10, prediction time domain h' of passing through to set, hidden state q based on airborne vehicle current time, acquisition future time period boat Position prediction value O of pocket；

Step C, real-time flight conflict monitoring and alarm module set up from airborne vehicle the most dynamically to discrete conflict logic Observer, by the continuous dynamic mapping of Air Traffic System be discrete observation value express conflict situation；When system likely When violating air traffic control rules, the Hybrid dynamics behavior implementing monitoring to air traffic hybrid system, provide for controller Warning information timely；

Step D, Solving Flight Conflicts 4D track optimization module are ensureing that system meets aircraft performance and regulation rule about Under the conditions of bundle, by selecting different object functions of freeing, use Model Predictive Control Theory method, calculate airborne vehicle conflict solution De-4D flight path；And by data communication module, airborne vehicle conflict Resolution 4D flight path is sent to Airborne Terminal module and performs；

Step E, Airborne Terminal module receive and perform the 4D track data that control terminal module is issued.

Further, in step B, the value of described cluster number M' is 4, and the value of hidden state number N' is 3, when parameter updates Section ζ ' is 30 seconds, and T' is 10, it was predicted that time domain h' is 300 seconds.

Further, the B8 of step B specifically refers to: the flight path sequence data length owing to being obtained is dynamically change, In order to real-time tracking airborne vehicle flight path state change, it is necessary to initial flight path HMM parameter lambda '=(π, A, B) on the basis of, it is readjusted, in order to speculate the airborne vehicle position in certain moment following more accurately；Every period ζ ', depend on T' the observation (o according to up-to-date acquisition₁,o₂,…,o_T') to flight path HMM parameter lambda '=(π, A, B) carry out weight New estimation.

The B10 of step B specifically refers to: every the periodHMM parameter lambda according to up-to-date acquisition '=(π, A, B) and nearest H conception of history measured value (o₁,o₂,…,o_H), hidden state q based on airborne vehicle current time, predicted by setting Time domain h', obtains airborne vehicle position prediction value O at future time period h' at moment t.

Further, the periodIt it is 4 seconds.

Further, the specific implementation process of described step D is as follows:

Step D1, to Solving Flight Conflicts process model building: conflict Resolution flight path is considered as three sections of smooth curves of continuous print, gives Surely free the beginning and end of flight path, according to flight path restrictive condition, set up and comprise acceleration, climb or rate of descent, turning rate Multivariate optimum conflict Resolution model；

Step D2, to conflict Resolution variable bound modeling under different flying conditions: wherein t need to implement conflict Resolution boat The variable bound of pocket k can be described as: a_k(t)≤a_M、ω_k(t)≤ω_M、γ_k(t)≤γ_M, a_M、ω_M、γ_MIt is respectively maximum Acceleration, turning rate and climb or rate of descent；

Step D3, termination reference point locations P of setting airborne vehicle collision avoidance planning, collision avoidance planning control time domain Θ, track are pre- Survey time domain

Step D4, at each sampling instant, the running status current based on airborne vehicle and historical position observation sequence, obtain The numerical value of spatial domain wind field variable；

Step D5, be set in given optimizing index function on the premise of, based on cooperative collision avoidance trajectory planning thought, pass through Give different weights to each airborne vehicle and incorporate real-time wind field variable filtering numerical value, obtaining the collision avoidance rail of each airborne vehicle Mark and collision avoidance control strategy and each airborne vehicle only implement its first Optimal Control Strategy in Rolling Planning is spaced；

Step D6, in next sampling instant, repeat step D4 to D5 and free terminal until each airborne vehicle all arrives it.

Further, in step D3: terminating reference point locations P and be the next way point of airborne vehicle, collision avoidance is planned Controlling time domain Θ is 300 seconds, trajectory predictions time domainIt it is 300 seconds；

The detailed process of step D4 is as follows:

D4.1) stop position of airborne vehicle is set as track reference coordinate initial point；

D4.2) when airborne vehicle is in straight running condition and at the uniform velocity turning running status, build spatial domain wind field and linearly filter (t+ Δ t)=F (t) x (t)+w (t) and z (t)=H (t) x (t)+v (t) obtains wind field variable value, wherein Δ t table to wave pattern x Showing the sampling interval, x (t) represents the state vector of t, and z (t) represents that the observation vector of t, F (t) and H (t) represent respectively State-transition matrix and output calculation matrix, w (t) and v (t) represents that system noise vector sum measures noise vector respectively；In boat When pocket is in speed change turning running status, build spatial domain wind field nonlinear filtering wave pattern

X (t+ Δ t)=Ψ (t, x (t), u (t))+w (t), z (t)=Ω (t, x (t))+v (t) and u (t)=[ω_a(t), γ_a(t)]^T, wherein Ψ () and Ω () represents state-transition matrix and output calculation matrix, ω respectively_a(t) and γ_a(t) point Biao Shi turning rate and rate of acceleration；

D4.3) numerical value of wind field variable is obtained according to constructed Filtering Model；

The detailed process of step D5 is as follows: order

$d_{it}^2=\left|\right|P_i\left(t\right)-P_i^f|{|}_2^2={(x_{it}-x_i^f)}^2+{(y_{it}-y_i^f)}^2,$

WhereinRepresent t airborne vehicle i present position P_i(t) and next way point P_i ^fBetween distance square, P_i(t)=(x_it,y_it),So priority index of t airborne vehicle i may be set to:

$L_{it}=100\frac{d_{it}^{-2}}{\underset{i=1}{\overset{n_t}{\Σ}}{d}_{it}^{-2}},$

Wherein n_tThe airborne vehicle number of conflict is there is, from the implication of priority index, aviation in representing t spatial domain Device is the nearest apart from its terminal, and its priority is the highest；

Set optimizing index

, wherein i ∈ I (t) represent airborne vehicle code and I (t)=1,2 ..., n_t, P_i(t+s Δ t) represents that airborne vehicle is at moment (t+s Δ T) position vector, P_i ^fRepresent next way point of airborne vehicle i, u_iRepresent the optimal control sequence of airborne vehicle i to be optimized, Q_it For positive definite diagonal matrix, its diagonal element is airborne vehicle i priority index L in t_it, and

Further, the airborne vehicle Lothrus apterus 4D flight path of described step A generates in accordance with the following methods:

Step A1, the aircraft states that carries out shift and model, and according to the flying height section of airborne vehicle in flight plan, set up At the Petri network model of different legs transfer: E=, (g, G, Pre, Post are m) the airborne vehicle stage to shift mould to single airborne vehicle Type, wherein g represents that flight leg, G represent the transfer point of flight status parameter in vertical section, Pre and Post represents boat respectively To annexation before and after section and way point,Represent the mission phase residing for airborne vehicle；

Step A2, to set up airborne vehicle full flight profile, mission profile hybrid model as follows,

v_H=κ (v_CAS,Mach,h_p,t_LOC),

v_GS=μ (v_CAS,Mach,h_p,t_LOC,v_WS, α),

Wherein v_CASFor calibrated airspeed, Mach is Mach number, h_pFor pressure altitude, α is the angle of wind direction forecast and air route, v_WSFor wind speed predicted value, t_LOCFor temperature forecast value, v_HFor altitude rate, v_GSFor ground velocity；

Step A3, the mode of hybrid system emulation is used to speculate and solve flight path: to use the method for time subdivision, utilize shape State continually varying characteristic Recursive Solution any time airborne vehicle is in a certain mission phase voyage away from reference pointAnd heightWherein J₀For initial time airborne vehicle away from reference point Voyage, Δ τ is the numerical value of time window, and J (τ) is the τ moment airborne vehicle voyage away from reference point, h₀For initial time airborne vehicle away from ginseng The height of examination point, h (τ) is the τ moment airborne vehicle height away from reference point, thereby it is assumed that the 4D flight path obtaining single airborne vehicle；

Step A4, to many airborne vehicles coupling model implement Lothrus apterus allotment: reach the time in cross point in advance according to two airborne vehicles, According to air traffic control principle, the airborne vehicle 4D flight path being unsatisfactory for space requirement near cross point is carried out quadratic programming, To Lothrus apterus 4D flight path.

Further, described step B monitoring, air traffic control radar is monitored data and automatic dependent surveillance by data fusion module Data merge, and generate airborne vehicle real-time flight path information, the most in accordance with the following methods:

Step B1, by coordinate unit and time unification；

Step B2, use closest data association algorithm to be associated by the point belonging to same target, extract targetpath； Step B3, by the track data that extracts from automatic dependent surveillance system and air traffic control radar respectively from different space-time ginsengs

Examine coordinate system conversion, be registered to the space-time reference coordinate system that control terminal is unified；

Step B4, the correlation coefficient of two flight paths of calculating, if correlation coefficient is less than a certain predetermined threshold value, then it is assumed that two boats Mark is uncorrelated；Otherwise these two flight paths are correlated with, and can merge；

Step B5, relevant flight path is merged.

Further, relevant flight path is merged by described step B5, use weighting based on the sampling period to put down All algorithms, its weight coefficient determines according to sampling period and precision of information, recycling Weighted Average Algorithm by associated from Dynamic dependent surveillance flight path and air traffic control radar Track Fusion are system flight path.

Further, the specific implementation process of described step C is as follows:

Step C1, structure conflict hypersurface collection of functions based on regulation rule: set up hypersurface collection of functions in order to reflect system The contention situation of system, wherein, continuous function relevant to single airborne vehicle in conflict hypersurfaceIt is that I type is super bent Face, the continuous function relevant to two frame airborne vehiclesFor Type-II hypersurface；

Step C2, foundation are by the observer of airborne vehicle continuous state to discrete conflict situation: need to build according to control specification Vertical observer, the collision event that observation system system is passed through hypersurface and produced, in order to corresponding control decision made by controller Instruction；Observer ξ is used for the consecutive variations of aircraft position in observation system and produces collision event, claimsIt is I Type observer,For Type-II observer；

Step C3, design are from the discrete watch-dog of conflict to conflict Resolution means, and this discrete watch-dog can be described as functionWherein S is the space of observer observation vector generated, and D is the space of all decision vector d generateds；Work as observer Discrete observation vector show when a certain unexpected state occurs, send corresponding alarm at once.

The present invention has positive effect: a kind of air traffic control method of (1) present invention is at airborne vehicle real-time track During supposition, having incorporated the impact of random factor, it is extraneous random that the rolling track used speculates that scheme can be extracted in time The changing condition of factor, improves the accuracy that airborne vehicle track speculates.

(2) a kind of air traffic control method of the present invention is during airborne vehicle conflict Resolution, has incorporated high-altitude wind field Impact, the rolling used frees trajectory planning scheme and can adjust in time according to the change of wind field in high-altitude and free track, Improve the robustness of airborne vehicle conflict Resolution.

(3) a kind of air traffic control method of the present invention is that airborne vehicle configures accurate blank pipe interval, strictly controls boat The pocket time window by way point, reduces traffic flow randomness, improves spatial domain safety.

(4) a kind of air traffic control method of the present invention is high to reckoning and the Trajectory Prediction precision of flight profile, mission profile, and then Conflict dissolution ability and automatization level are improved, reduces the workload of controller.

(5) a kind of air traffic control method of the present invention is no longer limited to keep between the safety between single airborne vehicle Every, but from the traffic flow in spatial domain is macroscopically implemented effectively control, control work can more be transferred to airborne vehicle and rise Fly the moment, sequence of marching into the arena, vile weather change the aspects such as boat.

(6) a kind of air traffic control method airborne vehicle based on different performance index optimum of the present invention free flight path can To significantly increase the economy that airborne vehicle runs, and the utilization rate in spatial domain.

Accompanying drawing explanation

Fig. 1 is the composition schematic diagram of the air traffic control system of the present invention；

Fig. 2 is that Airborne Terminal module forms schematic diagram；

Fig. 3 is that data communication module forms schematic diagram；

Fig. 4 is for monitoring data fusion module composition schematic diagram；

Fig. 5 generates method flow schematic diagram for Lothrus apterus 4D flight path before flight；

Fig. 6 is short-term 4D flying track conjecture method flow schematic diagram in-flight；

Fig. 7 is airborne vehicle flight path conflict monitoring and alarm method schematic flow sheet；

Fig. 8 is that 4D route optimization method schematic flow sheet freed by airborne vehicle.

Detailed description of the invention

(embodiment 1)

The air traffic control system run based on 4D flight path of the present embodiment, as it is shown in figure 1, include Airborne Terminal module 101, data communication module 102, supervision data fusion module 103 and control terminal module 104.Concrete to each several part below Embodiment is described in detail respectively.

1. Airborne Terminal module

Airborne Terminal module 101 is that pilot obtains ground control order, reference 4D flight path, and input flight intent Interface, the most still gathers the interface of current aerospace device position data.

As in figure 2 it is shown, its specific embodiments is as follows:

Airborne Terminal module 101 receives following information input: (1) ADS-B information acquisition unit 201 passes through Airborne GPS The aircraft position gathered is vectorial, velocity vector, and the catchword of this airborne vehicle, passes to machine by information and data after coding Carry data communication module 102；(2) airborne vehicle driver needs the flight intent inconsistent with ground control order, passes through people Machine inputting interface, and agreement ground controller can with identify form pass to airborne data communication by information and data Module 102.Additionally Airborne Terminal module 101 realizes following information output: (1) passes through terminal display, receives and shows The air traffic control instruction that pilot can identify；(2) the Lothrus apterus 4D boat that generate front with the flight of explicitly facial canal terminal is received Mark, and when the optimum calculated after ground line end-probing to conflict frees 4D flight path.

2. data communication module

Data communication module 102 can realize vacant lot bidirectional data communication, it is achieved airborne real time position data and flight intent The downlink transfer of data cell 202 and ground control command unit 203, and the uplink with reference to 4D flight path unit 204.

As it is shown on figure 3, its specific embodiments is as follows:

Downlink data communication: Airborne Terminal 101 passes through airborne secondary radar answering machine by aircraft identification mark and 4D position Confidence ceases, and other additional datas, as the information such as flight intent, flight speed, meteorology are transferred to ground secondary radar (SSR), secondary radar receive after data message is resolved, and be transferred to central data process assembly 301 decode, by refer to Track data interface is made to be transferred to control terminal 104；Upstream data communication: ground control terminal 104 is by instruction track data Interface, after central data processes assembly 301 coding, the inquisitor of ground secondary radar just ground control order or reference 4D Flight path information is transmitted and shows at Airborne Terminal 101.

3. monitor data fusion module

Monitor that data fusion module 103 realizes air traffic control radar and monitors the fusion with automatic dependent surveillance ADS-B data, for pipe The flight path of short-term 4D in-flight in terminal module 104 processed generates submodule and real-time flight conflict monitoring provides with alarm submodule Flight path information in real time.

As shown in Figure 4, its specific embodiments is as follows:

(1) at pretreatment stage by coordinate unit and time unification, it is assumed that extract from ADS-B and air traffic control radar respectively Data are the coordinate (such as longitude, latitude, height above sea level) of series of discrete point, each point correspondence acquisition time；(2) use closest The point belonging to same target is associated by data association algorithm, extracts targetpath；(3) will be respectively from ADS-B and blank pipe thunder The track data reaching middle extraction converts from different space-time reference coordinate system, is registered to the unified space-time of control terminal with reference to sitting Mark system；(4) correlation coefficient of two flight paths is calculated, if correlation coefficient is less than a certain predetermined threshold value, then it is assumed that two flight paths are not Relevant, otherwise these two flight paths are correlated with, and can merge；(5) relevant flight path is merged.Due to ADS-B and blank pipe The precision of radar is different with the sampling period, native system use Weighted Average Algorithm based on the sampling period, its weight coefficient according to Sampling period and precision of information determine, recycling Weighted Average Algorithm is by associated ADS-B flight path and air traffic control radar flight path It is fused to system flight path.

4. control terminal module

Control terminal module 104 include flying front Lothrus apterus 4D flight path generate, in-flight short-term 4D flight path generate, fly in real time Row conflict monitoring and alarm, these four submodules of Solving Flight Conflicts 4D track optimization.

(1) before flight, Lothrus apterus 4D flight path generates

The flight plan obtained according to Flight Data Processing System (FDP) and world area forecast system (WAFS) are issued Wind, the GRIB lattice point forecast data of temperature, set up the hybrid model of stratification to Air Traffic System, by system in peace The evolution of total state, describes the time locus of state evolution, generates airborne vehicle flight path.

As it is shown in figure 5, its specific implementation process is as follows:

First, aircraft states transfer modeling is carried out.Airborne vehicle shows as between leg dynamic along the process of track flight State handoff procedure, according to the flying height section of airborne vehicle in flight plan, sets up single airborne vehicle in the transfer of different legs (g, G, Pre, Post, m) be airborne vehicle stage metastasis model to Petri network model: E=, and wherein g represents that flight leg, G represent vertical Flight status parameter in straight section (include air speed, highly, configuration) transfer point, Pre and Post represents leg and air route respectively To annexation before and after Dian,Represent the mission phase residing for airborne vehicle.

Secondly, airborne vehicle full flight profile, mission profile hybrid model is set up.Airborne vehicle flight in single leg is considered as even Continuous process, according to particle energy model, derivation airborne vehicle the different operation phase with meteorological condition under airborne vehicle kinetics Equation, v_H=κ (v_CAS,Mach,h_p,t_LOC), v_GS=μ (v_CAS,Mach,h_p,t_LOC,v_WS, α), wherein v_CASFor calibrated airspeed, Mach is Mach number, h_pFor pressure altitude, α is the angle of wind direction forecast and air route, v_WSFor wind speed predicted value, t_LOCPre-for temperature Report value, v_HFor altitude rate, v_GSFor ground velocity.

Then, the mode using hybrid system emulation speculates and solves flight path.Use the method for time subdivision, utilization state Continually varying characteristic Recursive Solution any time airborne vehicle is in a certain mission phase voyage away from reference pointAnd heightWherein J₀For initial time airborne vehicle away from reference point Voyage, Δ τ is the numerical value of time window, and J (τ) is the τ moment airborne vehicle voyage away from reference point, h₀For initial time airborne vehicle away from ginseng The height of examination point, h (τ) is the τ moment airborne vehicle height away from reference point, thereby it is assumed that the 4D flight path obtaining single airborne vehicle.

Finally, many airborne vehicles coupling model is implemented Lothrus apterus allotment.Reach the time in cross point according to two airborne vehicles in advance, press According to air traffic control principle, the airborne vehicle 4D flight path being unsatisfactory for space requirement is carried out quadratic programming, obtain near cross point Lothrus apterus 4D flight path.

(2) short-term 4D flight path generates in-flight

The real-time track data of airborne vehicle is obtained after implementing to merge according to control radar and automatic dependent surveillance system ADS-B, Utilize HMM, thus it is speculated that the airborne vehicle 4D track in following 5 minutes windows.

As shown in Figure 6, its specific implementation process is as follows:

First, to airborne vehicle track data pretreatment, according to acquired airborne vehicle original discrete two-dimensional position sequence x= [x₁,x₂,…,x_n] and y=[y₁,y₂,…,y_n], use first-order difference method discrete to its airborne vehicle carrying out processing acquisition new Position sequence Δ x=[Δ x₁,Δx₂,…,Δx_n-1] and Δ y=[Δ y₁,Δy₂,...,Δy_n-1], wherein Δ x_b=x_b+1-x_b, Δy_b=y_b+1-y_b(b=1,2 ..., n-1).

Secondly, airborne vehicle track data is clustered.To airborne vehicle discrete two-dimensional position sequence Δ x and Δ y new after processing, By setting cluster number M', genetic algorithm for clustering is used respectively it to be clustered.

Then, HMM is utilized to carry out parameter training the airborne vehicle track data after cluster.At inciting somebody to action Airborne vehicle running orbit data Δ x and Δ y after reason is considered as the aobvious observation of hidden Markov models, by setting hidden status number Mesh N' and parameter update period ζ ', according to T' nearest position detection value and use B-W algorithm to roll the up-to-date hidden Ma Er of acquisition Section's husband's model parameter λ ': the flight path sequence data length owing to being obtained is dynamically change, in order to real-time tracking airborne vehicle navigates Mark state change, it is necessary to initial flight path HMM parameter lambda '=(π, A, B) on the basis of it is adjusted again Whole, in order to speculate the airborne vehicle position in certain moment following more accurately.Every period ζ ', according to T' observation of up-to-date acquisition Value (o₁,o₂,…,o_T') to flight path HMM parameter lambda '=(π, A, B) reappraise.

Again and, according to HMM parameter, use Viterbi algorithm to obtain corresponding to current time observation Hidden state q.

Finally, every the periodHMM parameter lambda according to up-to-date acquisition '=(π, A, B) and nearest H Conception of history measured value (o₁,o₂,…,o_H), hidden state q based on airborne vehicle current time, by setting prediction time domain h', at moment t Obtain airborne vehicle position prediction value O at future time period h'.

The value of described cluster number M' is 4, and the value of hidden state number N' is 3, and it is 30 seconds that parameter updates period ζ ', and T' is 10, it was predicted that time domain h' is 300 seconds, the periodIt it is 4 seconds.

(3) real-time flight conflict monitoring and alarm

When system likely the state violating safe condition collection occurs, implement condition monitoring by controller, to aviation Effective measure of control implemented by device, it is to avoid the generation of flight collision.

As it is shown in fig. 7, its specific implementation process is as follows:

First, conflict hypersurface collection of functions based on regulation rule is constructed.The violation of air traffic control constraint can It is considered as controlled device (the multi rack airborne vehicle of the control zone flight) event that composition system is passed through hypersurface and produced, sets up super bent Surface function collection is in order to reflect the contention situation of system.Wherein, relevant to single airborne vehicle in conflict hypersurface continuous functionIt is I type hypersurface, and by the continuous function relevant to two frame airborne vehiclesSuper bent for Type-II Face.

Then, set up by the observer of airborne vehicle continuous state to discrete conflict situation.Need to set up according to control specification Observer, the collision event that observation system system is passed through hypersurface and produced, in order to controller is made corresponding control decision and referred to Order.Observer ξ is used for the consecutive variations of aircraft position in observation system and produces collision event, claimsIt it is I type Observer,For Type-II observer.

Finally, design is from the discrete watch-dog of conflict to conflict Resolution means.When the discrete observation vector of observer shows When a certain unexpected state occurs, send corresponding alarm at once.This discrete watch-dog can be described as functionIts Middle S is the space of observer observation vector generated, and D is the space of all decision vector d generateds.

(4) Solving Flight Conflicts 4D track optimization

Under conditions of ensureing to make system meet control specification, by selecting different object functions of freeing, use Excellent control theory method so that the control input that controller is given can reach optimum.

As shown in Figure 8, its specific implementation process is as follows:

Step D1, to Solving Flight Conflicts process model building: conflict Resolution flight path is considered as three sections of smooth curves of continuous print, gives Surely free the beginning and end of flight path, according to flight path restrictive condition, set up and comprise acceleration a_i(t), climb or rate of descent γ_i (t), turning rate ω_iThe multivariate optimum conflict Resolution model of (t).

Step D2, to conflict Resolution variable bound modeling under different flying conditions: wherein t need to implement conflict Resolution boat The variable bound of pocket k can be described as: a_k(t)≤a_M、ω_k(t)≤ω_M、γ_k(t)≤γ_M, a_M、ω_M、γ_MIt is respectively maximum Acceleration, turning rate and climb or rate of descent.

Step D3, termination reference point locations P of setting airborne vehicle collision avoidance planning, collision avoidance planning control time domain Θ, track are pre- Survey time domainTerminating reference point locations P and be the next way point of airborne vehicle, collision avoidance planning control time domain Θ is 300 seconds, rail Mark prediction time domainIt it is 300 seconds.

Step D4, at each sampling instant t, the running status current based on airborne vehicle and historical position observation sequence, obtain Taking the numerical value of spatial domain wind field variable, its detailed process is as follows:

D4.1) stop position of airborne vehicle is set as track reference coordinate initial point；

D4.2) when airborne vehicle is in straight running condition and at the uniform velocity turning running status, build spatial domain wind field and linearly filter (t+ Δ t)=F (t) x (t)+w (t) and z (t)=H (t) x (t)+v (t) obtains wind field variable value, wherein Δ t table to wave pattern x Showing the sampling interval, x (t) represents the state vector of t, and z (t) represents that the observation vector of t, F (t) and H (t) represent respectively State-transition matrix and output calculation matrix, w (t) and v (t) represents that system noise vector sum measures noise vector respectively；In boat When pocket is in speed change turning running status, build spatial domain wind field nonlinear filtering wave pattern

X (t+ Δ t)=Ψ (t, x (t), u (t))+w (t), z (t)=Ω (t, x (t))+v (t) and u (t)=[ω_a(t), γ_a(t)]^T, wherein Ψ () and Ω () represents state-transition matrix and output calculation matrix, ω respectively_a(t) and γ_a(t) point Biao Shi turning rate and rate of acceleration；

D4.3) numerical value of wind field variable is obtained according to constructed Filtering Model.

Step D5, be set in given optimizing index function on the premise of, based on cooperative collision avoidance trajectory planning thought, pass through Give different weights to each airborne vehicle and incorporate real-time wind field variable filtering numerical value, obtaining the collision avoidance rail of each airborne vehicle Mark and collision avoidance control strategy and each airborne vehicle only implement its first Optimal Control Strategy, detailed process in Rolling Planning is spaced As follows: order

WhereinRepresent t airborne vehicle i present position P_i(t) and next way point P_i ^fBetween distance square, P_i(t)=(x_it,y_it),So priority index of t airborne vehicle i may be set to:

$L_{it}=100\frac{d_{it}^{-2}}{\underset{i=1}{\overset{n_t}{\Σ}}{d}_{it}^{-2}},$

Wherein n_tThe airborne vehicle number of conflict is there is, from the implication of priority index, aviation in representing t spatial domain Device is the nearest apart from its next way point, and its priority is the highest.

Set optimizing index

, wherein i ∈ I (t) represent airborne vehicle code and I (t)=1,2 ..., n_t, P_i(t+s Δ t) represents that airborne vehicle is at moment (t+s Δ T) position vector, P_i ^fRepresent next way point of airborne vehicle i, u_iRepresent the optimal control sequence of airborne vehicle i to be optimized, Q_it For positive definite diagonal matrix, its diagonal element is airborne vehicle i priority index L in t_it, and

Step D6, in next sampling instant, repeat step D4 to D5 and free terminal until each airborne vehicle all arrives it.

Airborne Terminal module receives and performs the 4D track data that control terminal module is issued.

Obviously, above-described embodiment is only for clearly demonstrating example of the present invention, and not to the present invention The restriction of embodiment.For those of ordinary skill in the field, can also be made it on the basis of the above description The change of its multi-form or variation.Here without also cannot all of embodiment be given exhaustive.And these belong to this What bright spirit was extended out obviously changes or changes among still in protection scope of the present invention.

Claims (1)

1. an air traffic control method is implemented by air traffic control system, and described air traffic control system includes airborne Terminal module, data communication module, supervision data fusion module and control terminal module；Monitor that data fusion module is for real Existing air traffic control radar monitors the fusion of data and automatic dependent surveillance data, provides real-time flight path information for control terminal module；Its It is characterised by:

Described control terminal module includes following submodule:

Lothrus apterus 4D flight path generation module before flight, according to flight plan and the forecast data of world area forecast system, sets up Airborne vehicle kinetic model, then sets up flight path conflict according to flight collision Coupling point and allocates theoretical model in advance, generate airborne vehicle Lothrus apterus 4D flight path；

Short-term 4D flight path generation module in-flight, according to the real-time flight path information monitoring that data fusion module provides, utilizes hidden horse Er Kefu model, thus it is speculated that the airborne vehicle 4D track in following certain time window；

Real-time flight conflict monitoring and alarm module, for setting up the dynamic observation to discrete conflict logic from airborne vehicle Device, is the conflict situation that discrete observation value is expressed by the continuous dynamic mapping of Air Traffic System；When system likely violates sky During middle traffic control rule, the Hybrid dynamics behavior implementing monitoring to air traffic hybrid system, provide timely for controller Warning information；

Solving Flight Conflicts 4D track optimization module, is ensureing that system meets under aircraft performance and regulation rule constraints, By selecting different object functions of freeing, use Model Predictive Control Theory method, calculate airborne vehicle conflict Resolution 4D flight path； And by data communication module, airborne vehicle conflict Resolution 4D flight path is sent to Airborne Terminal module and performs；

Described air traffic control method includes following several step:

Before step A, flight, Lothrus apterus 4D flight path generation module is according to flight plan and the forecast data of world area forecast system, Set up airborne vehicle kinetic model, and foundation flight collision Coupling point is set up flight path conflict and allocated theoretical model in advance, generates aviation Device Lothrus apterus 4D flight path；

Air traffic control radar is monitored that data merge with automatic dependent surveillance data by step B, supervision data fusion module, generates boat Pocket real-time flight path information is also supplied to control terminal module；The flight path generation module of short-term 4D in-flight in control terminal module The airborne vehicle 4D track in following certain time window is speculated according to airborne vehicle real-time flight path information and history flight path information；Described depend on Concrete reality according to the airborne vehicle 4D track in airborne vehicle real-time flight path information and the following certain time window of history flight path information supposition Execute process as follows:

Step B6, to airborne vehicle track data pretreatment, according to acquired airborne vehicle original discrete two-dimensional position sequence x= [x₁,x₂,…,x_n] and y=[y₁,y₂,…,y_n], use first-order difference method discrete to its airborne vehicle carrying out processing acquisition new Position sequence △ x=[△ x₁,△x₂,…,△x_n-1] and △ y=[△ y₁,△y₂,…,△y_n-1], wherein △ x_b=x_b+1-x_b, △y_b=y_b+1-y_b, b=1,2 ..., n-1 "；

Step B7, airborne vehicle track data is clustered, to new airborne vehicle discrete two-dimensional position sequence △ x and △ y after processing, logical Cross setting cluster number M', use genetic algorithm for clustering respectively it to be clustered；

Step B8, to cluster after airborne vehicle track data utilize HMM to carry out parameter training, by will process After airborne vehicle running orbit data △ x and △ y be considered as the aobvious observation of hidden Markov models, by setting hidden state number N' and parameter update period ζ ', according to T' nearest position detection value and use B-W algorithm to roll the up-to-date hidden Ma Erke of acquisition Husband's model parameter λ '；

Step B9, according to HMM parameter, use that Viterbi algorithm obtains corresponding to current time observation is hidden State q；

Step B10, by set prediction time domain h', hidden state q based on airborne vehicle current time, obtain future time period airborne vehicle Position prediction value O；

Step C, real-time flight conflict monitoring set up the dynamic sight to discrete conflict logic from airborne vehicle with alarm module Survey device, be the conflict situation that discrete observation value is expressed by the continuous dynamic mapping of Air Traffic System；When system is likely violated During air traffic control rules, the Hybrid dynamics behavior implementing monitoring to air traffic hybrid system, provide timely for controller Warning information；

Step D, Solving Flight Conflicts 4D track optimization module are ensureing that system meets aircraft performance and regulation rule constraint bar Under part, by selecting different object functions of freeing, use Model Predictive Control Theory method, calculate airborne vehicle conflict Resolution 4D Flight path；And by data communication module, airborne vehicle conflict Resolution 4D flight path is sent to Airborne Terminal module and performs；

Step E, Airborne Terminal module receive and perform the 4D track data that control terminal module is issued；

The specific implementation process of described step C is as follows:

Step C1, structure conflict hypersurface collection of functions based on regulation rule: set up hypersurface collection of functions in order to reflect system Contention situation, wherein, continuous function relevant to single airborne vehicle in conflict hypersurfaceIt is I type hypersurface, with The continuous function that two frame airborne vehicles are relevantFor Type-II hypersurface；

Step C2, foundation are by the observer of airborne vehicle continuous state to discrete conflict situation: need to set up according to control specification to see Survey device, the collision event that observation system system is passed through hypersurface and produced, in order to corresponding control decision instruction made by controller； Observer ξ is used for the consecutive variations of aircraft position in observation system and produces collision event, claimsIt is that I type is seen Survey device,For Type-II observer；

Step C3, design are from the discrete watch-dog of conflict to conflict Resolution means, and this discrete watch-dog can be described as functionWherein S is the space of observer observation vector generated, and D is the space of all decision vector d generateds；Work as observer Discrete observation vector show when a certain unexpected state occurs, send corresponding alarm at once.