CN105225541A - Based on the method for Trajectory Prediction in short-term that blank pipe historical data is excavated - Google Patents
Based on the method for Trajectory Prediction in short-term that blank pipe historical data is excavated Download PDFInfo
- Publication number
- CN105225541A CN105225541A CN201510718268.3A CN201510718268A CN105225541A CN 105225541 A CN105225541 A CN 105225541A CN 201510718268 A CN201510718268 A CN 201510718268A CN 105225541 A CN105225541 A CN 105225541A
- Authority
- CN
- China
- Prior art keywords
- flight path
- experience
- track points
- trajectory prediction
- aircraft
- 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.)
- Pending
Links
Landscapes
- Traffic Control Systems (AREA)
Abstract
A kind of method of Trajectory Prediction in short-term excavated based on blank pipe historical data.First it extract the aircraft track data of air traffic control system record, sort out process, and similar flight path is formed flight path group; Secondly by eliminating the redundant information of each flight path of aircraft, extracting and obtaining the crucial track points flight path group that can characterize flight path group information; Then time-space relation is carried out to crucial track points flight path group, carry out cluster analysis more afterwards, form the crucial track points of experience, and then form experience flight path; The last Trajectory Prediction in short-term carrying out aircraft based on experience flight path, and utilize aircraft actual motion flight path to upgrade experience flight path.The present invention with true blank pipe data for data source, illustrate that this method can get rid of the harmful effect of flight path to Trajectory Prediction that peel off by experiment, accurate Trajectory Prediction is made according to known flight path information, may be used on the aspects such as routeing, air traffic control and spatial domain supervision, there is Trajectory Prediction accuracy high, the simple advantage of parameter request.
Description
Technical field
The invention belongs to Trajectory Prediction technical field, particularly relate to a kind of method of Trajectory Prediction in short-term excavated based on blank pipe historical data.
Background technology
Along with the development of air-transport industry, spatial domain restriction becomes one of principal element of restriction Civil Aviation ATM development, therefore adopts flight planning to be difficult to gradually meet the demands in conjunction with the way to manage of blank pipe dynamic adaptation.For tackling this problem, the researchist of US and European has carried out the research of the air traffic control system (ATCS) of future generation based on 4D flight path operational mode, i.e. U.S.'s air transport system of future generation (NextGenerationAirTransportationSystem, and single European air traffic control system (SingleEuropeanSkyATMResearch, SESAR) NextGen).
The Trajectory Prediction run based on aircraft has critical role in air traffic control automation system of new generation, and become research emphasis, Chinese scholars achieves many achievements in this field.Prevost uses the aircraft states method of estimation based on EKF filter, carries out Trajectory Prediction by current state and flight model.The people such as Chester adopt the time-to-climb table in flying quality handbook, realize aircraft flying track conjecture by Aerodynamics Model and kinematical equation.Zhu Cheng towns etc., in conjunction with blank pipe system and aircraft handling characteristics, adopt process synthesis method to realize four-dimensional flying track conjecture.In short-term Trajectory Prediction, Peng Ying etc. propose Conjecture method of dynamic flying track, dynamically infer based on loxodrome, merge the dynamic data such as radar, telegram, can infer and aircraft flight profiles and timing node information.Above method may be summarized to be three classes: (1) is based on the printenv algorithm for estimating of Kalman filter and neural network etc.; (2) based on the Forecasting Methodology of aircraft model and flight simulation; (3) based on the Trajectory Prediction method of flight intent.Method based on Kalman and neural network makes prediction accuracy be difficult to promote due to Limited information; Forecasting Methodology based on flight simulation needs a large amount of flight parameter and database support, not easily obtains; Forecasting Methodology based on flight intent not easily combines the other factors of impact flight.
Along with the rise of data mining technology, historical data is excavated as Trajectory Prediction provides support.GarielM etc. use clustering method sorting track data and eliminating extracts typical flight path after departing from flight path, monitor and prediction, but flight path information loss is more, lacks height and temporal information for aircraft.SongY extracts typical flight path and is applied to Trajectory Prediction model from history track data, improves Trajectory Prediction accuracy rate, but its method is simply and only as forecast model reference.He Wen etc. extract path rule by trajectory time cluster, exclusive PCR factor, but the temporal information only considering flight path.Bend editing distance (TimeWarpEditDistance, TWED) service times such as TangXM for similarity measurement improved K-means clustering algorithm, extract flight profile, mission profile prediction flight path from flight path cross-sectional data, but only relate to flight profile, mission profile prediction.The clustering method that LEEJG etc. propose based on flight path section excavates flight path information, obtain public sub-flight path for prediction, but computing is comparatively complicated.Clustering method in above-mentioned document all adopts the similarity measurement based on the overall situation, and in practical application, parameter is not easily determined.DuanL proposes the Spatial Clustering (LocalDensityBasedSpatialClusteringAlgorithmwithNoise, LDBSCAN) based on local density and is applied to the object detection that peels off, and solves difficult parameters with the problem determined.
The Trajectory Prediction excavated based on history track data mainly faces following problem: the other factors of (1) impact flight, as pilot's custom, weather etc., is not easily fused in prediction flight path; (2) corresponding relation is lacked between history track data; (3) based on the Trajectory Prediction of historical data, the flight path that peels off can produce interference to predicting the outcome; (4) how from track data, the data representing flight path operational mode and rule are extracted, and for Trajectory Prediction; (5) under the prerequisite reducing information loss, simplify history track data, reduce intractability.
Summary of the invention
In order to solve the problem, the object of the present invention is to provide a kind of based on historical data, merge flight influence factor (flight custom, type etc.), prediction accurately and parameter request simply based on the method for Trajectory Prediction in short-term of blank pipe historical data excavation.
In order to achieve the above object, the method for Trajectory Prediction in short-term excavated based on blank pipe historical data provided by the invention comprises the following step carried out in order:
(1) the aircraft track data of air traffic control system record is extracted and sorts out process, and similar flight path is formed the S1 stage of flight path group;
(2) redundant information of each flight path in removal process (1) gained flight path group, and extract the S2 stage of the crucial track points flight path group that can characterize each signature of flight path;
(3) utilize the crucial track points flight path group of time-space relation technology to step (2) gained to process, obtain the S3 stage of the crucial track points flight path group after time-space relation;
(4) the crucial track points flight path group after the time-space relation obtained step (3) carries out cluster analysis, forms the crucial track points of experience, and forms the S4 stage of experience flight path in order by the crucial track points of experience;
(5) according to real-time flight path information, from step (4) acquired results, select the most similar experience flight path as prediction reference, Trajectory Prediction is in short-term carried out to the aircraft of real time execution, and utilizes actual flight path to upgrade the S5 stage of corresponding experience flight path.
In step (1), described extracts the aircraft track data of air traffic control system record and sorts out process, and the method similar flight path being formed flight path group is: the aircraft track data of utilization system record comprise aircraft code, time and aircraft run between be interposed between interior information, similarity detection is carried out to aircraft flight path, the flight path of height correlation is classified as similar, forms corresponding flight path group.
In step (2), the redundant information of each flight path in described removal process (1) gained flight path group, and the method extracting the crucial track points flight path group that can characterize each signature of flight path is: suitably ignore the track points in straight part, only consider the crucial track points occurring to turn.
In step (3), the described crucial track points flight path group of time-space relation technology to step (2) gained that utilize processes, and the method obtaining the crucial track points flight path group after time-space relation is: on space-time, carry out registration to similar crucial track points flight path group by carving method when the evaluation space degree of correlation, uniform sampling.
In step (4), crucial track points flight path group after the described time-space relation obtained step (3) carries out cluster analysis, the crucial track points of formation experience, and the method forming experience flight path in order by the crucial track points of experience is: according to the cluster feature of aircraft flight path, LDBSCAN algorithm is utilized to carry out cluster analysis to step (3) gained crucial track points flight path group, the crucial track points of formation experience, and form experience flight path in an orderly manner by the crucial track points of experience.
In step (5), described according to real-time flight path information, from step (4) acquired results, select the most similar experience flight path as prediction reference, Trajectory Prediction is in short-term carried out to the aircraft of real time execution, and the method utilizing actual flight path to upgrade corresponding experience flight path is according to real-time flight path information, from step (4) acquired results, select the most similar experience flight path as prediction reference, according to the standard deviation of the crucial track points of the experience obtained as weighting standard, process is weighted based on the error between real-time flight path and experience flight path, form the Trajectory Prediction in short-term along the change of experience flight path, and utilize the deviation size between actual flight path and experience flight path to set experience flight path update scheme.
The method of Trajectory Prediction in short-term excavated based on blank pipe historical data provided by the invention can get rid of the flight path impact that peels off, and improves Trajectory Prediction precision; Prediction flight path and history flight path space-time similarity high, and parameter is easily determined.Have the following advantages: (1) flight path simplifies simultaneously, reduce data processing load; (2) with history track data for foundation, comprise impact flight various factors, as weather, flight custom etc.; (3) get rid of the informational influence that peels off, improve prediction flight path confidence level; (4) parameter is easily determined; (5) Trajectory Prediction is effective.This method can provide reference for air traffic control system, aircraft supervision.
Accompanying drawing explanation
Fig. 1 is the method flow diagram of Trajectory Prediction in short-term excavated based on blank pipe historical data provided by the invention;
Fig. 2 is a flight path group groups of samples (in rectangle frame) curve map;
Fig. 3 is the degree that peels off of object and the relation schematic diagram of its LOF;
Fig. 4 is Trajectory Prediction schematic diagram;
Fig. 5 is the experience flight path curve map of flight UPS9 near International airport, strand;
Fig. 6 is the criterion distance dygoram of flight UPS9 experience flight path;
Fig. 7 is a true flight path of flight path group 1 and corresponding prediction flight path curve map;
Fig. 8 is a true flight path of flight path group 2 and corresponding prediction flight path curve map.
Embodiment
Below in conjunction with accompanying drawing and instantiation, the method for Trajectory Prediction in short-term excavated based on blank pipe historical data provided by the invention is described in detail.
Fig. 1 is the method flow diagram of Trajectory Prediction in short-term excavated based on blank pipe historical data provided by the invention.
As shown in Figure 1, the method for Trajectory Prediction in short-term excavated based on blank pipe historical data provided by the invention comprises the following step carried out in order:
(1) the aircraft track data of air traffic control system record is extracted and sorts out process, and similar flight path is formed the S1 stage of flight path group:
Aircraft track data due to air traffic control system record is the information of whole aircraft in the whole spatial domain of real time record, and information is by sector, store by the scan period.The flight path information of all aircrafts is mixed in together, therefore, needs to extract the flight path information of aircraft, sort out process.
Discrete Data Integration is orderly time series, in order to characterize flight path by the aircraft code information comprised in the aircraft track data of utilization system record and temporal information.The flight path extracted, according to index in short-term, as 30,20,15 minutes, and in conjunction with aircraft interval index, carries out similarity detection to the flight path of aircraft, is classified as similar and forms flight path group by the flight path of height correlation.
(2) redundant information of each flight path in removal process (1) gained flight path group, and extract the S2 stage of the crucial track points flight path group that can characterize each signature of flight path:
Because the aircraft flight path of air traffic control system record exists a large amount of redundant informations, when directly utilizing the original flight path of aircraft to process, redundant information can consume a large amount of storage resources and computational resource.Therefore, before carrying out data mining, the redundant information eliminating flight path is as much as possible needed, to improve the efficiency of system.Civil aircraft normally relies on flight planning to run, and it turns near crucial way point, and other position is all in rectilinear flight state, so turning can be set to crucial track points, and flight path and then crucial track points sequence can be reduced to
and as crucial track points flight path group, n
ifor crucial flight path is counted, can information loss be reduced like this, reconstruct flight path to greatest extent, and reduce data processing amount.
Crucial track points differentiates by angle character.With flight path T
ifirst point
as first crucial track points, for point
with
form one section of great-circle track respectively, two flight path azimuthangles are respectively
for crucial track points decision threshold θ
tif,
then will
be considered as crucial track points, add K
i.θ
tshould try one's best little of fully to catch crucial track points, but also will avoid because of θ
tvalue is too small and make the point of catching too much.
(3) utilize the crucial track points flight path group of time-space relation technology to step (2) gained to process, obtain the S3 stage of the crucial track points flight path group after time-space relation:
Above-mentioned steps (1) is that similar flight path is formed flight path group, and after the crucial track points of step (2) is extracted, flight path group is characterized by crucial track points.Because Different Flight there are differences in time, and spatially also there is certain difference due to the impact of the factors such as meteorology.Therefore, before carrying out data mining for crucial track points flight path group, need the time-space relation carrying out crucial track points flight path group.
Status in flight path group between each crucial track points flight path is identical, therefore needs when carrying out time-space relation to find benchmark flight path.The present invention, first for two flight paths of aircraft, adopts spatial correlation to evaluate its spatial registration level, and utilizes corresponding registration result to unite the start time of one or two flight path.Then, then carry out registration by next flight path and the above results, and so forth, until complete the registration of all flight paths in flight path group.Then flight path in flight path group is adopted to the moment of the mode uniform sampling point of interpolation.
Fig. 2 is a flight path group groups of samples (in rectangle frame) curve map; As shown in Figure 2, round dot is flight path sampled point, and the sampled point in rectangle frame is same sampling instant, and sample line and history track data intersect the sampled point obtained and form a groups of samples.The position coordinates of sampled point can be obtained by intersection calculations.It should be noted that flight path section and sample line are all circular arc on great circle path and non-rectilinear, 2 distances are the angular distance on great circle path, need to be distinguished during calculating sampling point coordinate.
(4) the crucial track points flight path group after the time-space relation obtained step (3) carries out cluster analysis, forms the crucial track points of experience, and forms the S4 stage of experience flight path in order by the crucial track points of experience:
Certainly exist the flight path that peels off departing from flight path universal law in history track data, these flight paths will inevitably disturb Trajectory Prediction, therefore need to get rid of.In a groups of samples, the sampled point Relatively centralized of normal flight path, the sampled point departing from flight path then can become the object that peels off.The flight path sampled point that peels off can be distinguished by suitable clustering method, and be rejected.
Be gathered in in cluster according to normal sample point more, and the characteristic distributions that the flight path sampled point that peels off can depart from bunch, the density clustering method that adopts differentiates the object that peels off more.Most is representational is density-based spatial clustering (Density-BasedSpatialClusteringofApplicationswithNoise, DBSCAN) algorithm, but its parameter sensitivity and not easily determining, effect is not ideal enough.The present invention adopts LDBSCAN algorithm, this algorithm is by the factor (LocalOutlierFactor that locally peels off, LOF) combine with DBSCAN clustering algorithm, take LOF as similarity measurement, reduce the requirement of algorithm to input parameter, overcome the parameter sensitivity of DBSCAN algorithm and the shortcoming not easily determined, more feasible.
LOF, by propositions such as BreunigMM, characterizes the local of an object and to peel off degree, and the local Outlier Data utilizing LOF to detect cannot to monitor out based on distance method, is applicable to the situation that Data distribution8 is uneven.Be the distribution situation of one group of random sample data and the LOF value of each object in Fig. 3, can find out, real point object Relatively centralized, its LOF value is close to 1; The degree that peels off of star-shaped object is comparatively obvious, and its LOF value is obviously greater than 1, and the degree that peels off is more obvious, and LOF value is larger.
LDBSCAN algorithm is proposed by DuanLian, the method with the LOF of data object for the module that peels off, the object that LOF value is less than threshold value is defined as kernel object, copy DBSCAN algorithm to define direct density on this basis can reach, local density can reach and the concept such as local density is connected, and then carries out cluster with the cluster principle of similar DBSCAN algorithm.The parameter of LDBSCAN algorithm comprises b
u, t, k, k
c, wherein b
ufor the threshold value in kernel object definition procedure, k calculates the minimum neighborhood number of objects in LOF process, k
cfor the minimum neighborhood number of objects in cluster process.The crucial track points of crucial track points flight path group after the time-space relation obtain step (3) formation experience after LDBSCAN cluster analysis, arranges crucial for experience track points in an orderly manner and can obtain experience flight path.
(5) according to real-time flight path information, from step (4) acquired results, select the most similar experience flight path as prediction reference, Trajectory Prediction is in short-term carried out to the aircraft of real time execution, and utilizes actual flight path to upgrade the S5 stage of corresponding experience flight path:
According to real-time flight path information, the most similar experience flight path can be selected as prediction reference from step (4) acquired results, according to the standard deviation of the crucial track points of the experience obtained as weighting standard, based on the error between real-time flight path and experience flight path, be weighted process, form the Trajectory Prediction in short-term along the change of experience flight path.
History track data has different dispersion degrees at diverse location, corresponding experience flight path also can change in the error of diverse location thereupon, be embodied in history flight path more discrete, history flight path sampled point will be larger to the average of the spacing of the crucial track points of corresponding experience and standard deviation.If the xth groups of samples of history flight path is
the crucial track points of corresponding experience is f
x, then the standard deviation sigma of the spacing of the crucial track points of experience and sampling track points
xcan be expressed as:
Wherein d
ifor
to f
xangular distance, unit is rad.By comparing discovery, history flight path is to the variation tendency of the spacing of experience flight path and σ
xvariation tendency match, be similar to proportional relation.Such as, in Fig. 4, for two track points of arbitrary flight path
the crucial track points of experience of its same sample plane is respectively
then have:
According to this rule next track points s to known flight path
3predict:
(1) by σ
2/ d
22 '=σ
3/ d
33 ', try to achieve d
33 ';
(2) by the crucial track points s ' of experience
3d is extended along sample line direction
33 'distance, must predict track points s
3latitude coordinates;
(3) s
3moment t
3=t
2+ (t '
3-t '
2).
Can be predicted according to the flight path of current flight path state to following a period of time by above-mentioned steps.
In addition, new flight path can provide new breath for experience flight path, but the information that how to effectively utilize is guarantee experience flight path is tending towards optimum guarantee gradually.According to spatial auto-correlation principle, spatially the closer to things or phenomenon more similar.Therefore for the actual flight path that distance experience flight path is far away, less weight should be occupied; Otherwise, close together should have larger weight.Therefore, the present invention uses inverse distance weight (InverseDistanceWeighting, IDW) weight allocation of new actual flight path in flight path group is realized, process is weighted to the deviation between new actual flight path and experience flight path, and then utilize the deviation after weighting to set experience flight path update scheme, thus guarantee the stability of experience flight path.
Experimental result
The method of Trajectory Prediction in short-term excavated based on blank pipe historical data provided by the invention can be further illustrated by following experiment.
The present invention carries out Trajectory Prediction in short-term with the ADS-B track data of 1 to 31 October in 2013 near Tianjin Binhai International airport for data source and tests.Extract according to similar flight path, sort out treatment technology, from mass historical data, extract the two class track data of flight UPS9.Track data shows this flight and runs along two intersection flight paths respectively at Different periods, is designated as flight path group 1 and flight path group 2 respectively.After two flight path group data are carried out crucial track points extraction, time-space relation, cluster analysis, obtain two experience flight paths, be designated as experience flight path 1 and experience flight path 2 respectively, sequence length is respectively 383,963, specifically as shown in Figure 5.
Specific experiment optimum configurations:
(1) crucial track points is extracted: by crucial track points decision threshold θ
tbe set to 0.025rad (about 1.43 °);
(2) LDBSCAN cluster analysis: k=7, k
c=7, b
u=2, t=0.2.
Fig. 5 is the history flight path of flight UPS9 and the longitude-latitude view of experience flight path.Solid line and dotted line represent the history flight path of flight path group 1 and flight path group 2 respectively; "×" curve and "○" curve represent experience flight path 1 and experience flight path 2 respectively.Show two groups of history track data in figure and all there is the flight path that significantly peels off, and experience flight path successfully avoid the impact that departs from of the flight path that peels off, and along the area operation that history flight path density is maximum, and accurately can represent the operation details of flight path.Article two, the goodness of fit of experience flight path to history track data is respectively 0.9802,0.9939, therefore, it is possible to the operational mode of accurate description history flight path and universal law, has the similarity of height with history flight path.
As shown in Figure 6, the time in figure is relative time to the criterion distance difference curve of flight UPS9 two experience flight paths and history flight path, is namely zero moment position with the reference position of experience flight path.Contrast Fig. 5, the turn fractions dot density of flight path is comparatively large, and straight part dot density is less, especially experience flight path 2, and two end portions is kept straight on, and the two ends of corresponding standard deviation curve also show as straight line.Find by contrasting with track data, in Fig. 5, flight path group 1 historical data is polymerized (low longitude is to high longitude) in time gradually, and in corresponding Fig. 6, the criterion distance difference of experience flight path 1 declines in time gradually; In Fig. 5 flight path group 2 historical data in time be gradually polymerized-dispersion-polymerization (low latitude is to high latitude), in Fig. 6 the criterion distance difference of experience flight path 2 reduce in time-increase-reduce.To sum up, criterion distance difference curve accurately can reflect the error characteristics of experience flight path in each stage.
With two experience flight paths and standard deviation curve thereof for reference, respectively get one section of actual flight path and carry out Trajectory Prediction experiment as initial flight path, Trajectory Prediction result as shown in Figure 7,8.For ease of display, suitable dilution is carried out to the density of track points.In figure, " * " curve is experience flight path, "○" curve is an actual flight path, " △ " curve is one section of actual flight path, it can be used as initial flight path and predicts flight path according to this initial flight path, and "+" curve is the prediction flight path obtained according to initial flight path.Can be found out by Fig. 7,8, article two, predict that flight path all shows excellence on longitude, Position Latitude, prediction flight path is all identical in position, operational mode with corresponding actual flight path, also can have good accuracy at flight path turning, and visible prediction flight path has good prediction effect.
Experimental result shows, the method of Trajectory Prediction in short-term excavated based on blank pipe historical data provided by the invention can get rid of the interference of the flight path that peels off, merge flight influence factor, flight path in known flight path following a period of time can be gone out by Accurate Prediction, have the advantages that accuracy is high, parameter request is few.
Claims (6)
1. based on the method for Trajectory Prediction in short-term that blank pipe historical data is excavated, it is characterized in that, the described method of Trajectory Prediction in short-term based on the excavation of blank pipe historical data comprises the following step carried out in order:
(1) the aircraft track data of air traffic control system record is extracted and sorts out process, and similar flight path is formed the S1 stage of flight path group;
(2) redundant information of each flight path in removal process (1) gained flight path group, and extract the S2 stage of the crucial track points flight path group that can characterize each signature of flight path;
(3) utilize the crucial track points flight path group of time-space relation technology to step (2) gained to process, obtain the S3 stage of the crucial track points flight path group after time-space relation;
(4) the crucial track points flight path group after the time-space relation obtained step (3) carries out cluster analysis, forms the crucial track points of experience, and forms the S4 stage of experience flight path in order by the crucial track points of experience;
(5) according to real-time flight path information, from step (4) acquired results, select the most similar experience flight path as prediction reference, Trajectory Prediction is in short-term carried out to the aircraft of real time execution, and utilizes actual flight path to upgrade the S5 stage of corresponding experience flight path.
2. the method for Trajectory Prediction in short-term excavated based on blank pipe historical data according to claim 1, it is characterized in that: in step (1), described extracts the aircraft track data of air traffic control system record and sorts out process, and the method similar flight path being formed flight path group is: the aircraft track data of utilization system record comprise aircraft code, time and aircraft run between be interposed between interior information, similarity detection is carried out to aircraft flight path, the flight path of height correlation is classified as similar, forms corresponding flight path group.
3. the method for Trajectory Prediction in short-term excavated based on blank pipe historical data according to claim 1, it is characterized in that: in step (2), the redundant information of each flight path in described removal process (1) gained flight path group, and the method extracting the crucial track points flight path group that can characterize each signature of flight path is: suitably ignore the track points in straight part, only consider the crucial track points occurring to turn.
4. the method for Trajectory Prediction in short-term excavated based on blank pipe historical data according to claim 1, it is characterized in that: in step (3), the described crucial track points flight path group of time-space relation technology to step (2) gained that utilize processes, and the method obtaining the crucial track points flight path group after time-space relation is: on space-time, carry out registration to similar crucial track points flight path group by carving method when the evaluation space degree of correlation, uniform sampling.
5. the method for Trajectory Prediction in short-term excavated based on blank pipe historical data according to claim 1, it is characterized in that: in step (4), crucial track points flight path group after the described time-space relation obtained step (3) carries out cluster analysis, the crucial track points of formation experience, and the method forming experience flight path in order by the crucial track points of experience is: according to the cluster feature of aircraft flight path, LDBSCAN algorithm is utilized to carry out cluster analysis to step (3) gained crucial track points flight path group, the crucial track points of formation experience, and form experience flight path in an orderly manner by the crucial track points of experience.
6. the method for Trajectory Prediction in short-term excavated based on blank pipe historical data according to claim 1, it is characterized in that: in step (5), described according to real-time flight path information, from step (4) acquired results, select the most similar experience flight path as prediction reference, Trajectory Prediction is in short-term carried out to the aircraft of real time execution, and the method utilizing actual flight path to upgrade corresponding experience flight path is according to real-time flight path information, from step (4) acquired results, select the most similar experience flight path as prediction reference, according to the standard deviation of the crucial track points of the experience obtained as weighting standard, process is weighted based on the error between real-time flight path and experience flight path, form the Trajectory Prediction in short-term along the change of experience flight path, and utilize the deviation size between actual flight path and experience flight path to set experience flight path update scheme.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510718268.3A CN105225541A (en) | 2015-10-29 | 2015-10-29 | Based on the method for Trajectory Prediction in short-term that blank pipe historical data is excavated |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510718268.3A CN105225541A (en) | 2015-10-29 | 2015-10-29 | Based on the method for Trajectory Prediction in short-term that blank pipe historical data is excavated |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105225541A true CN105225541A (en) | 2016-01-06 |
Family
ID=54994468
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510718268.3A Pending CN105225541A (en) | 2015-10-29 | 2015-10-29 | Based on the method for Trajectory Prediction in short-term that blank pipe historical data is excavated |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105225541A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106485282A (en) * | 2016-10-19 | 2017-03-08 | 中国人民解放军海军航空工程学院 | Foreign peoples's flight path method for measuring similarity of space-time restriction coupling |
| CN106846919A (en) * | 2017-01-16 | 2017-06-13 | 南京航空航天大学 | A kind of four-dimensional dynamic track prediction method based on ADS B information updatings |
| CN107424441A (en) * | 2017-06-08 | 2017-12-01 | 中国民航大学 | Based on Hotelling ' s T2The airborne vehicle flight path detection of change-point and method of estimation of statistic |
| CN107832778A (en) * | 2017-10-13 | 2018-03-23 | 中国舰船研究设计中心 | A kind of same target recognition methods based on spatial synthesis similarity |
| CN108257421A (en) * | 2017-12-22 | 2018-07-06 | 广州市中南民航空管通信网络科技有限公司 | A kind of flight Trajectory Prediction method, electronic equipment and storage medium |
| CN109191922A (en) * | 2018-09-03 | 2019-01-11 | 北京航空航天大学 | A kind of extensive four-dimensional dynamic track prediction method and device |
| CN110335507A (en) * | 2019-06-12 | 2019-10-15 | 中国电子科技集团公司第二十八研究所 | Flight operation situation law analytical method based on blank pipe track big data |
| CN110807954A (en) * | 2019-11-12 | 2020-02-18 | 上海眼控科技股份有限公司 | Method and device for identifying abnormal track point |
| CN111968365A (en) * | 2020-07-24 | 2020-11-20 | 武汉理工大学 | Non-signalized intersection vehicle behavior analysis method and system and storage medium |
| CN113611158A (en) * | 2021-06-30 | 2021-11-05 | 四川大学 | Aircraft trajectory prediction and altitude deployment method based on airspace situation |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000331300A (en) * | 1999-05-24 | 2000-11-30 | Nec Software Kyushu Ltd | Air traffic flow managing system and air traffic flow managing method |
| CN101794296A (en) * | 2010-01-13 | 2010-08-04 | 中国电子科技集团公司第五十四研究所 | Excavating method based on air activity target data |
| CN102853836A (en) * | 2012-09-10 | 2013-01-02 | 电子科技大学 | Feedback weight fusion method based on track quality |
| KR101417519B1 (en) * | 2012-12-27 | 2014-07-16 | 인하대학교 산학협력단 | Air traffic control integrated system |
| CN104504939A (en) * | 2015-01-07 | 2015-04-08 | 江苏理工学院 | Aircraft track forecasting method of air traffic control system |
-
2015
- 2015-10-29 CN CN201510718268.3A patent/CN105225541A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000331300A (en) * | 1999-05-24 | 2000-11-30 | Nec Software Kyushu Ltd | Air traffic flow managing system and air traffic flow managing method |
| CN101794296A (en) * | 2010-01-13 | 2010-08-04 | 中国电子科技集团公司第五十四研究所 | Excavating method based on air activity target data |
| CN102853836A (en) * | 2012-09-10 | 2013-01-02 | 电子科技大学 | Feedback weight fusion method based on track quality |
| KR101417519B1 (en) * | 2012-12-27 | 2014-07-16 | 인하대학교 산학협력단 | Air traffic control integrated system |
| CN104504939A (en) * | 2015-01-07 | 2015-04-08 | 江苏理工学院 | Aircraft track forecasting method of air traffic control system |
Non-Patent Citations (2)
| Title |
|---|
| 吴鹍 等: "基于数据挖掘的四维飞行轨迹预测模型", 《计算机应用》 * |
| 徐涛 等: "基于航迹点法向距离的航迹聚类研究", 《***工程与电子技术》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106485282A (en) * | 2016-10-19 | 2017-03-08 | 中国人民解放军海军航空工程学院 | Foreign peoples's flight path method for measuring similarity of space-time restriction coupling |
| CN106846919A (en) * | 2017-01-16 | 2017-06-13 | 南京航空航天大学 | A kind of four-dimensional dynamic track prediction method based on ADS B information updatings |
| CN107424441B (en) * | 2017-06-08 | 2020-05-01 | 中国民航大学 | Based on Hotelling's T2Aircraft track inflection point detection and estimation method of statistics |
| CN107424441A (en) * | 2017-06-08 | 2017-12-01 | 中国民航大学 | Based on Hotelling ' s T2The airborne vehicle flight path detection of change-point and method of estimation of statistic |
| CN107832778A (en) * | 2017-10-13 | 2018-03-23 | 中国舰船研究设计中心 | A kind of same target recognition methods based on spatial synthesis similarity |
| CN107832778B (en) * | 2017-10-13 | 2021-07-13 | 中国舰船研究设计中心 | Same target identification method based on spatial comprehensive similarity |
| CN108257421A (en) * | 2017-12-22 | 2018-07-06 | 广州市中南民航空管通信网络科技有限公司 | A kind of flight Trajectory Prediction method, electronic equipment and storage medium |
| CN109191922A (en) * | 2018-09-03 | 2019-01-11 | 北京航空航天大学 | A kind of extensive four-dimensional dynamic track prediction method and device |
| CN110335507A (en) * | 2019-06-12 | 2019-10-15 | 中国电子科技集团公司第二十八研究所 | Flight operation situation law analytical method based on blank pipe track big data |
| CN110335507B (en) * | 2019-06-12 | 2021-01-26 | 中国电子科技集团公司第二十八研究所 | Flight operation situation rule analysis method based on air traffic control track big data |
| CN110807954A (en) * | 2019-11-12 | 2020-02-18 | 上海眼控科技股份有限公司 | Method and device for identifying abnormal track point |
| CN111968365A (en) * | 2020-07-24 | 2020-11-20 | 武汉理工大学 | Non-signalized intersection vehicle behavior analysis method and system and storage medium |
| CN111968365B (en) * | 2020-07-24 | 2022-02-15 | 武汉理工大学 | Non-signalized intersection vehicle behavior analysis method and system and storage medium |
| CN113611158A (en) * | 2021-06-30 | 2021-11-05 | 四川大学 | Aircraft trajectory prediction and altitude deployment method based on airspace situation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105225541A (en) | Based on the method for Trajectory Prediction in short-term that blank pipe historical data is excavated | |
| Wang et al. | A hybrid machine learning model for short-term estimated time of arrival prediction in terminal manoeuvring area | |
| Chen et al. | Travel time prediction system based on data clustering for waste collection vehicles | |
| US11501039B2 (en) | Optimizing aircraft flows at airports using data driven predicted capabilities | |
| Jacquillat et al. | Dynamic control of runway configurations and of arrival and departure service rates at JFK airport under stochastic queue conditions | |
| CN103530704B (en) | A kind of air dynamic traffic volume in terminal airspace prognoses system and method thereof | |
| JP6034130B2 (en) | Method and system for estimating aircraft parameters | |
| US20180012499A1 (en) | System and Method for Predicting Aircraft Taxi Time | |
| CN110709908A (en) | Computer system and method for state prediction of traffic system | |
| JP2015125775A (en) | System and method for multi-task learning system for prediction of demand on system | |
| US9507053B2 (en) | Using aircraft trajectory data to infer atmospheric conditions | |
| CN112561191A (en) | Prediction model training method, prediction method, device, apparatus, program, and medium | |
| Mukherjee et al. | Predicting ground delay program at an airport based on meteorological conditions | |
| Avery et al. | Data-driven modeling and prediction of the process for selecting runway configurations | |
| Achenbach et al. | Prescriptive analytics in airline operations: Arrival time prediction and cost index optimization for short-haul flights | |
| Avery et al. | Predicting airport runway configuration: A discrete-choice modeling approach | |
| Kang et al. | Improving airline fuel efficiency via fuel burn prediction and uncertainty estimation | |
| Lin et al. | Approach for 4-d trajectory management based on HMM and trajectory similarity | |
| CN109410576A (en) | Road condition analyzing method, apparatus, storage medium and the system of multisource data fusion | |
| Yang et al. | Empirical exploration of air traffic and human dynamics in terminal airspaces | |
| Yang | Practical method for 4-dimentional strategic air traffic management problem with convective weather uncertainty | |
| JP2017033198A (en) | Space-time variable prediction device and program | |
| Hrastovec et al. | Machine learning model for aircraft performances | |
| CN111047107B (en) | Road traffic time prediction method, device, electronic equipment and storage medium | |
| Wang et al. | A high-precision method of flight arrival time estimation based on xgboost |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160106 |