CN108764560A - Aircraft scene trajectory predictions method based on shot and long term Memory Neural Networks - Google Patents
Aircraft scene trajectory predictions method based on shot and long term Memory Neural Networks Download PDFInfo
- Publication number
- CN108764560A CN108764560A CN201810495952.3A CN201810495952A CN108764560A CN 108764560 A CN108764560 A CN 108764560A CN 201810495952 A CN201810495952 A CN 201810495952A CN 108764560 A CN108764560 A CN 108764560A
- Authority
- CN
- China
- Prior art keywords
- trajectory predictions
- scene
- data
- aircraft
- neural networks
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 17
- 230000007787 long-term memory Effects 0.000 title claims abstract description 9
- 238000012549 training Methods 0.000 claims abstract description 22
- 230000007774 longterm Effects 0.000 claims abstract description 8
- 238000005070 sampling Methods 0.000 claims description 19
- 238000012360 testing method Methods 0.000 claims description 10
- 238000003062 neural network model Methods 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 238000013480 data collection Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 abstract description 9
- 238000005516 engineering process Methods 0.000 abstract description 5
- 238000001514 detection method Methods 0.000 abstract description 4
- 238000004088 simulation Methods 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 7
- 230000015654 memory Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 208000025174 PANDAS Diseases 0.000 description 1
- 208000021155 Paediatric autoimmune neuropsychiatric disorders associated with streptococcal infection Diseases 0.000 description 1
- 240000000220 Panda oleosa Species 0.000 description 1
- 235000016496 Panda oleosa Nutrition 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- General Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Operations Research (AREA)
- Evolutionary Computation (AREA)
- Quality & Reliability (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of aircraft scene trajectory predictions method based on shot and long term Memory Neural Networks, LSTM neural networks are combined with polynomial fitting method to implement trajectory predictions technology, by the way that the upsample period is arranged, it can theoretically predict the position of any time in long period 60 seconds, but predicted macrocyclic position, the training data quality generated by pretreatment is low, cause precision of prediction too low, the detection of conflict is slided for scene, also without practical function, so the position of any time is relatively suitable in long-term 30 seconds in prediction.The present invention has the characteristics that historical trace by LSTM neural networks, it can be according to the context of track sets, the scene motion state of recessiveness simulation aircraft, can be used for predicting airport taxiway, on runway aircraft future time section position, aircraft scene is avoided to slide conflict, it lays the groundwork for real-time route planning, ensures airport security, efficiently runs.
Description
Technical field
The present invention relates to airport aircraft scene trajectory predictions technologies.
Background technology
With the high speed sustainable development of World Airways transport service, airdrome scene traffic is further busy, especially in airdrome scene
During traffic control, the resources such as scene space, time, manpower are not fully utilized, the aircraft slide road of initial plan
Diameter, there are still the potential security risks for sliding conflict in operational process, directly influence the operational safety and efficiency of airdrome scene.
It may determine that the geographical location of aircraft future time section using trajectory predictions technology, aircraft can also be estimated and reach critical path
The time of mouth and time departure reduce scene traffic jam, avoid sliding conflict, shorten the total coasting time of aircraft, it is ensured that
Airport Expwy operates, and improves service quality.
Currently, traditional based on trajectory predictions sides such as kinetic model, Kalman filtering algorithm, Hidden Markov Model
Method is not easy the aircraft performance parameter obtained dependent on actual motion, and is grasped by airport surface traffic control, aircraft driver in the process
Make to be intended to and the influence of meteorology, the aircraft kinetic model of foundation influence the accuracy of trajectory predictions.In addition, history data set
The state vector space of middle various dimensions, it is computationally intensive, excessive historical information cannot be preserved, medium- and long-term forecasting, shadow are unsuitable for
Ring the real-time of trajectory predictions.
Since there are very strong correlation and dependences between the position sequence of aircraft slide movement, by means of time sequence
The recessive relationship between sequential value can be analyzed and be characterized to row method, can according to the historical position sequence that aircraft slide moves
To predict the position of future time instance.And shot and long term Memory Neural Networks (Long-Short Term Memory Networks,
LSTM) be in artificial intelligence field processing time sequence problem important sharp weapon.But it can only be realized with LSTM following certain discrete
The position prediction at moment, this moment were determined by the sampling period of training data, when cannot achieve long-term arbitrary in future
The position prediction at quarter.
Invention content
The present invention provides a kind of aviation based on the LSTM neural networks with decay window that can realize medium- and long-term forecasting
Device scene trajectory predictions method.
The present invention is to solve above-mentioned technical problem the technical scheme adopted is that the boat based on shot and long term Memory Neural Networks
Pocket scene trajectory predictions method, includes the following steps:
Step 1, the history taxi data collection for obtaining aircraft, including longitude, latitude and speed data, are arranged upsample
Period pre-processes taxi data sequence, is divided into training data and test data;
Step 2, LSTM neural network model of the structure with decay window, input training data, and Configuration network parameter is completed
The training of model;
LSTM neural network models under step 3, input test data to Different sampling period obtain predicted value, anti-normalizing
After change, in addition basic item when first-order difference is handled obtains trajectory predictions position, the trajectory predictions position is by longitude and latitude
Composition;
Step 4, by the trajectory predictions position under Different sampling period, form trajectory predictions by sampling period incremental sequence
As basic point by first tracing point in trajectory predictions sequence scene phase is calculated with remaining tracing point respectively in sequence
It adjusts the distance and (for sliding curve trajectory predictions, azimuth can be further calculated), by scene relative distance and time using more
Item formula approximating method is handled, and is obtained trajectory predictions Fitting curve equation, can be obtained the scene of medium-term and long-term interior any time
Relative distance.
The position of certain discrete instants can only be individually predicted using LSTM neural networks, and discrete instants are by training number
According to sampling period determine;And individually use the method for fitting of a polynomial that can predict the position of the following any time, but miss
Difference is too big.The present invention combines LSTM neural networks with polynomial fitting method to implement trajectory predictions technology, is passed by setting
Increase the sampling period, can theoretically predict the position of any time in long period 60 seconds, but predicted macrocyclic position, passes through
It is low to pre-process the training data quality generated, causes precision of prediction too low, the detection of conflict is slided for scene, also without practical
Effect, so the position of any time is relatively suitable in long-term 30 seconds in prediction.
The invention has the advantages that have the characteristics that historical trace by LSTM neural networks, it can be according to track
The context of sequence, the scene motion state of recessiveness simulation aircraft, can be used for predicting airport taxiway, aircraft on runway
The position of future time section avoids aircraft scene from sliding conflict, lays the groundwork for real-time route planning, ensures airport security, height
The operation of effect ground.
Description of the drawings
Fig. 1 is specific implementation mode flow chart.
Fig. 2 is supervised learning sequence conversion operation flow chart.
Fig. 3 is aircraft scene motion state change schematic diagram.
Fig. 4 is polynomial fitting curve figure.
Specific implementation mode
As shown in Figure 1, for choosing certain domestic airport, implementation process is broadly divided into following steps:
Step 1, the history taxi data collection (longitude, latitude, speed) for obtaining aircraft, taxi data sequence is carried out etc.
Pretreatment away from the conversion of sampling, first-order difference, normalization and supervised learning sequence, is divided into training data and test data:
Step 1.1 chooses the historical trajectory data on a certain line sliding road according to history taxi data, using as follows
Formula:
Tk=kT, k ∈ N* (1)
Wherein k is decimation factor, N*For positive integer, T is the period of scene surveillance radar acquisition trajectories data.Equidistant sampling
N track point sequence is obtained, is expressed as:
[(x1,y1,v1),(x2,y2,v2),…,(xn,yn,vn)]
(xt,yt,vt) longitude, latitude, speed of t moment tracing point, t=1,2 ..., n are indicated respectively;
It is poor that the basic item of adjacent moment in the point sequence of track is made, i.e., first-order difference is handled;
x′t=xt+1-xt,y′t=yt+1-yt,v′t=vt+1-vt (2)
N-1 new track point sequence is obtained, is expressed as:
[(x′1,y′1,v′1),(x'2,y'2,v'2),…,(x'n-1,y'n-1,v'n-1)]
Step 1.2 is used the track point sequence after first-order difference using the sklearn tools in machine learning library
MinMaxScaler () function, which is normalized, to be zoomed in [- 1,1] range, and preserves the output parameter scaler of function, then
Data after normalization are converted into supervised learning sequence, operating process such as Fig. 2 with the pandas tools in machine learning library
Shown, n_vars indicates that the Characteristic Number of input data, n_in indicate that list entries number, n_out indicate forecasting sequence number,
Input Sequence and Forecast two data objects of Sequece, Shift are initialized using DataFrame () function
() function recycles structure list entries data and forecasting sequence label data respectively, and Concat () function is used in combination to splice, and composition is most
Whole supervised learning tracing point sequence data, i.e. training dataset, sample format are as shown in table 1.
1 supervised learning serial data format of table
xt-3 | yt-3 | vt-3 | xt-2 | yt-2 | vt-2 | xt-1 | yt-1 | vt-1 | xt | yt | vt | xt+1 | yt+1 | vt+1 |
x1 | y1 | v1 | x2 | y2 | v2 | x3 | y3 | v3 | x4 | y4 | v4 | x5 | y5 | v5 |
x2 | y2 | v2 | x3 | y3 | v3 | x4 | y4 | v4 | x5 | y5 | v5 | x6 | y6 | v6 |
x3 | y3 | v3 | x4 | y4 | v4 | x5 | y5 | v5 | x6 | y6 | v6 | x7 | y7 | v7 |
x4 | y4 | v4 | x5 | y5 | v5 | x6 | y6 | v6 | x7 | y7 | v7 | x8 | y8 | v8 |
Data set is divided into training dataset and test data set with 80%, 20% ratio, training dataset is used for
Training pattern, test data set are used for assessment models.
Step 2 builds the shot and long term Memory Neural Networks model with decay window, Configuration network parameter, training pattern:
Step 2.1 introduces shot and long term Memory Neural Networks LSTM, utilizes input gate it, forget door ftWith out gate otTo control
System increases, abandons and forgets information, and specific effect is as follows:
Input gate it:Control information increases to the degree of storage unit;
Forget door ft:The output of the storage unit of last moment and the input at current time are controlled, when passing to current together
Carve the degree that the information in storage unit abandons;
Out gate ot:Information in control current time storage unit passes to current hidden state htThe journey that middle information abandons
Degree.
In t moment, the parameter more new relation of propagated forward and back-propagating is as follows, wherein the definition of symbol such as table 2
It is shown.
it=σ (Wi·[ht-1,Xt]+bi)
ft=σ (Wf·[ht-1,Xt]+bf) (3)
ot=σ (Wo·[ht-1,Xt]+bo)
ct=ft*ct-1+it*tanh(Wc·[ht-1,Xt]+bc) (4)
ht=ot*tanh(ct)
2 LSTM network structure basic parameters of table
Step 2.2 analyzes the variation of scene aircraft motion state, as shown in figure 3, having memory according to LSTM neural networks
The characteristics of property, for example, in time period t2~t3When interior progress position prediction, t1~t2The data flow generated in period can be to working as
The disturbance for the accelerated motion state that the motion state at preceding moment creates a deceitful impression causes to predict that error increases, influences model instruction
Experienced precision;Consider to introduce Attenuation Memory Recursive mechanism, the data flow in the hidden layer of LSTM neural networks is remembered according to training set
The exponential decay of window sliding is specifically drawn in hidden layer to weaken influence of the historical data to current time motion state
Enter attenuation coefficient λ ∈ (0,1), it is proposed that value 0.9 acts on the location mode in hidden layer:
Wherein, t indicates data moment nearest in Attenuation Memory Recursive sliding window, ctIndicate the location mode of current t moment,
W indicates the length of Attenuation Memory Recursive sliding window.
Step 2.3 uses deep learning frame Keras, builds LSTM training patterns, inputs training data, uses
Iteration weight matrix in ModelCheckpoint () function reservation training process, Configuration network parameter, as shown in table 3:
3 LSTM network parameters of table configure
LSTM neural network models under step 3 input test data to Different sampling period obtain predicted value, anti-normalizing
After change, in addition basic item when first-order difference is handled obtains trajectory predictions position, the trajectory predictions position is by longitude and latitude
Composition:
Test data is inputted training pattern by step 3.1, obtains predicted valueWith the scaler renormalizations of preservation
After obtainIn addition basic item (x when first-order difference is handledl,yl) obtain predicted position:
Again by root-mean-square error (Root Mean Square Error, RMSE) assessment models, cycle executes step
2.3,3.1 until it is T to obtain predetermined period1When, all predicted positions for reaching root-mean-square error threshold value are averaged to obtain
Future position
Step 3.2 incrementally updates the sampling period, repeats steps 1 and 2,3.1, obtains Different sampling period TkUnder track it is pre-
Measuring pointK=1,2 ..., P, P indicate the total number in the sampling period of setting.
Step 4 is obtained using the future position of LSTM neural network models under polynomial fitting method processing Different sampling period
Obtain trajectory predictions Fitting curve equation:
Step 4.1 obtains the trajectory predictions sequence under the upsample period according to step 3
Choose sampling period T1Track sets pointFor basic point, and calculated and the rail under other sampling periods using following formula
Mark sequence of pointsBetween longitude and latitude scene relative distance s, k=2,3 ..., P:
Wherein, d is the median that data calculate, and r=6371km is earth radius, and experimental data is as shown in table 4.
Scene relative distance under 4 different moments of table
Step 4.2 seeks following m using the functional relation of method processing the scene relative distance s and time t of fitting of a polynomial
Order polynomial:
sm(t)=a0+a1t+…+amtm (8)
Wherein, ai, i ∈ m are multinomial coefficient, are carried out to 4 data of table using the sklearn tool storage rooms in python multinomial
Formula solves, and obtains three times, four times, five multi-form matched curves, as shown in Figure 4.The fitting effect of five powers as can be seen from Figure
Fruit is more preferable, solves polynomial function and is:
S (t)=0.00401 × t5-0.172×t4+2.53×t3-14.2×t2+34.6×t-15.6 (9)
The relative distance that scene aircraft any time t in line sliding road can be obtained by above formula, due to the example
It is line sliding road trajectory predictions, it is not necessary to computer azimuth angle;If for sliding curve trajectory predictions, according to two trajectory predictions
Longitude, the latitude of sequence of points can calculate scene relative distance and azimuth simultaneously, are more advantageous to detection scene and slide conflict.
Theoretically, by the way that upsample factor k is arranged, the position of any time in long period 60 seconds can be predicted, but pre-
Macrocyclic position was surveyed, the training data quality generated by pretreatment is low, causes precision of prediction too low, scene is slided
The detection of conflict, also without practical function, so the position of long-term (in 30 seconds) any time is relatively suitable in prediction.
Claims (5)
1. the aircraft scene trajectory predictions method based on shot and long term Memory Neural Networks, which is characterized in that include the following steps:
Step 1, the history taxi data collection for obtaining aircraft, including longitude, latitude and speed data, are arranged the upsample period
Taxi data sequence is pre-processed, training data and test data are divided into;
Step 2, structure shot and long term Memory Neural Networks LSTM neural network models, input training data, Configuration network parameter is complete
At the training of model;
LSTM neural network models under step 3, input test data to Different sampling period obtain predicted value, renormalization
Afterwards, in addition basic item when first-order difference is handled obtains trajectory predictions position, the trajectory predictions position is by longitude and latitude group
At;
Step 4, by the trajectory predictions position under Different sampling period, form trajectory predictions sequence by sampling period incremental sequence
Row, as basic point by first tracing point in trajectory predictions sequence, it is opposite to be calculated scene with remaining tracing point respectively
Distance handles scene relative distance with the time using polynomial fitting method, obtains trajectory predictions Fitting curve equation,
It can obtain the scene relative distance of medium-term and long-term interior any time.
2. method as described in claim 1, which is characterized in that introduce attenuation coefficient λ ∈ in the hidden layer of LSTM neural network models
(0,1) acts on the location mode in hidden layer:
Wherein, t indicates data moment nearest in sliding window, ctIndicate that the location mode of current t moment, W are indicated according to track
The adjustable sliding window length of Variation Features of data.
3. method as claimed in claim 2, which is characterized in that λ attenuation coefficient values are 0.9, W values are 10.
4. method as described in claim 1, which is characterized in that pretreatment includes equidistant sampling, first-order difference, normalizing in step 1
Change and supervised learning sequence is converted;
Input test data to LSTM neural network models obtain predicted value in step 3, then will be added after predicted value renormalization
Basic item when first-order difference processing obtains trajectory predictions position.
5. method as described in claim 1, which is characterized in that step 4 fitting of a polynomial is fitted using five powers.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810495952.3A CN108764560B (en) | 2018-05-22 | 2018-05-22 | Aircraft scene trajectory prediction method based on long-short term memory neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810495952.3A CN108764560B (en) | 2018-05-22 | 2018-05-22 | Aircraft scene trajectory prediction method based on long-short term memory neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108764560A true CN108764560A (en) | 2018-11-06 |
CN108764560B CN108764560B (en) | 2022-03-15 |
Family
ID=64007537
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810495952.3A Active CN108764560B (en) | 2018-05-22 | 2018-05-22 | Aircraft scene trajectory prediction method based on long-short term memory neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108764560B (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109769285A (en) * | 2019-02-18 | 2019-05-17 | 西安交通大学 | A kind of method for routing of the inter-vehicular communication based on position prediction |
CN109814101A (en) * | 2019-01-25 | 2019-05-28 | 北京邮电大学 | A kind of method and device of aircraft position prediction |
CN110232169A (en) * | 2019-05-09 | 2019-09-13 | 北京航空航天大学 | Track denoising method based on two-way length memory models and Kalman filtering in short-term |
CN110334167A (en) * | 2019-06-26 | 2019-10-15 | 浪潮软件集团有限公司 | Positional shift method for early warning and device based on neural network track |
CN110443448A (en) * | 2019-07-01 | 2019-11-12 | 华中科技大学 | A kind of aircraft seat in the plane classification prediction technique and system based on two-way LSTM |
CN110471055A (en) * | 2019-07-08 | 2019-11-19 | 岭澳核电有限公司 | Flying object trajectory predictions method, apparatus, readable storage medium storing program for executing and terminal device |
CN110795522A (en) * | 2019-11-06 | 2020-02-14 | 中国人民解放军战略支援部队信息工程大学 | Method and device for predicting track position of mobile user |
CN110826788A (en) * | 2019-10-30 | 2020-02-21 | 南京智慧航空研究院有限公司 | Airport scene variable slide-out time prediction method based on big data deep learning |
CN111027692A (en) * | 2019-11-04 | 2020-04-17 | 中国电子科技集团公司第十四研究所 | Target motion situation prediction method and device |
CN111310965A (en) * | 2019-10-30 | 2020-06-19 | 北京航空航天大学 | Aircraft track prediction method based on LSTM network |
CN111965639A (en) * | 2020-09-09 | 2020-11-20 | 扬州船用电子仪器研究所(中国船舶重工集团公司第七二三研究所) | Radar signal analysis method based on bidirectional long-and-short-term memory neural network |
CN112327903A (en) * | 2020-09-15 | 2021-02-05 | 南京航空航天大学 | Aircraft trajectory generation method based on deep mixed density network |
CN112329974A (en) * | 2020-09-03 | 2021-02-05 | 中国人民公安大学 | LSTM-RNN-based civil aviation security event behavior subject identification and prediction method and system |
CN114187783A (en) * | 2021-12-06 | 2022-03-15 | 中国民航大学 | Method for analyzing and predicting potential conflicts in airport flight area |
CN114298196A (en) * | 2021-12-22 | 2022-04-08 | 中国民航科学技术研究院 | Automatic judgment method for operation modes of multiple runways of airport |
CN117076839A (en) * | 2023-10-17 | 2023-11-17 | 中国民用航空总局第二研究所 | Airport aircraft track dynamic prediction method based on dual incremental neural network |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060155432A1 (en) * | 2005-01-07 | 2006-07-13 | United Technologies Corporation | Methods and systems for monitoring atmospheric conditions, predicting turbulent atmospheric conditions and optimizing flight paths of aircraft |
WO2016156236A1 (en) * | 2015-03-31 | 2016-10-06 | Sony Corporation | Method and electronic device |
CN106096767A (en) * | 2016-06-07 | 2016-11-09 | 中国科学院自动化研究所 | A kind of link travel time prediction method based on LSTM |
CN106571068A (en) * | 2015-01-07 | 2017-04-19 | 江苏理工学院 | Method for predicting track of aircraft based on 4D track operation |
CN107492113A (en) * | 2017-06-01 | 2017-12-19 | 南京行者易智能交通科技有限公司 | A kind of moving object in video sequences position prediction model training method, position predicting method and trajectory predictions method |
-
2018
- 2018-05-22 CN CN201810495952.3A patent/CN108764560B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060155432A1 (en) * | 2005-01-07 | 2006-07-13 | United Technologies Corporation | Methods and systems for monitoring atmospheric conditions, predicting turbulent atmospheric conditions and optimizing flight paths of aircraft |
CN106571068A (en) * | 2015-01-07 | 2017-04-19 | 江苏理工学院 | Method for predicting track of aircraft based on 4D track operation |
WO2016156236A1 (en) * | 2015-03-31 | 2016-10-06 | Sony Corporation | Method and electronic device |
CN106096767A (en) * | 2016-06-07 | 2016-11-09 | 中国科学院自动化研究所 | A kind of link travel time prediction method based on LSTM |
CN107492113A (en) * | 2017-06-01 | 2017-12-19 | 南京行者易智能交通科技有限公司 | A kind of moving object in video sequences position prediction model training method, position predicting method and trajectory predictions method |
Non-Patent Citations (2)
Title |
---|
ABDELRAHMAN ELSAID: "Using LSTM recurrent neural networks to predict excess vibration events in aircraft engines", 《2016 IEEE 12TH INTERNATIONAL CONFERENCE ON E-SCIENCE (E-SCIENCE)》 * |
翟岱亮 等: "基于自适应IMM的高超声速飞行器轨迹预测", 《航空学报》 * |
Cited By (26)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109814101A (en) * | 2019-01-25 | 2019-05-28 | 北京邮电大学 | A kind of method and device of aircraft position prediction |
CN109769285A (en) * | 2019-02-18 | 2019-05-17 | 西安交通大学 | A kind of method for routing of the inter-vehicular communication based on position prediction |
CN110232169A (en) * | 2019-05-09 | 2019-09-13 | 北京航空航天大学 | Track denoising method based on two-way length memory models and Kalman filtering in short-term |
CN110232169B (en) * | 2019-05-09 | 2022-01-04 | 北京航空航天大学 | Track denoising method based on bidirectional long-time and short-time memory model and Kalman filtering |
CN110334167B (en) * | 2019-06-26 | 2023-06-09 | 浪潮软件集团有限公司 | Position deviation early warning method and device based on neural network track |
CN110334167A (en) * | 2019-06-26 | 2019-10-15 | 浪潮软件集团有限公司 | Positional shift method for early warning and device based on neural network track |
CN110443448A (en) * | 2019-07-01 | 2019-11-12 | 华中科技大学 | A kind of aircraft seat in the plane classification prediction technique and system based on two-way LSTM |
CN110443448B (en) * | 2019-07-01 | 2022-03-29 | 华中科技大学 | Bidirectional LSTM-based airplane position classification prediction method and system |
CN110471055A (en) * | 2019-07-08 | 2019-11-19 | 岭澳核电有限公司 | Flying object trajectory predictions method, apparatus, readable storage medium storing program for executing and terminal device |
CN110471055B (en) * | 2019-07-08 | 2021-03-30 | 岭澳核电有限公司 | Method and device for predicting flight path, readable storage medium and terminal equipment |
CN111310965A (en) * | 2019-10-30 | 2020-06-19 | 北京航空航天大学 | Aircraft track prediction method based on LSTM network |
CN110826788A (en) * | 2019-10-30 | 2020-02-21 | 南京智慧航空研究院有限公司 | Airport scene variable slide-out time prediction method based on big data deep learning |
CN111027692A (en) * | 2019-11-04 | 2020-04-17 | 中国电子科技集团公司第十四研究所 | Target motion situation prediction method and device |
CN110795522B (en) * | 2019-11-06 | 2023-01-06 | 中国人民解放军战略支援部队信息工程大学 | Method and device for predicting track position of mobile user |
CN110795522A (en) * | 2019-11-06 | 2020-02-14 | 中国人民解放军战略支援部队信息工程大学 | Method and device for predicting track position of mobile user |
CN112329974A (en) * | 2020-09-03 | 2021-02-05 | 中国人民公安大学 | LSTM-RNN-based civil aviation security event behavior subject identification and prediction method and system |
CN112329974B (en) * | 2020-09-03 | 2024-02-27 | 中国人民公安大学 | LSTM-RNN-based civil aviation security event behavior subject identification and prediction method and system |
CN111965639A (en) * | 2020-09-09 | 2020-11-20 | 扬州船用电子仪器研究所(中国船舶重工集团公司第七二三研究所) | Radar signal analysis method based on bidirectional long-and-short-term memory neural network |
CN111965639B (en) * | 2020-09-09 | 2024-04-09 | 扬州船用电子仪器研究所(中国船舶重工集团公司第七二三研究所) | Radar signal analysis method based on bidirectional long-short-term memory neural network |
CN112327903A (en) * | 2020-09-15 | 2021-02-05 | 南京航空航天大学 | Aircraft trajectory generation method based on deep mixed density network |
CN114187783A (en) * | 2021-12-06 | 2022-03-15 | 中国民航大学 | Method for analyzing and predicting potential conflicts in airport flight area |
CN114187783B (en) * | 2021-12-06 | 2023-10-31 | 中国民航大学 | Method for analyzing and predicting potential conflict in airport flight area |
CN114298196A (en) * | 2021-12-22 | 2022-04-08 | 中国民航科学技术研究院 | Automatic judgment method for operation modes of multiple runways of airport |
CN114298196B (en) * | 2021-12-22 | 2022-06-17 | 中国民航科学技术研究院 | Automatic judgment method for operation modes of multiple runways of airport |
CN117076839A (en) * | 2023-10-17 | 2023-11-17 | 中国民用航空总局第二研究所 | Airport aircraft track dynamic prediction method based on dual incremental neural network |
CN117076839B (en) * | 2023-10-17 | 2023-12-26 | 中国民用航空总局第二研究所 | Airport aircraft track dynamic prediction method based on dual incremental neural network |
Also Published As
Publication number | Publication date |
---|---|
CN108764560B (en) | 2022-03-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108764560A (en) | Aircraft scene trajectory predictions method based on shot and long term Memory Neural Networks | |
CN103336863B (en) | The flight intent recognition methods of flight path observed data of flying based on radar | |
CN114048889B (en) | Aircraft trajectory prediction method based on long-term and short-term memory network | |
Lawrance et al. | Path planning for autonomous soaring flight in dynamic wind fields | |
CN111598473B (en) | Multi-platform combined task planning method for complex observation task | |
CN105988146A (en) | Application data processing method of spaceborne microwave radiometer | |
CN104049639B (en) | A kind of unmanned boat antisurge based on support vector regression controls apparatus and method | |
CN111695299A (en) | Mesoscale vortex trajectory prediction method | |
CN109242886A (en) | A kind of modeling of space cluster target trajectory and forecasting procedure | |
CN113569465B (en) | Flight path vector and target type joint estimation system and estimation method based on deep learning | |
CN114842681A (en) | Airport scene flight path prediction method based on multi-head attention mechanism | |
Yang et al. | LSTM-based deep learning model for civil aircraft position and attitude prediction approach | |
CN110532665A (en) | A kind of mobile object dynamic trajectory prediction technique under scheduled airline task | |
CN113589272A (en) | Automatic generation method for target tracking equipment on-duty log | |
CN109558934A (en) | A kind of unmanned plane interpolation moth flutters flame low-level penetration method | |
Bastas et al. | Data driven aircraft trajectory prediction with deep imitation learning | |
Han et al. | Short-term 4D trajectory prediction based on LSTM neural network | |
CN114610070A (en) | Unmanned aerial vehicle-cooperated wind power plant intelligent inspection method | |
Wu et al. | UAV Path Planning Based on Multicritic‐Delayed Deep Deterministic Policy Gradient | |
Jiang et al. | Research on method of trajectory prediction in aircraft flight based on aircraft performance and historical track data | |
CN117217089A (en) | Track prediction method based on two-way long-short-term memory network | |
CN115390584B (en) | Multi-machine collaborative searching method | |
Ravela | Mapping coherent atmospheric structures with small unmanned aircraft systems | |
Yin et al. | Context-aware Aircraft Trajectory Prediction with Diffusion Models | |
Xu et al. | Aircraft trajectory prediction for terminal airspace employing social spatiotemporal graph convolutional network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |