CN112232567A - Method for predicting influence probability of unknown intention flight path on air route - Google Patents

Method for predicting influence probability of unknown intention flight path on air route Download PDF

Info

Publication number
CN112232567A
CN112232567A CN202011115262.4A CN202011115262A CN112232567A CN 112232567 A CN112232567 A CN 112232567A CN 202011115262 A CN202011115262 A CN 202011115262A CN 112232567 A CN112232567 A CN 112232567A
Authority
CN
China
Prior art keywords
probability
course
flight
influence
route
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
Application number
CN202011115262.4A
Other languages
Chinese (zh)
Inventor
傅永强
王也
毛继志
朱永文
王长春
何魏巍
唐治理
仝佳露
谢华
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing Smart Aviation Research Institute Co ltd
Original Assignee
Nanjing Smart Aviation Research Institute Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Nanjing Smart Aviation Research Institute Co ltd filed Critical Nanjing Smart Aviation Research Institute Co ltd
Priority to CN202011115262.4A priority Critical patent/CN112232567A/en
Publication of CN112232567A publication Critical patent/CN112232567A/en
Pending legal-status Critical Current

Links

Images

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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
    • G06F18/24155—Bayesian classification
    • 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
    • G06Q50/10—Services
    • G06Q50/26—Government or public services

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Artificial Intelligence (AREA)
  • Development Economics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Marketing (AREA)
  • Educational Administration (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Health & Medical Sciences (AREA)
  • Probability & Statistics with Applications (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a prediction method of influence probability of an unknown flight path on an airway, which comprises the steps of preprocessing an acquired flight data sample, selecting a candidate input variable with high correlation contribution as an input characteristic variable of a prediction model by analyzing a correlation coefficient between a possible candidate input variable and a possible output variable of the sample, fusing a plurality of course difference historical data by adopting an AR model, and giving an estimated value of course change at the next moment; taking the estimated value output by the AR model, the differential data of the current course and the differential data of the current speed as input, and adopting a naive Bayes classifier to process the course state at the next moment
Figure DDA0002728807480000011
Carrying out probability prediction; finally according to the third stepHeading state of next moment of measured flight trajectory of unknown intention
Figure DDA0002728807480000012
And calculating the predicted probability of the influence of the unknown flight trajectory on the specified route. The method can effectively predict the influence prediction probability of the aircraft with unknown intention on the flight route of the civil aircraft as early warning judgment.

Description

Method for predicting influence probability of unknown intention flight path on air route
Technical Field
The invention belongs to the technical field of aviation control, and particularly relates to a method for predicting influence probability of an unknown-intention flight trajectory on an airway.
Background
In recent years, the flying amount of aircrafts with unknown intentions in ocean areas displayed by radar screens in the tri-regional control areas is continuously increased, so that the original activity range and the original flying range of civil aircrafts are limited and influenced, and the control workload is greatly improved. The flight plan of a certain type of aircraft is not available, and the track formed by the great uncertainty and the uncertainty of the flight activity is called an unknown intention flight track, which poses serious threats to the civil aviation flight safety. In order to accurately evaluate the influence of such an aircraft, it is necessary to predict its unknown intended flight trajectory, and further study the predicted probability of the influence of each trajectory on the route flight.
Systems based on track prediction algorithms are constantly being researched and developed. Typically, in an integrated system environment, a pilot and a controller jointly determine the future trajectory of an aircraft, which must also be followed when the intent is clear. The intent information enables uncertainty in future operation of the aircraft to be reduced, and the extraction of the intent information, also known as intent prediction, is of widespread interest. The traditional Intent prediction Algorithm (Intent Inference Algorithm, IIA) uses trajectory correlation to determine which Intent model best describes the true aircraft Intent by a discrete set of aircraft Intent models. The IIA algorithm may smoothly increase the amount of intent information obtained from pilot behavior, flight plans, and environmental information. However, the method has disadvantages that the prediction delay may occur due to the complexity of the method; secondly, because the method mainly plans the track prediction, the current model of the aircraft cannot be better tracked, and the prediction precision is possibly reduced.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a method for predicting the influence probability of an unknown flight path on an air route.
The invention is realized by the following technical scheme:
a method for predicting influence probability of an unknown-intention flight trajectory on an airway includes the following steps:
the method comprises the following steps: firstly, preprocessing an acquired flight data sample, and homogenizing the sampling time interval of the data sample;
step two: selecting the input variable to be selected with high correlation contribution as the input characteristic variable of the prediction model by analyzing the correlation coefficient between the possible input variable to be selected and the output variable of the sample, wherein the selected input characteristic variable of the prediction model comprises the following steps: historical differential sequence data of the heading, differential data of the current speed and differential data of the current heading;
step three: fusing a plurality of course difference historical data by adopting an AR model, and giving an estimated value of course change at the next moment; taking the estimated value output by the AR model, the differential data of the current course and the differential data of the current speed as input, and adopting a naive Bayes classifier to process the course state at the next moment
Figure BDA0002728807460000021
Carrying out probability prediction;
step four: according to the next moment of the unknown flight trajectory predicted in the third stepTo the state
Figure BDA0002728807460000022
Calculating the predicted probability of the influence of the unknown intended flight trajectory on the designated route:
course state of aircraft at next moment
Figure BDA0002728807460000023
Under the condition of (1), its course change angle is Delta SkThe state of probability distribution of, i.e.
Figure BDA0002728807460000024
Assuming that the flight path is of a kind not known to be intended
Figure BDA0002728807460000025
The heading transition condition under the flight intention follows normal distribution, namely:
Figure BDA0002728807460000026
wherein the mean and standard deviation parameters in the probability density distribution function are derived from sample statistics in the training data set, i.e.:
Figure BDA0002728807460000027
Figure BDA0002728807460000028
if the aircraft flies in a predicted state in the next period of time, and if the flight track with unknown intention passes through a known civil aviation route at the next certain time, the aircraft is considered to possibly influence the civil aviation route to be evaluated in the next period of time;
according to the hypothesis, when the changed course is in an opening angle formed by the current point and a connecting line of the air route to be evaluated, the flight of the aircraft can possibly influence the civil aviation route;
setting the course angle formed by the connection line of the current track position and two end points of the route as alpha respectively1And alpha2The current course angle is alpha0Then, the probability that the unknown intended flight trajectory affects the civil aviation route is:
Figure BDA0002728807460000031
the predicted probability of the effect of the unintended aircraft on the designated route is shown as,
Figure BDA0002728807460000032
wherein the content of the first and second substances,
Figure BDA0002728807460000033
indicating when the Bayes classifier predicts a state of
Figure BDA0002728807460000034
The probability of the time of day is,
Figure BDA0002728807460000035
belonging to the sequence of left steering L, straight driving R and right steering St.
In the above technical solution, in the first step, the sampling time interval of the data samples is homogenized by using a linear interpolation method.
In the above technical solution, the possible input variables to be selected of the sample include: the speed at the current moment, the course at the current moment, the flight altitude, the horizontal and vertical coordinates of the aircraft and the short-term historical change sequence of the course.
In the above technical solution, in step three, the heading state predicted by the naive Bayes classifier at the next time includes: left steering, straight traveling and right steering, and calculating the probability of each state.
In the technical scheme, when the air route is spliced by a plurality of sections, each section of air routeThe line influence prediction probability is solved by the formulaiCalculating the influence prediction probability on the whole route by the following formula:
P＝1-Π(1-Pi)。
the invention has the advantages and beneficial effects that:
the invention establishes an airplane course intention conjecture model based on an AR model and a Bayes classifier, firstly, preprocessing the current sample data, and homogenizing the sample sampling time by using a linear interpolation mode; on the basis, a plurality of differential variables are selected as input variables of a prediction model by analyzing correlation coefficients among input and output variables, such as course historical differential data and speed differential data; fusing a plurality of course difference historical data by adopting an AR model, and giving an estimated value of course change at the next moment; in order to fully utilize the detected data, for example, the differential value of the speed may reflect the deceleration action before the airplane turns, on the basis of the AR model, the estimated output of the AR model, the differential data of the current heading and the differential data of the speed are used as input, and a naive Bayes classifier is adopted to predict the possible situation (left turning, straight going or right turning) of the heading at the next moment. The final result shows that the course intention prediction model established by the invention has good performance for predicting the course intention.
On the basis that a Bayes model predicts the course state of an unknown aircraft at the next moment, a prediction probability model of the influence of the unknown flight path on the air route is established, the model can effectively predict the prediction probability of the influence of the unknown aircraft on the flight route of the civil aircraft in a short time, and can be used as early warning judgment.
Drawings
FIG. 1 is a graph of correlation coefficients between a feature set of input variables and output variables in the present invention.
Fig. 2 is a diagram of the prediction result of the AR model in the present invention.
FIG. 3 is a graph of the comparison between the model prediction results and the actual labels of the present invention.
For a person skilled in the art, other relevant figures can be obtained from the above figures without inventive effort.
Detailed Description
In order to make the technical solution of the present invention better understood, the technical solution of the present invention is further described below with reference to specific examples.
A method for predicting influence probability of an unknown-intention flight trajectory on an airway includes the following steps:
the method comprises the following steps: firstly, the flight data samples are preprocessed, and the sampling time intervals of the data samples are homogenized. Specifically, the method comprises the following steps:
the flight data of the aircraft are recorded by time sequence data with non-uniform time intervals, and in the existing research, the sampling frequency is generally required to be kept constant in the continuous state discretization process, so the sampling time intervals of the data samples are firstly homogenized.
Considering that all the collected data have the characteristic of good continuity and the existing sampling interval is small enough, the samples are repaired by adopting a linear interpolation mode, and the time interval after the repair is 5s in the embodiment, wherein x istSamples at time t after repair, tk,tk+1Two sampling time points adjacent to t.
Step two: selecting the candidate input variables with high correlation contribution as input characteristic variables of the prediction model by analyzing correlation coefficients between the possible candidate input variables and output variables (namely change values of the aircraft course at the next moment) of the sample, wherein the selected input characteristic variables of the prediction model comprise: historical differential sequence data for the heading, differential data for the current speed, and differential data for the current heading. Specifically, the method comprises the following steps:
possible candidate input variables for the sample include: the speed at the current moment, the course at the current moment, the flight altitude, the horizontal and vertical coordinates of the aircraft and the short-term historical change sequence of the course.
According to the flying characteristics of the aircraft, carrying out differential processing on the input variables to be selected to obtain backward differential data of each variable at the time k and in a period before, and using the backward differential data as an input variable set of the model to be selected, namely:
Figure BDA0002728807460000051
wherein
Figure BDA0002728807460000052
And (3) representing the value of the ith variable at the time k as a candidate variable of the model input characteristic. When k is 10, the correlation coefficient between the input variable feature set and the output variable is shown in fig. 1.
As can be seen from fig. 1, the most contributing factor to the output is the historical heading difference data, the autocorrelation coefficient of which is much larger than that of other variables, and it can be seen from the figure that the input variable and the output variable have a significant positive correlation in the time period k-3 and before, and have a negative correlation in the time period k-4 and after. I.e. it is possible that the aircraft can perform course adjustment around 25 s. In addition, the correlation between the speed difference and the output variable is larger than that of other variables; in addition, considering that the autocorrelation coefficient of the course difference at the current moment and the course difference at the next moment is obviously greater than other values, the historical difference sequence, the current speed difference and the current course difference of the course are used as input characteristic variables of the prediction model in the invention.
Step three, fusing a plurality of course difference historical data by adopting an AR model, and giving an estimated value of course change at the next moment; and taking the estimated value output by the AR model, the differential data of the current course and the differential data of the current speed as input, and predicting the possible condition (left steering, straight running or right steering) of the course at the next moment by adopting a naive Bayes classifier. Specifically, the method comprises the following steps:
since the autocorrelation of the heading difference data is strong, the autocorrelation is contrary to the independence between the assumed input features in the naive Bayes model, and for the Bayes model, the distribution state of the data points in the space needs to be estimated, the whole data set is very unbalanced, the straight samples are far away from the redundant steering samples, and in conclusion, the number reduction is realizedThe data distribution estimation in the Bayesian model is very beneficial according to the dimension; therefore, the heading difference historical data features need to be fused into one to adapt to the subsequent Bayesian model for prediction. Therefore, the embodiment adopts an AR model (Auto-Regression), takes the course difference historical time sequence as the input of the AR model, and further outputs the estimated value of course change at the next moment
Figure BDA0002728807460000053
The estimated value is used as a new predicted heading difference parameter, and the model is shown as (2):
Figure BDA0002728807460000061
according to the ACF scaling method for the AR model, a position where the autocorrelation coefficient of the model approaches 0 may be taken as the order of the model, and in the embodiment, the change of the autocorrelation coefficient is analyzed, and the order of the AR model is set to 10. The optimization solution yields the results shown in table 1:
TABLE 1 AR model coefficients
Figure BDA0002728807460000062
The regression results of the AR model are shown in fig. 2, and it can be seen from fig. 2 that the established AR model has a good prediction effect on the output data, and its RMSE index value is 5.9560, where 95% of the points fall between two red dotted lines.
Next, the estimated value output by the AR model
Figure BDA0002728807460000063
Differential data of current course
Figure BDA0002728807460000064
And difference data Δ v of the current speedkAs an input parameter of naive Bayes, a naive Bayes classifier is adopted to carry out the possible situations (left steering, straight going or right steering) of the course at the next momentTo) to predict.
Bayes classifiers are a generic term for a class of classification algorithms that are all based on Bayes' theorem, as given in equation (3). The classification principle is that the prior probability of a certain object is calculated by using a Bayesian formula, namely the probability that the object belongs to a certain class, and the class with the maximum posterior probability is selected as the class to which the object belongs. The nature of the classification problem is to calculate the conditional probability that a sample belongs to each class on the basis of prior knowledge, namely model input, and under a general condition, a machine learning algorithm based on data is difficult to obtain the probability density distribution of the data in a high-dimensional space, so that a statistical learning method is difficult to directly apply to obtain the corresponding conditional probability density distribution. The na iotave Bayes theorem, in turn, translates this conditional probability computation problem well into the computation of conditional probabilities between variables in a low-dimensional space.
Figure BDA0002728807460000065
In the above formula, the first and second carbon atoms are,
Figure BDA0002728807460000066
c represents a sample of classification (course classification: left steering, right steering and straight going), P (C) is a prior probability function, namely a sample data statistical value classified as C in historical data, and P (omega | C) is a conditional probability function of the sample omega belonging to the class label C, or is called a likelihood function; p (Ω) is an evidence factor for probability normalization. For a given sample data set, the evidence factor is independent of class labels, so the estimate P (C | Ω) can be converted to estimate the prior probability function P (C) and the likelihood function P (Ω | C) based on the training data R.
For Naive Bayes Classifier (Naive Bayes Classifier), Naive means that the input feature attributes are assumed to be independent from each other in the modeling process. Namely, satisfies formula (4):
Figure BDA0002728807460000071
for the continuous problem, a probability density distribution function is generally used to describe the distribution of conditional probability. A commonly used distribution model is a gaussian distribution model, that is, assuming that the distribution of the samples follows gaussian distribution, the likelihood function is estimated as shown in equation (5):
Figure BDA0002728807460000072
wherein:
Figure BDA0002728807460000073
Figure BDA0002728807460000074
wherein DcRepresenting the number of collections classified as C. Although the above conditions are very strict and the data set in general cannot meet the above requirements, in practical applications, a naive Bayes classifier can often obtain a good classification result. Therefore, the invention adopts the classifier to predict and classify the change of the aircraft course.
Simulation of experiment
In an embodiment, since the classifier provides classified data, the output prediction data needs to be discretized. It is known from practical physics that an output of 0 indicates that the aircraft is traveling straight, a heading of greater than 0 indicates a turn to the right, and a value of less than 0 indicates a turn to the left. They are therefore first classified herein into three categories according to a given threshold (threshold): greater than threshold of 1, less than-threshold of-1, and others of 0, then the threshold parameter indicates the degree of turn, the greater the parameter, the greater the turn.
The model parameters for the model estimation are shown in Table 2, where the ith row and the jth column indicate the mean parameter and variance parameter of the Gaussian distribution of the jth variable under the ith class of conditions.
TABLE 2 na iotave Bayes model parameters
Figure BDA0002728807460000081
Wherein-1 represents a left steering state of the flight trajectory without intention, 0 represents a right steering state, and 1 represents a straight-ahead state. Fig. 3 shows the comparison between the prediction result of the partial model and the actual label, wherein the dotted line represents the actual label, and when 1 is reached, it represents that the current event actually occurs, and the solid line represents the probability that the event may occur given the model. As can be seen from the figure, the prediction result of the model has good prediction effect and can be well used for the prediction of the aircraft navigation intention.
And fourthly, further establishing a prediction model of the influence probability of the unknown aircraft on the air route (civil aviation route) on the basis that the Bayes classifier is used for predicting the course change state of the unknown aircraft at the next moment of the flight track in the third step. Specifically speaking:
the state of course change of the aircraft at the next moment without intention is
Figure BDA0002728807460000082
(
Figure BDA0002728807460000083
Including left steering, straight traveling, right steering, etc.) with a heading change angle of Δ SkThe state of probability distribution of, i.e.
Figure BDA0002728807460000084
Assuming that the flight path is of a certain kind without intention
Figure BDA0002728807460000085
The heading transition condition under the flight intention follows normal distribution, namely:
Figure BDA0002728807460000086
wherein the mean and standard deviation parameters in the probability density distribution function are derived from sample statistics in the training data set, i.e.:
Figure BDA0002728807460000087
Figure BDA0002728807460000088
based on the above assumption of the model of equation 8, it is further assumed that the aircraft is flying in the predicted state for the next period of time. If the unknown flight trajectory passes through a known civil route at a subsequent time, it is assumed that the aircraft may have an impact on the civil route to be evaluated for a subsequent period of time.
According to the hypothesis, when the changed course is in an opening angle formed by the current point and a connecting line of the to-be-evaluated air route, the flight of the aircraft can influence the civil air route. Setting the course angle formed by the connection line of the current track position and the end point of the air route as alpha respectively1And alpha2The current course angle is alpha0Then the probability of the unknown intended flight trajectory affecting the flight path is:
Figure BDA0002728807460000091
the predicted probability of the effect of the unintended aircraft on the designated route is shown as equation (11), where
Figure BDA0002728807460000092
Indicating when the Bayes classifier predicts a state of
Figure BDA0002728807460000093
Probability of time (available from example step three),
Figure BDA0002728807460000094
belonging to the (left turn, straight, right turn) sequence.
Figure BDA0002728807460000095
When the route is spliced by a plurality of sections, the influence prediction probability of each section of route is solved by the formula (11) to obtain PiThe predicted probability of influence on the whole route can be calculated by equation (12):
P＝1-Π(1-Pi) (12)
the invention has been described in an illustrative manner, and it is to be understood that any simple variations, modifications or other equivalent changes which can be made by one skilled in the art without departing from the spirit of the invention fall within the scope of the invention.

Claims (5)

1. A method for predicting influence probability of an unknown-intention flight path on an air route is characterized by comprising the following steps:
the method comprises the following steps: firstly, preprocessing an acquired flight data sample, and homogenizing the sampling time interval of the data sample;
step two: selecting the input variable to be selected with high correlation contribution as the input characteristic variable of the prediction model by analyzing the correlation coefficient between the possible input variable to be selected and the output variable of the sample, wherein the selected input characteristic variable of the prediction model comprises the following steps: historical differential sequence data of the heading, differential data of the current speed and differential data of the current heading;
step three: fusing a plurality of course difference historical data by adopting an AR model, and giving an estimated value of course change at the next moment; taking the estimated value output by the AR model, the differential data of the current course and the differential data of the current speed as input, and adopting a naive Bayes classifier to process the course state at the next moment
Figure FDA0002728807450000011
Carrying out probability prediction;
step four: predicting the course state of the flight track with unknown intention at the next moment according to the step three
Figure FDA0002728807450000012
Calculating the predicted probability of the influence of the unknown intended flight trajectory on the designated route:
course state of aircraft at next moment
Figure FDA0002728807450000013
Under the condition of (1), its course change angle is Delta SkThe state of probability distribution of, i.e.
Figure FDA0002728807450000014
Assuming that the flight path is of a kind not known to be intended
Figure FDA0002728807450000015
The heading transition condition under the flight intention follows normal distribution, namely:
Figure FDA0002728807450000016
wherein the mean and standard deviation parameters in the probability density distribution function are derived from sample statistics in the training data set, i.e.:
Figure FDA0002728807450000017
Figure FDA0002728807450000018
if the aircraft flies in a predicted state in the next period of time, and if the flight track with unknown intention passes through a known civil aviation route at the next certain time, the aircraft is considered to possibly influence the civil aviation route to be evaluated in the next period of time;
according to the hypothesis, when the changed course is in an opening angle formed by the current point and a connecting line of the air route to be evaluated, the flight of the aircraft can possibly influence the civil aviation route;
setting the course angle formed by the connection line of the current track position and two end points of the route as alpha respectively1And alpha2The current course angle is alpha0Then, the probability that the unknown intended flight trajectory affects the civil aviation route is:
Figure FDA0002728807450000021
the predicted probability of the effect of the unintended aircraft on the designated route is shown as,
Figure FDA0002728807450000022
wherein the content of the first and second substances,
Figure FDA0002728807450000023
indicating when the Bayes classifier predicts a state of
Figure FDA0002728807450000024
The probability of the time of day is,
Figure FDA0002728807450000025
belonging to the sequence of left steering L, straight driving R and right steering St.
2. The method of predicting probability of influence of an unintended flight trajectory on an airway as defined in claim 1, wherein: in the first step, the sampling time interval of the data samples is homogenized by using a linear interpolation mode.
3. The method of predicting probability of influence of an unintended flight trajectory on an airway as defined in claim 1, wherein: possible candidate input variables for the sample include: the speed at the current moment, the course at the current moment, the flight altitude, the horizontal and vertical coordinates of the aircraft and the short-term historical change sequence of the course.
4. The method of predicting probability of influence of an unintended flight trajectory on an airway as defined in claim 1, wherein: in step three, the heading state of the naive Bayes classifier at the next moment comprises the following steps: left steering, straight traveling and right steering, and calculating the probability of each state.
5. The method of predicting probability of influence of an unintended flight trajectory on an airway as defined in claim 1, wherein: in the fourth step, when the flight path is spliced by a plurality of sections, the influence prediction probability of each section of flight path is solved by the formulaiCalculating the influence prediction probability on the whole route by the following formula:
P＝1-Π(1-Pi)。
CN202011115262.4A 2020-10-16 2020-10-16 Method for predicting influence probability of unknown intention flight path on air route Pending CN112232567A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011115262.4A CN112232567A (en) 2020-10-16 2020-10-16 Method for predicting influence probability of unknown intention flight path on air route

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011115262.4A CN112232567A (en) 2020-10-16 2020-10-16 Method for predicting influence probability of unknown intention flight path on air route

Publications (1)

Publication Number Publication Date
CN112232567A true CN112232567A (en) 2021-01-15

Family

ID=74117851

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011115262.4A Pending CN112232567A (en) 2020-10-16 2020-10-16 Method for predicting influence probability of unknown intention flight path on air route

Country Status (1)

Country Link
CN (1) CN112232567A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114926795A (en) * 2022-07-19 2022-08-19 深圳前海中电慧安科技有限公司 Method, device, equipment and medium for determining information relevance

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102621987A (en) * 2011-01-28 2012-08-01 波音公司 Providing data for predicting aircraft trajectory
CN103336863A (en) * 2013-06-24 2013-10-02 北京航空航天大学 Radar flight path observation data-based flight intention recognition method
CN104091470A (en) * 2014-07-15 2014-10-08 南京大学 Channel traffic information prediction method and application based on multidata fusion

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102621987A (en) * 2011-01-28 2012-08-01 波音公司 Providing data for predicting aircraft trajectory
CN103336863A (en) * 2013-06-24 2013-10-02 北京航空航天大学 Radar flight path observation data-based flight intention recognition method
CN104091470A (en) * 2014-07-15 2014-10-08 南京大学 Channel traffic information prediction method and application based on multidata fusion

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
张余: "洋区不明意图飞行轨迹对空中交通的影响研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑(月刊)》, no. 2, pages 031 - 1432 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114926795A (en) * 2022-07-19 2022-08-19 深圳前海中电慧安科技有限公司 Method, device, equipment and medium for determining information relevance
CN114926795B (en) * 2022-07-19 2022-11-15 深圳前海中电慧安科技有限公司 Method, device, equipment and medium for determining information relevance

Similar Documents

Publication Publication Date Title
CN108594788B (en) Airplane actuator fault detection and diagnosis method based on deep random forest algorithm
Zeng et al. A deep learning approach for aircraft trajectory prediction in terminal airspace
CN107016193B (en) Expected following distance calculation method in driver following behavior analysis
Gadepally et al. A framework for estimating driver decisions near intersections
CN112242059B (en) Intelligent decision-making method for unmanned vehicle based on motivation and risk assessment
Janakiraman Explaining aviation safety incidents using deep temporal multiple instance learning
CN109109863B (en) Intelligent device and control method and device thereof
Shangguan et al. A proactive lane-changing risk prediction framework considering driving intention recognition and different lane-changing patterns
Li et al. Development and evaluation of two learning-based personalized driver models for pure pursuit path-tracking behaviors
CN114048889A (en) Aircraft trajectory prediction method based on long-term and short-term memory network
Jeong et al. Bidirectional long shot-term memory-based interactive motion prediction of cut-in vehicles in urban environments
CN109977546B (en) Four-dimensional track online anomaly detection method based on unsupervised learning
Zhu et al. Interaction-aware cut-in trajectory prediction and risk assessment in mixed traffic
CN115195748A (en) Data-driven intelligent automobile personification decision planning system and method
Gadepally et al. Driver/vehicle state estimation and detection
Li et al. Modeling vehicle merging position selection behaviors based on a finite mixture of linear regression models
Remmen et al. Cut-in scenario prediction for automated vehicles
Kong et al. Bayesian deep learning for aircraft hard landing safety assessment
WO2023187117A1 (en) Simulation-based testing for robotic systems
Zhang et al. Spatiotemporal interaction pattern recognition and risk evolution analysis during lane changes
Cui et al. Intelligent health management of fixed-wing UAVs: A deep-learning-based approach
Wheeler et al. A probabilistic framework for microscopic traffic propagation
Deshmukh et al. Data-driven precursor detection algorithm for terminal airspace operations
CN112232567A (en) Method for predicting influence probability of unknown intention flight path on air route
Wu et al. Aircraft flight regime recognition with deep temporal segmentation neural 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
CB03 Change of inventor or designer information

Inventor after: Fu Yongqiang

Inventor after: Wang Ye

Inventor after: Mao Jizhi

Inventor after: He Weiwei

Inventor after: Tong Jialu

Inventor after: Xie Hua

Inventor before: Fu Yongqiang

Inventor before: Wang Ye

Inventor before: Mao Jizhi

Inventor before: Zhu Yongwen

Inventor before: Wang Changchun

Inventor before: He Weiwei

Inventor before: Tang Zhili

Inventor before: Tong Jialu

Inventor before: Xie Hua

CB03 Change of inventor or designer information