CN106249232B - Method for tracking target based on target state of motion information data associating policy - Google Patents
Method for tracking target based on target state of motion information data associating policy Download PDFInfo
- Publication number
- CN106249232B CN106249232B CN201610721016.0A CN201610721016A CN106249232B CN 106249232 B CN106249232 B CN 106249232B CN 201610721016 A CN201610721016 A CN 201610721016A CN 106249232 B CN106249232 B CN 106249232B
- Authority
- CN
- China
- Prior art keywords
- target
- mark
- tracking
- point mark
- moment
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/66—Radar-tracking systems; Analogous systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses the method for tracking target based on target state of motion information data associating policy, including：Step 1, targetpath to be selected is created；Step 2, targetpath to be selected is handled into line trace, and rational official goals flight path is established according to target initial conditions；Step 3, estimation is filtered to tracking target and official goals flight path to handle, obtain flight path state estimation using Kalman filter model.The least meeting distance point time TCPA of tracking target and official goals is calculated by ship collision prevention radar measurement model, TCPA values are compared with target disengaging time thresholding, target time to approach thresholding and target overlapping time thresholding, obtain residing movement state information between target；Step 4：The associating policy and association algorithm that tracking Targets Dots flight path is determined according to movement state information residing between target finally obtain tracking target and put the incidence relation of mark.
Description
Technical field
The invention belongs to radar target tracking technologies, more particularly to based on target state of motion information data associating policy
Method for tracking target.
Background technology
Data correlation process is to determine the process of the measurement information that sensor receives and target source correspondence, through
Target starting, target maintains and target terminating procedures.Currently, common data association algorithm includes two major classes in engineering, first
Class is studied newest measurement set, as Singer and Sea proposes optimal nearest neighbor algorithm, basic principle selection and mesh
The nearest point mark of predicted position is marked as relating dot.When target signal is higher or target sparse, the algorithm keeps track effect
Preferably.And low-resolution radar investigative range is wide, tracking destination number is more.In addition radar resolution is low, azimuthal measurement low precision, mesh
Intensive trend is easily presented for mark or target echo is overlapped, target overtakes, target crossover phenomenon.When target is in heavy dense targets area
Or under target echo Overlay scenes, the optimal nearest neighbor algorithm target following effect based on single location information is poor.Second class
Data association algorithm is that all measurement set pervious to current time are studied, the more hypothesis algorithms proposed such as Reid
(MHT).It is assume that algorithm using the metric data of follow-up multiple scan periods, realizes the stabilization of target in heavy dense targets region more
Tracking.Donald B.Reid assume that algorithm simulation simulates tracking effect (An when two target echoes are overlapped more using
Algorithm for Tracking Multiple Target.IEEE Transctions on Automatic Control,
Vol.AC-24, NO.6, December 1979.).Theoretically, assume algorithm using the posterior probability for assuming flight path branch more
To solve the ambiguity of compact district or clutter area target data association.Masamichi Kojima prove that more hypothesis algorithms are one
Kind solves optimal algorithm (the A Study of Target Tracking of the Multiple Targets Data Association accuracy under complex environment
Using Track-oriented Multiple Hypothesis Tracking.SICE, 1998.5:29-31.).But MHT is calculated
Assume that flight path number of branches exponentially rises pass with false alarm rate, number of targets and handled multiple scan period numbers caused by method
System, therefore the calculation amount of algorithm and amount of storage are all very big, to be applied in real time in extra large radar target tracking engineering still
It acquires a certain degree of difficulty.
Invention content
It is an object of the invention to identify target local environment using target state of motion information, believed according to target environment
Breath adaptively uses different data correlation strategy and data association algorithm, improves Targets Dots track association accuracy, especially
It is to improve multiple target the ability of tracking in complex scenes such as to meet, intersect.
Realize that the technical solution of the object of the invention is：Target based on target state of motion information data associating policy
Tracking, steps are as follows：
Include the following steps：
Step 1, targetpath to be selected is created：Centered on tracking target, r be the Searching point in the circle regional extent of radius
Mark information, search radius r are related with the radar antenna period.The radar antenna period is longer, and search radius is bigger.Under normal circumstances may be used
According to tracking position of object, 1~7Km is set.If point mark belongs to the region, targetpath to be selected is created using this mark；
Otherwise targetpath to be selected is not created；
Step 2, targetpath to be selected is handled into line trace：Within the detections of radar period, if targetpath to be selected meets
Targetpath originates constraints, then targetpath to be selected is switched to official goals flight path, otherwise delete targetpath to be selected；
Step 3, estimation is filtered to tracking target and official goals flight path to handle, obtain using Kalman filter model
Flight path state estimation.When calculating the least meeting distance point for tracking target and official goals by ship collision prevention radar measurement model
Between TCPA (time to closest point of approach), TCPA values and target disengaging time thresholding, target are approached
Time threshold and target overlapping time thresholding are compared, and obtain tracking target and the mutual residing movement shape of official goals
State information, such as separation, close or overlap condition；
Step 4：Tracking target point is determined according to tracking target and the mutual residing movement state information of official goals
The associating policy and association algorithm of mark flight path finally obtain tracking target and put the incidence relation of mark, to realize target following.
Steps are as follows for targetpath to be selected in step 1 of the present invention within the scope of foundation tracking target area, it is assumed that tracking mesh
Mark distance, the orientation references s of AA, wA.With (sA,wA) centered on, r is that radius determines regional extent.For falling into the region model
Point mark in enclosing establishes targetpath to be selected.
Step 2 of the present invention includes the following steps：
Step 2-1, it is assumed that in t1Moment creates target starting point to be selected, from t2Moment, using optimal nearest neighbor algorithm
A mark track association is carried out to target to be selected：
Wherein, z indicates radar detection value, S-1New breath covariance, d are filtered for target to be selected2(z) it is new breath weighted norm,Indicate estimation of the k moment to k+1 moment target prediction states；
If d2(z)≤γ is set up, then judges target association to be selected to available point mark, otherwise, it is determined that Targets Dots to be selected are lost
It loses, γ is distance threshold value.If target association to be selected records target association point mark information to be selected, mesh to be selected to available point mark
It includes point mark radial distance g to mark relating dot mark information1, point mark orientation g2, point mark energy g3, point mark orientation extend g4, point mark distance
Extend g5With a mark time g6, the target association point mark information to be selected is formed into relating dot mark feature vector G, is shown below：
G=[g1 g2 g3 g4 g5 g6]；
Step 2-2 estimates the filter state and predicted state of target to be selected according to Kalman filter formulation group：
K (k)=P (k | k-1) H'(k) [H (k) P (k | k-1) H'(k)+Rk]-1
P (k | k-1)=Φ (k | k-1) P (k-1 | k-1) Φ ' (k | k-1)+Q (k)
P (k | k)=[I-K (k) H (k)] P (k | k-1)
Wherein, Z (k) expressions k moment sensor detection informations, and Φ (k | k-1) indicate that the state at k-1 moment to k moment turns
Matrix is moved, H (k) indicates that measurement matrix, U (k) indicate that input control item matrix, u (k) indicate known input or control signal, R
(k) indicating that zero-mean, White Gaussian measure the covariance of noise, Q (k) is the covariance of zero-mean, White Gaussian process noise,Indicate the state vector at target k moment,Indicate prediction of the target at the k-1 moment to the k moment, K (k) tables
Showing k moment filtering gains, P (k | k) indicates the covariance matrix at k moment, P (k | k-1) indicate that prediction covariance matrix, I are single
Bit matrix, H'(k) be measurement matrix transposed matrix, Φ ' (k | k-1) is the transposed matrix of state-transition matrix.
Step 2-3 repeats step 2-1~step 2-2, until tnMoment, N are the detections of radar period, it is assumed that target to be selected
M relating dot mark is shared within N number of detections of radar period, relating dot mark sequence is expressed as [G1,G2...,GM]；
Step 2-4 calculates target proximity association point trace to be selected to range rate according to following formula：
Wherein, gi1Indicate first feature of i-th of radar scanning period target association point mark to be selected, that is, put trace to away from
From gj1Indicate first feature of j-th of radar scanning period target association point mark to be selected, gi6Indicate i-th of radar scanning week
The 6th feature of target association point mark is selected in expectation, that is, puts mark time, gj6Indicate j-th of radar scanning period target association to be selected
6th feature of point mark, i=j+1, j=1,2 ..., M-1.Such as g11Indicate the of first scan period target association point mark
1 feature, g21Indicate the 1st feature of second scan period target association point mark.tk+1-tkIndicate vk+1With vkBetween when
Between poor, vk, vk+1Indicate k, k+1 moment target radial is apart from speed, k=1,2 ..., n-2, akIndicate radial distance change rate；
Step 2-5 calculates the rate of azimuth change of target proximity association point mark to be selected according to following formula：
Azik=gi2-gj2,
Azik'=Azik+1-Azik,
Wherein, gi2And gj2Orientation values (i.e. second spy of i-th of radar scanning target cycle relating dot mark is indicated respectively
Sign) and j-th of radar scanning target cycle relating dot mark orientation values, AzikIndicate j-th of radar scanning target cycle point mark
Bearing variation value, Azik' indicate rate of azimuth change,
Step 2-6 calculates separately to obtain n target range rate of change value and Orientation differences according to step 2-4 and step 2-5
The root mean square of radial distance change rate and the root mean square RMSE of rate of azimuth change is calculated as follows in rate value：
Wherein, XiIndicate target range change rate or rate of azimuth change,Be target radial range rate mean value or
The mean value of person's rate of azimuth change；
Step 2-7 calculates the detection probability p of target association point mark to be selected：
Step 2-8, if target radial range rate root mean square thresholding is RdG, target bearing change rate root mean square thresholding is
RaG, target detection probability thresholding is Pd.If the root mean square RMSE of the radial distance change rate of target proximity association point mark to be selectedd、
The root mean square RMSE of rate of azimuth changeaIt is satisfied by the following conditions with detection probability p：
RMSEd≤RdG,
RMSEd≤RaG,
p≤Pd,
Targetpath to be selected is then switched into official goals flight path, and target following is maintained according to Kalman filter model, together
Relating dot mark feature vector during the tracking of Shi Jilu targetpaths.
Step 3 of the present invention includes the following steps：
Step 3-1, according to the method described in step 2-1~step 2-7, using Kalman filter model to tracking target and
Official goals flight path is filtered estimation processing, obtains flight path state distance, orientation, course and speed of a ship or plane estimated value.Utilize ship
Radar for collision avoidance measurement model, tracking target and the distance of official goals, orientation, course, the speed of a ship or plane, calculate tracking target and formal mesh
Target least meeting distance point time TCPA：Plane coordinate system XOY is established, Y-axis is direct north, if tracking target S1Positioned at seat
Mark origin O and with course d1Speed of a ship or plane V1Uniform motion tracks the official goals S detected around target2With course d2Speed of a ship or plane V2It is even
Speed movement, T1Moment, official goals S2Opposite tracking target S1Position be A points, relative distance R, relative bearing θ, AA'
For two target virtual courses, the angle of AA' and Y-axis is φr, the opposite speed of a ship or plane is Vr, φ is calculated using following formularAnd Vr：
After t moment, official goals S is calculated2Opposite tracking target S1Displacement：
Official goals S is calculated by Δ X and Δ Y2Opposite tracking target S1Distance R (t) after t moment and orientation θ (t)：
Wherein, Xt0And Yt0T moment official goals S is indicated respectively2Initial position abscissa and t moment official goals S2's
Initial position ordinate, Δ X and Δ Y indicate target S respectively2The position of X-direction moves in t moment position and Y-direction movement
It sets, in practical engineering application, by t moment radar to target S2Detection, can get location information R (t) and θ (t).Pass through
Following formula calculates the least meeting distance point time TCPA of tracking target and official goals：
TCPA=R (t) | cos (φr-θ(t))|/Vr,
Step 3-2, if GT1For target echo overlapping time, GT2For target echo time to approach.If TCPA absolute values are less than
GT1, then judge to track target and official goals for state can be met；If TCPA absolute values are more than GT1Less than GT2, then judge to track mesh
Mark is proximity state with official goals；If TCPA absolute values are more than GT2, then judge to track target with official goals to detach shape
State.
Step 4 of the present invention includes the following steps：
Step 4-1 carries out tracking target using optimal nearest neighbor algorithm if tracking target and target are in discrete state
Data correlation；
Step 4-2 judges to track target and formal mesh if tracking target and official goals are in close or overlap condition
Whether mark is associated with common point mark.If different point marks is respectively associated in tracking target and official goals, at normal tracking
Reason；If tracking target and official goals relating dot mark be common point mark, judge following conditions whether and meanwhile set up：
Wherein,WithThe energy of the average energy value and official goals relating dot mark of tracking target association point mark is indicated respectively
Mean value is measured,WithThe echo bearing extension mean value of tracking target association point mark and returning for official goals relating dot mark are indicated respectively
Wave orientation extends mean value, EdFor the observable common point mark energy value of radar, ldFor radar observable common point mark orientation
Expanding value.α, β are proportionality coefficient, and value range is respectively 0.7≤α≤1 and 0.7≤β≤1.GEFor energy error thresholding, GlFor
Orientation extends error threshold, if condition is set up simultaneously, it is tracking target and official goals echo overlapping point mark to put mark, puts mark
It is not assigned to tracking target and official goals, target carries out extrapolation process respectively；If condition is invalid, put mark derive from
Track target or official goals, are calculated as follows ΔAAnd ΔB：
ΔAIndicate the similitude of the energy and tracking target energy mean value of common point mark, ΔBIndicate the energy of common point mark
With the similitude of official goals average energy value, if ΔA≤ΔB, point mark is tracking target association point mark, if ΔA≥ΔB, putting mark is
Official goals relating dot mark.
In step 4-1 of the present invention, when tracking target and official goals and being in discrete state, using optimal nearest neighbor algorithm into
Row data correlation.If tracking target association point mark energy feature collection is combined into { E1,E2,...,En, echo bearing extensive features sets
For { l1,l2,...,ln, n is the count value for tracking target association point mark.Official goals relating dot mark energy feature collection is combined into
{E'1,E2',...,Em', echo bearing expanded features are combined into { l1',l2',...,lm'}.M is official goals relating dot mark
Count value.Point mark energy feature and orientation extension feature can be used ripe point mark extractive technique and realize, that is, utilize target echo
Plot coherence is carried out in orientation, apart from upper continuity, obtains target bearing extension feature.Energy feature will be by that will participate in accumulation
The corresponding energy of azran carry out summation acquisition.
The average energy value of tracking target association point mark is calculated by following formulaWith the energy of official goals relating dot mark
Mean value
The echo bearing that tracking target association point mark is calculated by following formula extends mean valueWith official goals relating dot
The echo bearing of mark extends mean value
Advantageous effect：Compared with prior art, the present invention its remarkable advantage：(1) it during target data association, not only examines
Consider the state estimation of tracking target itself, it is also contemplated that the environmental information around tracking target establishes tracking target and week
The state of motion relationship between other targets is enclosed, the adaptive switching of a variety of data association algorithms is realized.(2) using between target
Situation information, which judges that target is in, complicated when tracking scene, the binding site mark diverse characteristics information assistance data such as can meet, intersect
Association algorithm reduces associated errors rate caused by being associated with based on single location information.(3) it compared with assuming algorithm, is based on more
Calculation amount, the amount of storage of the data correlation strategy process of target state of motion information are smaller, are suitble to engineer application.
Description of the drawings
The present invention is done with reference to the accompanying drawings and detailed description and is further illustrated, of the invention is above-mentioned
And/or otherwise advantage will become apparent.
Fig. 1 is to create object delineation to be selected.
Fig. 2 is that target to be selected turns official goals process chart.
Fig. 3 is ship collision prevention radar measurement model.
Fig. 4 is that target state of motion judges process chart.
Fig. 5 is that TCPA state of motion judges definition graph.
Fig. 6 is that multiple target overtakes tracking schematic diagram.
Fig. 7 is to track target situation change curve during multiple target overtakes tracking.
Fig. 8 is that multiple target intersects tracking schematic diagram.
Fig. 9 is to track target situation change curve during multiple target intersects tracking.
Specific implementation mode
The first step is illustrated with reference to Fig. 1 targetpath to be selected and creates process.Centered on tracking target A, r is the circle of radius
Searching point mark information in regional extent.Wherein, search radius r is related with the radar antenna period.The radar antenna period is longer, searches
Rope radius is bigger.Can 1~7Km be set according to tracking position of object under normal circumstances.If point mark falls into the region, create
Targetpath to be selected；Otherwise, targetpath to be selected is not created.(in radar detection object procedure, be vulnerable to clutter, target etc. because
Element influences, and synchronization multiple marks can occur in tracking target Bo Mennei.The meaning at two t1 moment is when identical in Fig. 1
In quarter, there are two marks, two mark time consistencies simultaneously in target Bo Mennei.Similarly, also there are two points in the t3 moment
Mark.)
Second step handles the targetpath to be selected within the scope of tracking target a-quadrant into line trace.In several detection cycles
It is interior, if goal satisfaction targetpath initial conditions to be selected, target to be selected is switched into official goals flight path B, is otherwise deleted to be selected
Targetpath, it is described in detail below.
(2-1)：t1Moment has created target starting point to be selected.From t2Moment utilizes optimal nearest neighbor algorithm formula pair
Target to be selected carries out a mark track association.Wherein, z indicates radar detection value, S-1New breath covariance, d are filtered for target to be selected2
(z) it is new breath weighted norm.
If d2(z)≤γ is set up, then target association to be selected is to available point mark, and otherwise, Targets Dots to be selected are lost.γ be away from
From threshold value.If target association to be selected records target association point mark information to be selected to point mark.Target association point mark information to be selected
Mainly it is made of six parts：Point mark radial distance g1, point mark orientation g2, point mark energy g3, point mark orientation extend g4, point mark distance
Extend g5, put mark time g6, relating dot mark feature vector G is formed, as shown in formula.
G=[g1 g2 g3 g4 g5 g6]
(2-2)：After target data association to be selected is disposed, according to Kalman filter formulation group, to target to be selected
Filter state, predicted state are estimated.
K (k)=P (k | k-1) H'(k) [H (k) P (k | k-1) H'(k)+Rk]-1
P (k | k-1)=Φ (k | k-1) P (k-1 | k-1) Φ ' (k | k-1)+Q (k)
P (k | k)=[I-K (k) H (k)] P (k | k-1)
(2-3)：(2-1)~(2-2) step is repeated until tNMoment, N are detection cycle.In view of radar false alarm probability and
Ensure that target effective relating dot mark sample number to be selected, N generally take 5~7.
(2-4)：It is illustrated in combination with fig. 2 the process that object judgement to be selected is official goals.Assuming that target to be selected is visited in N number of radar
It surveys relating dot mark sequence in the period and is represented by [G1,G2...,GM], M is target association point mark number to be selected, GiFor a mark feature
Vector.
(2-4-1)：Target consecutive points trace to be selected is calculated by formula to range rate.
G in formulaijIndicate j-th of feature of i-th of radar scanning target cycle relating dot mark, i=j+1, j=1,2 ...,
m-1。vk, vk+1Indicate k, k+1 moment target radial is apart from speed, k=1,2 ..., n-2, tk+1-tkFor vk+1With vkBetween when
Between poor, akFor target radial range rate value.
(2-4-2)：The rate of azimuth change of target consecutive points mark to be selected is calculated by formula.
Azik=gi2-gj2, i=j+1
Azik'=Azik+1-Azik
gijIndicate j-th of feature of i-th of radar scanning target cycle relating dot mark, Azik' it is rate of azimuth change.
(2-4-3)：Calculate the root-mean-square value of target radial range rate and rate of azimuth change to be selected.
X in formulaiIndicate target radial range rate value or rate of azimuth change value,It is target radial range rate
The mean value of mean value or rate of azimuth change,
(2-4-4)：Calculate the detection probability of Targets Dots to be selected.
If target radial range rate root mean square to be selected, rate of azimuth change root mean square and point mark detection probability be satisfied by with
Lower targetpath originates constraints：
RMSEd≤RdG
RMSEd≤RaG
p≤Pd
Then target to be selected switchs to official goals flight path.And target following is maintained according to Kalman filter model, it records simultaneously
Relating dot mark feature vector in object tracking process.
Third walks：Illustrate in conjunction with Fig. 3, the minimum of tracking target A and target B is calculated using ship collision prevention radar measurement model
Meeting distance point time (TCPA) method.In plane coordinate system XOY, Y-axis is direct north.If tracking target S1Positioned at coordinate original
Point, with course d1Speed of a ship or plane V1Uniform motion.The target S detected around tracking target2With course d2Speed of a ship or plane V2Uniform motion.T1When
It carves, S2Opposite S1Position be A points, relative distance R, relative bearing θ.AA' is two target virtual courses, and the angle with Y-axis is
φr, the opposite speed of a ship or plane is Vr.φ can be calculated by geometrical relationshiprAnd Vr,
After t moment, target S is calculated2Relative target S1Displacement,
By Δ X, Δ Y derives S2Opposite S1Distance and bearing,
In formula, Xt0And Yt0For t moment official goals S2Initial position, Δ X and Δ Y indicate target S respectively2In t moment
The position of interior X-direction and Y-direction movement.It, can be by t moment radar to target S in practical engineering application2Detection obtain position
Confidence ceases R (t) and θ (t).By relative distance and orientation, TCPA is calculated by formula,
TCPA=R (t) | cos (φr-θ(t))|/Vr
Illustrate to judge the state of motion flow chart between tracking target A and its surrounding objects B according to TCPA values in conjunction with Fig. 4.
GT1For target echo overlapping time, GT2For target echo time to approach.Target situation deterministic process is illustrated in conjunction with Fig. 5.If
GT1=20, GT2=40.If TCPA absolute values are less than GT1, then track target A and target B and be judged as that state can be met；If TCPA is exhausted
GT is more than to value1Less than GT2, then track target A and target B and be judged as proximity state；If TCPA absolute values are more than GT2, then track
Target A and target B is judged as discrete state.GT is arranged according to radar period in ordinary circumstance1And GT2.The present invention is in engineer application
Middle GT1It is set as 1 minute, GT2It is set as 5 minutes.Illustrate to track target situation song during multiple target overtakes tracking in conjunction with Fig. 6
The variation of line chart, Fig. 7 are the situation curve variation diagram for overtaking target 0009 batch.When the TCPA of tracking target 0009 batch is more than GT1
When, tracking 0009 situation curve of target is discrete state.Within the 10th to 35 scan period of radar, tracking target 0009 batch with just
Formula target echo does not detach, and 0009 batch of tracking target TCPA value is less than GT1When, situation curve holding can meet state.When radar exists
When the 36th scan period, tracking target completion overtakes process.0009 batch tracks target and official goals situation curve again more
It is newly discrete state.Illustrate that multiple target intersects the variation that tracking tracks target situation curve figure in the process in conjunction with Fig. 8, Fig. 9 is to hand over
Pitch the situation curve variation diagram of target 0015 batch.When the TCPA of tracking target 0015 batch is more than GT1When, track 0015 batch of state of target
Power curve is discrete state.When tracking target 0015 batch is close with 0018 batch of echo of official goals, TCPA value result of calculations are situated between
Enter GT1And GT2Between, tracking 0015 batch of situation curve of target is displayed in proximity to state.In the 57th to 77 scan period of radar, with
The 0018 batch of echo overlapping of track target 0015 and official goals, TCPA value result of calculations are less than GT1, 0015 batch of situation song of tracking target
Line is that can meet state.As radar scanning and target move, tracks target and official goals echo is gradually disengaged, track target
0015 situation curve is shown by that can meet, and is become close, is eventually become discrete state.
4th step determines the associating policy and association algorithm of tracking target A point mark flight paths according to target state of motion information.
If tracking target A and its surrounding objects B is in discrete state, illustrate that it is not in that can meet or hand over to track target A and target B
Pitch scene.It tracks target A and carries out data correlation by optimal nearest neighbor algorithm.If track target A and its surrounding objects B be in it is close or
When person's overlap condition, in fact it could happen that two kinds of point marks are associated with situation.The first, the different point mark of each auto correlation of target；Second, mesh
Mark identical mark of association.For the first situation, without considering point mark feature.The second situation need to consider a mark feature.It is false
If tracking target A point mark energy feature collection is combined into { E1,E2,...,En, orientation expanded features are combined into { l1,l2,...,ln, n
To track the count value of target association point mark.Target B point mark energy feature collection is combined into { E'1,E2',...,Em', orientation extension is special
Collection is combined into { l1',l2',...,lm', m is the count value for tracking target association point mark.
(4-1)：The average energy value of target A, target B relating dot marks are calculated separately by formula,
(4-2)：Target A is calculated separately by formula, the echo bearing of target B relating dot marks extends mean value.
(4-3)：By common point mark ENERGY Ed, orientation extend ldRespectively with tracking target A, target B average energy values
Orientation extends mean valueRow compares.If meeting following condition, illustrate that common point mark is two target echoes overlapping point mark, then it is public
Concurrent mark is not assigned to target A, target B, two target extrapolation process.Otherwise, it is handled by the 4th step.G in formulaEFor energy threshold, Gl
For energy threshold.Ordinary circumstance is configured G according to the target energy of radar detection and orientation extensionEAnd Gl。
(4-4)：Calculate following equalities.
If ΔA≤ΔB, point mark is tracking target A relating dot marks.If ΔA≥ΔB, point mark is target B relating dot marks.It is polynary
Characteristic correlating method is disposed.
Although the present invention is illustrated and has been described with regard to preferred embodiment, it is understood by those skilled in the art that
Without departing from scope defined by the claims of the present invention, variations and modifications can be carried out to the present invention.
Claims (4)
1. the method for tracking target based on target state of motion information data associating policy, which is characterized in that include the following steps：
Step 1, targetpath to be selected is created：Centered on to track target, r believes for Searching point mark in the circle regional extent of radius
Breath creates targetpath to be selected if point mark belongs to the region using this mark；Otherwise targetpath to be selected is not created；
Step 2, targetpath to be selected is handled into line trace：Within the detections of radar period, if targetpath to be selected meets target
Track initiation constraints, then switch to official goals flight path by targetpath to be selected, otherwise deletes targetpath to be selected；
Step 3, estimation is filtered to tracking target and official goals flight path to handle, keep away by ship using Kalman filter model
The least meeting distance point time TCPA that radar measurement model calculates tracking target and official goals is touched, by TCPA values and target point
It is compared from time threshold, target time to approach thresholding and target overlapping time thresholding, obtains tracking target and official goals
Residing movement state information between each other；
Step 4：Tracking Targets Dots boat is determined according to tracking target and the mutual residing movement state information of official goals
The associating policy and association algorithm of mark finally obtain tracking target and put the incidence relation of mark, to realize target following；
Step 2 includes the following steps：
Step 2-1, it is assumed that in t1Moment creates target starting point to be selected, from t2Moment is treated using optimal nearest neighbor algorithm
Target is selected to carry out a mark track association：
Wherein, z indicates radar detection value, S-1New breath covariance, d are filtered for target to be selected2(z) it is new breath weighted norm,Indicate estimation of the k moment to k+1 moment target prediction states；
If d2(z)≤γ is set up, then judges target association to be selected to available point mark, otherwise, it is determined that Targets Dots to be selected are lost, γ
For distance threshold value, if target association to be selected records target association point mark information to be selected, target association to be selected to available point mark
Point mark information includes point mark radial distance g1, point mark orientation g2, point mark energy g3, point mark orientation extend g4, point mark extended distance g5
With a mark time g6, the target association point mark information to be selected is formed into relating dot mark feature vector G, is shown below：
G=[g1 g2 g3 g4 g5 g6]；
Step 2-2 estimates the filter state and predicted state of target to be selected according to Kalman filter formulation：
K (k)=P (k | k-1) H'(k) [H (k) P (k | k-1) H'(k)+R (k)]-1
P (k | k-1)=Φ (k | k-1) P (k-1 | k-1) Φ ' (k | k-1)+Q (k)
P (k | k)=[I-K (k) H (k)] P (k | k-1)
Wherein, Z (k) indicates the co-ordinate position information of k moment sensors detection target, when Φ (k | k-1) indicates that the k-1 moment is to k
Target state transfer matrix is carved, H (k) indicates that target measurement matrix, U (k) indicate that input control item matrix, u (k) indicate
Know that input or control signal, R (k) indicate that zero-mean, White Gaussian measure the covariance of noise, Q (k) is zero-mean, white
The covariance of Gaussian process noise,Indicate the state estimation vector at target k moment,Indicate the k-1 moment
To the prediction of the dbjective state at k moment, K (k) indicates k moment filtering gains, and P (k | k) indicates the covariance matrix at k moment, P (k |
K-1) indicate that prediction covariance matrix, I are unit matrix, H'(k) be target measurement matrix transposed matrix, Φ ' (k | k-1) is
The transposed matrix of state-transition matrix；
Step 2-3 repeats step 2-1~step 2-2, until tNMoment, N are the detections of radar period, it is assumed that target to be selected is N number of
M relating dot mark is shared in the detections of radar period, relating dot mark sequence is expressed as [G1,G2...GM]；
Step 2-4 calculates target proximity association point trace to be selected to range rate according to following formula：
Wherein, gi1It indicates first feature of i-th of radar scanning period target association point mark to be selected, that is, puts mark radial distance,
gj1Indicate first feature of j-th of radar scanning period target association point mark to be selected, gi6Indicate i-th of radar scanning period
6th feature of target association point mark to be selected puts mark time, gj6Indicate j-th of radar scanning period target association point to be selected
6th feature of mark, i=j+1, j=1,2 ..., M-1, tk+1-tkIndicate vk+1With vkBetween time difference, vkIndicate the k moment
Target radial is apart from speed, k=1,2 ..., n-2, akIndicate target radial range rate；
Step 2-5 calculates the rate of azimuth change of target proximity association point mark to be selected according to following formula：
Azik=gi2-gj2,
Azik'=Azik+1-Azik,
Wherein, gi2And gj2Orientation values and j-th of the radar scanning week of i-th radar scanning target cycle relating dot mark are indicated respectively
The orientation values of phase target association point mark, AzikIndicate j-th of radar scanning target cycle point mark bearing variation value, Azik' expression side
Position change rate；
Step 2-6 calculates separately to obtain n target range rate of change value and rate of azimuth change according to step 2-4 and step 2-5
Value, is calculated as follows the root mean square of radial distance change rate and the root mean square RMSE of rate of azimuth change：
Wherein, XiIndicate target range change rate or rate of azimuth change,It is target radial range rate mean value or orientation
The mean value of change rate；
Step 2-7 calculates the detection probability p of target association point mark to be selected：
Step 2-8, if target radial range rate root mean square thresholding is RdG, target bearing change rate root mean square thresholding is RaG,
Target detection probability thresholding is PdIf the root mean square RMSE of the radial distance change rate of target proximity association point mark to be selectedd, orientation
The root mean square RMSE of change rateaIt is satisfied by the following conditions with detection probability p：
RMSEd≤RdG,
RMSEd≤RaG,
p≤Pd,
Targetpath to be selected is then switched into official goals flight path, and target following is maintained according to Kalman filter model, is remembered simultaneously
Record the relating dot mark feature vector during targetpath tracking.
2. according to the method described in claim 1, it is characterized in that, step 3 includes the following steps：
Step 3-1, when calculating the least meeting distance point for tracking target and official goals using ship collision prevention radar measurement model
Between TCPA：Plane coordinate system XOY is established, Y-axis is direct north, if tracking target S1Positioned at coordinate origin O and with course d1The speed of a ship or plane
V1Uniform motion tracks the official goals S detected around target2With course d2Speed of a ship or plane V2Uniform motion, T1Moment, official goals
S2Opposite tracking target S1Position be A points, relative distance R, relative bearing θ, AA' is two target virtual courses, AA' and Y
The angle of axis is φr, the opposite speed of a ship or plane is Vr, φ is calculated using following formularAnd Vr：
After t moment, official goals S is calculated2Opposite tracking target S1Displacement：
Official goals S is calculated by Δ X and Δ Y2Opposite tracking target S1Distance R (t) after t moment and orientation θ (t)：
Wherein, Xt0And Yt0T moment official goals S is indicated respectively2Initial position abscissa and t moment official goals S2Starting
Position ordinate, Δ X and Δ Y indicate target S respectively2The position of X-direction moves in t moment position and Y-direction movement, leads to
Cross the least meeting distance point time TCPA that following formula calculates tracking target and official goals：
TCPA=R (t) | cos (φr-θ(t))|/Vr；
Step 3-2, if GT1For target echo overlapping time, GT2For target echo time to approach, if TCPA absolute values are less than GT1,
Then judge to track target and official goals for state can be met；If TCPA absolute values are more than GT1Less than GT2, then judge track target with
Official goals are proximity state；If TCPA absolute values are more than GT2, then judge that it is discrete state to track target with official goals.
3. according to the method described in claim 2, it is characterized in that, step 4 includes the following steps：
Step 4-1 carries out data using optimal nearest neighbor algorithm if tracking target and target are in discrete state to tracking target
Association；
Step 4-2 needs to judge tracking target and formal mesh if tracking target and official goals are in close or overlap condition
Whether mark is associated with common point mark, if different point marks is respectively associated in tracking target and official goals, at normal tracking
Reason；If tracking target and official goals relating dot mark be common point mark, judge following conditions whether and meanwhile set up：
Wherein,WithThe energy of the average energy value and official goals relating dot mark that indicate tracking target association point mark respectively is equal
Value,WithThe echo side of echo bearing the extension mean value and official goals relating dot mark of tracking target association point mark is indicated respectively
Bits Expanding mean value, EdFor the observable common point mark energy value of radar, ldIt is extended for radar observable common point mark orientation
Value, α and β are proportionality coefficient, GEFor energy error thresholding, GlError threshold, if condition is set up simultaneously, point are extended for orientation
Mark is tracking target and an official goals echo overlapping point mark, and point mark is not assigned to tracking target and official goals, target respectively into
Row extrapolation process；If condition is invalid, mark is put from tracking target or official goals, Δ is calculated as followsA
And ΔB：
ΔAIndicate the similitude of the energy and tracking target energy mean value of common point mark, ΔBIndicate common point mark energy with just
The similitude of formula target energy mean value, if ΔA≤ΔB, point mark is tracking target association point mark, if ΔA＞ ΔsB, point mark is formal
Target association point mark.
4. according to the method described in claim 3, it is characterized in that, in step 4-1, when tracking target and official goals are in point
When from state, data correlation is carried out using optimal nearest neighbor algorithm, if tracking target association point mark energy feature collection is combined into { E1,
E2,...,En, echo bearing expanded features are combined into { l1,l2,...,ln, n is the count value for tracking target association point mark, just
Formula target association point mark energy feature collection is combined into { E'1,E2',...,Em', echo bearing expanded features are combined into { l1',
l2',...,lm', m is the count value of official goals relating dot mark, is carried out in orientation, apart from upper continuity using target echo
Plot coherence, obtains target bearing extension feature, and energy feature is carried out by the corresponding energy of the azran that will participate in accumulation
Summation obtains,
The average energy value of tracking target association point mark is calculated by following formulaWith the average energy value of official goals relating dot mark
The echo bearing that tracking target association point mark is calculated by following formula extends mean valueWith returning for official goals relating dot mark
Wave orientation extends mean value
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610721016.0A CN106249232B (en) | 2016-08-24 | 2016-08-24 | Method for tracking target based on target state of motion information data associating policy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610721016.0A CN106249232B (en) | 2016-08-24 | 2016-08-24 | Method for tracking target based on target state of motion information data associating policy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106249232A CN106249232A (en) | 2016-12-21 |
| CN106249232B true CN106249232B (en) | 2018-08-21 |
Family
ID=57595546
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610721016.0A Active CN106249232B (en) | 2016-08-24 | 2016-08-24 | Method for tracking target based on target state of motion information data associating policy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106249232B (en) |
Families Citing this family (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180275265A1 (en) * | 2017-03-23 | 2018-09-27 | GM Global Technology Operations LLC | Target tracking using region covariance |
| CN106932771A (en) * | 2017-03-30 | 2017-07-07 | 成都紫瑞青云航空宇航技术有限公司 | A kind of radar simulation targetpath tracking and system |
| CN107192998B (en) * | 2017-04-06 | 2019-05-31 | 中国电子科技集团公司第二十八研究所 | A kind of adapter distribution track data fusion method based on covariance target function |
| CN107896136B (en) * | 2017-10-27 | 2020-09-08 | 南京莱斯电子设备有限公司 | Radar track message encoding method |
| CN107911196B (en) * | 2017-10-27 | 2020-07-14 | 南京莱斯电子设备有限公司 | Radar track message transmission method |
| CN108344992B (en) * | 2017-12-20 | 2020-03-27 | 北京华航无线电测量研究所 | Multi-target tracking method for vehicle-mounted millimeter wave radar |
| CN108226863B (en) * | 2017-12-27 | 2021-12-03 | 武汉理工大学 | Monocular satellite tracking and positioning method |
| CN110031834A (en) * | 2018-01-12 | 2019-07-19 | 西安艾索信息技术有限公司 | A kind of improved multiple target radar track processing method |
| CN108344997B (en) * | 2018-01-16 | 2021-10-29 | 华域汽车***股份有限公司 | Road guardrail rapid detection method based on trace point characteristics |
| CN108303684B (en) * | 2018-01-31 | 2021-09-03 | 长沙深之瞳信息科技有限公司 | Multi-target tracking method for ground monitoring radar based on radial velocity information |
| CN108680918B (en) * | 2018-05-18 | 2020-08-28 | 森思泰克河北科技有限公司 | Speed measuring method and device applied to radar and electronic equipment |
| CN109001725B (en) * | 2018-06-07 | 2020-11-10 | 中国人民解放军海军工程大学 | Offshore unmanned ship offshore multi-target tracking method |
| CN109031269B (en) * | 2018-06-08 | 2020-07-07 | 上海西井信息科技有限公司 | Positioning method, system, equipment and storage medium based on millimeter wave radar |
| CN109509214B (en) * | 2018-10-15 | 2021-07-16 | 杭州电子科技大学 | Ship target tracking method based on deep learning |
| CN110400491A (en) * | 2019-06-10 | 2019-11-01 | 北京海兰信数据科技股份有限公司 | A kind of Open sea area multiple target auxiliary Decision of Collision Avoidance method and decision system |
| CN110646798B (en) * | 2019-09-23 | 2021-09-28 | 森思泰克河北科技有限公司 | Target track association method, radar and terminal equipment |
| CN110907903B (en) * | 2019-10-30 | 2023-08-01 | 中电科技（合肥）博微信息发展有限责任公司 | Self-adaptive tracking processing method based on multiple sensors |
| CN110940971B (en) * | 2019-11-06 | 2022-05-31 | 四川川大智胜软件股份有限公司 | Radar target point trace recording method and device and storage medium |
| CN111323757B (en) * | 2019-12-30 | 2022-04-05 | 北京海兰信数据科技股份有限公司 | Target detection method and device for marine radar |
| CN111289954B (en) * | 2020-03-31 | 2022-03-15 | 四川长虹电器股份有限公司 | Point cloud division and track matching method for millimeter wave radar target tracking |
| CN111707997B (en) * | 2020-06-03 | 2022-05-27 | 南京慧尔视智能科技有限公司 | Radar target tracking method and device, electronic equipment and storage medium |
| CN111751816B (en) * | 2020-07-07 | 2023-03-21 | 北京无线电测量研究所 | Radar track starting method based on trace point characteristic information |
| CN112363167A (en) * | 2020-11-02 | 2021-02-12 | 重庆邮电大学 | Extended target tracking method based on fusion of millimeter wave radar and monocular camera |
| CN112505681B (en) * | 2020-11-02 | 2022-11-25 | 南京理工大学 | Four-side two-dimensional phased array radar multi-target tracking processing method based on TAS |
| CN112800349B (en) * | 2021-02-02 | 2022-02-18 | 中华人民共和国广东海事局 | Method, device, equipment and medium for acquiring motion state of aquatic moving target |
| CN113744309B (en) * | 2021-08-11 | 2024-05-24 | 中国科学院空天信息创新研究院 | Maneuvering target tracking method based on motion state change perception |
| CN113702940B (en) * | 2021-09-18 | 2022-06-21 | 中国人民解放军63921部队 | Spatial cluster target resolution method based on multi-element characteristic information hierarchical fusion and application |
| CN113721237B (en) * | 2021-11-02 | 2022-02-08 | 南京雷电信息技术有限公司 | Multi-membership-degree target intelligent matching algorithm |
| CN117554922B (en) * | 2024-01-12 | 2024-03-26 | 航天宏图信息技术股份有限公司 | Method and device for associating target tracks, electronic equipment and computer storage medium |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7142981B2 (en) * | 2003-08-05 | 2006-11-28 | The Boeing Company | Laser range finder closed-loop pointing technology of relative navigation, attitude determination, pointing and tracking for spacecraft rendezvous |
| US7417583B2 (en) * | 2006-10-30 | 2008-08-26 | Raytheon Company | Methods and apparatus for providing target altitude estimation in a two dimensional radar system |
| CN103727931A (en) * | 2013-12-31 | 2014-04-16 | 南京理工大学 | Improved logic-based track initiation method |
| CN105388472B (en) * | 2015-11-20 | 2018-02-13 | 中国电子科技集团公司第二十八研究所 | One kind is to the method for tracking target that is overlapping in extra large radar system |
-
2016
- 2016-08-24 CN CN201610721016.0A patent/CN106249232B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN106249232A (en) | 2016-12-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106249232B (en) | Method for tracking target based on target state of motion information data associating policy | |
| CN109100714B (en) | Low-slow small target tracking method based on polar coordinate system | |
| Newman et al. | Towards constant-time SLAM on an autonomous underwater vehicle using synthetic aperture sonar | |
| CN106407677B (en) | A kind of multi-object tracking method in the case of missing measurement | |
| CN109633590B (en) | Extended target tracking method based on GP-VSMM-JPDA | |
| CN106980114A (en) | Target Track of Passive Radar method | |
| CN103729859A (en) | Probability nearest neighbor domain multi-target tracking method based on fuzzy clustering | |
| CN107192998A (en) | A kind of adapter distribution track data fusion method based on covariance target function | |
| KR101628154B1 (en) | Multiple target tracking method using received signal strengths | |
| CN108957437B (en) | Radar target tracking method based on environmental situation analysis | |
| CN107861123A (en) | A kind of through-wall radar is under complex environment to the method for multiple mobile object real-time tracking | |
| CN105301584B (en) | The IPPHDF multiple-moving target tracking methods of fuzzy distance solution simultaneously | |
| CN108734725A (en) | Probabilistic contractor couple based on Gaussian process extends method for tracking target | |
| CN107436434B (en) | Track starting method based on bidirectional Doppler estimation | |
| CN108614580A (en) | Layered obstacle avoidance control method in target tracking of unmanned aerial vehicle | |
| CN109858526A (en) | Sensor-based multi-target track fusion method in a kind of target following | |
| CN108645412A (en) | A kind of adaptive track initiation method of multisensor | |
| CN109521420A (en) | Based on the matched multi-object tracking method of multiple features | |
| CN107037423A (en) | Multi-object tracking method is filtered with reference to the PHD of amplitude information | |
| CN114114242A (en) | Radar target tracking method based on point-selected flight path and flight path extrapolation | |
| CN105353352B (en) | The MM PPHDF multiple-moving target tracking methods of improved search strategy | |
| CN109509207A (en) | The method that a kind of pair of point target and extension target carry out seamless tracking | |
| Qin et al. | Research on information fusion structure of radar and AIS | |
| Chen et al. | Research on AIS and radar information fusion method based on distributed Kalman | |
| CN106772357A (en) | AI PHD wave filters under signal to noise ratio unknown condition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |