CN106990398A - A kind of body of revolution fine motion feature awareness extracting method - Google Patents

A kind of body of revolution fine motion feature awareness extracting method Download PDF

Info

Publication number
CN106990398A
CN106990398A CN201610041784.1A CN201610041784A CN106990398A CN 106990398 A CN106990398 A CN 106990398A CN 201610041784 A CN201610041784 A CN 201610041784A CN 106990398 A CN106990398 A CN 106990398A
Authority
CN
China
Prior art keywords
fine motion
target
detection
motion feature
fitting
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201610041784.1A
Other languages
Chinese (zh)
Other versions
CN106990398B (en
Inventor
陈怡君
张群
李治安
罗迎
马志强
封同安
李开明
王恺
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.)
Air Force Engineering University of PLA
Original Assignee
Air Force Engineering University of PLA
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 Air Force Engineering University of PLA filed Critical Air Force Engineering University of PLA
Priority to CN201610041784.1A priority Critical patent/CN106990398B/en
Publication of CN106990398A publication Critical patent/CN106990398A/en
Application granted granted Critical
Publication of CN106990398B publication Critical patent/CN106990398B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • 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
    • G01S7/415—Identification of targets based on measurements of movement associated with the target

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

A kind of body of revolution fine motion feature awareness extracting method is provided, including：The first step：Target scattering point is obtained apart from trace information using tracking before detection；Second step：Fine motion parameter is fitted using apart from trace information, in turn, the state during being tracked before detection is updated according to fine motion parameter estimation result and shifts set function；3rd step：Setting decision rule is considered in terms of target detection and fine motion feature extraction two, to judge whether target detection and fine motion feature extraction tasks complete.This method can realize target fine motion feature extraction while detecting and tracking is carried out to target, improve radar operating frequency.

Description

A kind of body of revolution fine motion feature awareness extracting method
Technical field
The present invention relates to Signal and Information Processing technology, and in particular to a kind of body of revolution fine motion feature awareness is extracted Method.
Background technology
Since Chen V. C. are by micro-doppler concept introducing radar detection field, radar target fine motion feature extraction is just Cause the extensive concern of domestic and foreign scholars.Encourage text and wait in the village《Achievements of Target Characteristic with Micro-Motion》（Electronic letters, vol, 2007, 35(3): 520–525）In mention fine motion feature and be typically considered the unique motion feature that radar target has, This fine motion feature is extracted using high resolution radar and modern signal processing technology, can be the non-cooperative target of radar Mark detection provides new approach with identification.Gao H. W. etc. exist《Micro-Doppler signature extraction form ballistic target with micro-motion》（IEEE Transactions on Aerospace and Electronic Systems, 2010, 46(4): 1969-1982）In by setting up extraterrestrial target micro-motion model, analyze The micro-doppler characteristic of target echo under different fine motion forms, and verified by darkroom measurement data.Liu Y. X. etc. 《Estimation of micro-motion parameters based on micro-Doppler》（IET Signal Processing, 2010,4 (3): 213-217）In have studied time frequency distribution map into moving-target, and use parameter transformation method Realize effective extraction of target fine motion feature.Cao Wenjie etc. exists《Based on instantaneous Frequency Estimation to enter movable cone target micro- how general Strangle frequency extraction method》（Electronics and information journal, 2015,37 (5): 1091-1096）It is micro- how general that middle combination precession is modulated The characteristics of strangling frequency near sinusoidal changing rule, on the basis of echo-signal is segmented, it is proposed that based on instantaneous Frequency Estimation Enter moving-target micro-doppler frequency extraction method with random sampling uniformity, effectively increase carrying for target micro-doppler frequency Take precision.
However, existing radar target fine motion feature extracting method is carried out on the basis of target detection and tracking is completed , this is accomplished by distributing radar resource respectively to target detection, tracking and feature extraction.If can be by target detection, tracking It is combined with fine motion feature extraction, while target detection is completed, realizes the extraction of target fine motion feature, thunder will be effectively improved Up to resource utilization and the real-time of fine motion feature extraction.
In recent years, track algorithm achieves good effect on Faint target detection and tracking problem before detecting, the algorithm Accumulated by interframe, improve the detection probability of target, and target trajectory detecting target while recalling.If energy Body of revolution fine motion feature extraction processing is introduced during enough tracking before detection, utilizes target trajectory to extract fine motion Feature, in turn, instructs the state during being tracked before detection to shift according to the extraction result of fine motion feature and collects, be expected to significantly carry How high operation efficiency and resource utilization, while realizing the detection and fine motion feature extraction of body of revolution, but realize Radar can realize that target fine motion feature extraction also there are problems that while detect and track is carried out to target.
The content of the invention
It is an object of the invention to overcome above-mentioned weak point of the prior art to propose a kind of body of revolution fine motion Feature awareness extracting method.
The present invention is realized in the following way：
A kind of body of revolution fine motion feature awareness extracting method, comprises the following steps：
The first step：Target scattering point is obtained apart from trace information using tracking before detection；
Second step：Fine motion parameter is fitted using apart from trace information, in turn, is updated according to fine motion parameter estimation result before detection State transfer set function during tracking；
3rd step：Setting decision rule is considered in terms of target detection and fine motion feature extraction two, to judge target detection Whether completed with fine motion feature extraction tasks.
The first step is specifically included：
If circular cone target centroid is,Axle is in electromagnetic wave incident directionWithIn the plane that axle is constituted, according to right-handed helix Law is set upAxle, target is with angle of precession, precession frequencyAround z-axis precession, radar wave incident direction,For average-visual angle, target is made up of 3 scattering centers, respectively vertex of a cone scattering center, And cone bottom scattering centerWith, bottom radius of circle is, barycenter arrivesPoint distance is designated as, barycenter on earth circle it is vertical away from From being designated as；
Assuming that t=0 moment, Projection in plane withThe angle of axle is, then t,On radar line of sight Be projected as
WhereinFor target radial speed；
WithBeing projected as on radar line of sight
Wherein
Enter row distance to radar return data to handle to pulse pressure, if radarFrame（I.e.Individual pulse）Obtained target is one-dimensional Range Profile byIndividual resolution cell is constituted, theMeasuring value on individual resolution cell is, thenThe measuring value of frame For
BeforeThe measuring value of frame is represented by
If stateBefore correspondingFrame energy accumulation value is, then theFrame stateEnergy accumulation value It is represented by
Wherein,To can be transferred to stateAll possible statesSet；To accumulationTarget letter after frame NumberThresholding is set, and the state that thresholding is more than to energy accumulation functional value carries out flight path backtracking, involved by trace-back process State can be obtained by following formula
Wherein,For backtracking function, energy accumulated value is set to reach maximum dbjective state, mesh for recording each stage Mark status switch is the estimate of target range track.
The second step is specifically included：
Recall for each apart from track, respectively according to the vertex of a cone with cone two kinds of different distances of bottom scattering point with The changing rule of time fits two groups of parameters：、、、、、, and、、、、、、, define two groups of parameters error of fitting be respectively：
Wherein
Compare error of fittingWithIf,, then it is assumed that this is formed apart from track by summit scattering point, can extract Go out fine motion characteristic parameter、、、、、；It is on the contrary, then it is assumed that this is scattered apart from track by the slip for boring baseplane Point is formed, and can extract out fine motion characteristic parameter、、、、、、；
On this basis, it is anti-by setting up information before fine motion characteristic parameter extraction result and target detection between tracking process Feedback, updates dbjective state transfer function；
If theThe vertex of a cone scattering point fine motion characteristic parameter extracted during frame is、、、、、；Extract Cone bottom scattering point fine motion characteristic parameter be、、、、、、；Then for, it is special according to fine motion State transfer collection can be determined by levying parameter extraction resultFor
Wherein
For searching scope adaptive Tuning function, the size of detection range door during being tracked before control detection；FunctionWithFine motion characteristic parameter error of fitting is independent variable during frame, when error of fitting is larger, is illustrated to next frame distance The precision of information prediction is relatively low, and detection range door should suitably increase；When error of fitting is larger, illustrate to next frame range information The precision of prediction is higher, can suitably reduce detection range door to reduce algorithm operation quantity；Therefore,It should be increasing function, It is defined as
WhereinFor detection range door minimum value.
3rd step is specifically included：
Two detection thresholds are set：Low detection thresholdWith high detection thresholding；When energy accumulation value is more than high detection door LimitWhen, judgement target is present, and fitting obtains target fine motion characteristic parameter；When energy accumulation value is less than high detection thresholding When, judge that target whether there is according to fine motion feature extraction precision.
3rd step further comprises：
WhenFrame energy accumulation valueIt is more thanWhen, backtracking obtain target scattering point apart from track estimate, point Two groups of fine motion characteristic parameters are not fitted apart from track form according to vertex of a cone scattering point and cone bottom scattering point, compare two kinds of fittings Error, determines scattering point attribute, so as to obtainThe fine motion feature extraction result of scattering point during frame, and error of fitting is designated as, on this basis, shifted and collected using fine motion characteristic parameter extraction result more new state, to theFrame data carry out energy Accumulate and recall target range track, by curve matching and application condition, obtain theThe fine motion feature of scattering point during frame Result is extracted, and error of fitting is designated as；If error of fitting very little, and to the prediction foot of the next frame location information of target It is enough accurate, then it is assumed that fine motion feature extraction precision has reached requirement, while can determine that target is present, it is not necessary to carry out energy again and tire out Product；Otherwise it is assumed that cumlative energy is not enough, it is necessary to continue pairFrame data carry out energy accumulation and repeated the above steps；
Fine motion feature extraction precision judgment variables are defined as
WhereinIt is error of fitting thresholding,For according toThe fine motion feature extraction prediction of result of scattering center is obtained during frame Frame scattering center range information,ForObtained target range track is recalled after frame energy accumulation；When Energy accumulation value is more thanOr fine motion feature extraction precision judgment variables be 1 when, think target exist, realize to target Fine motion characteristic parameter extraction.
Preferably taken in the first step,,,,,,。
Preferably taken in the second step,。
Preferably taken in 3rd step,,。
Scattering point is referred to as " apart from track " apart from versus time curve in the present invention.On this basis, pass through Set up closed loop feedback target fine motion feature extraction is incorporated into before target detection during tracking, that is, utilize obtained target to dissipate Exit point apart from trace information be fitted fine motion characteristic parameter, in turn, according to fine motion characteristic parameter extraction result update detection before with State transfer set function in track algorithm.Finally, by set comprehensive decision rule, to judge that target detection and fine motion feature are carried Take whether task completes, so that target fine motion feature is effectively extracted while target detection, can be right so as to solve radar Target carries out realizing the problem of target fine motion feature extraction is present while detect and track.
Brief description of the drawings
Fig. 1 shows target geometrical model；
Fig. 2 shows a kind of indicative flowchart of body of revolution fine motion feature awareness extracting method；
Fig. 3 shows that target scattering point distance changes with time rule；
Fig. 4 is shown using tracking is obtained before detection three apart from track；
Fig. 5 shows each fine motion parameter fitting error apart from track；
Fig. 6 shows the fine motion parameter fitting error change process of each track；
Fig. 7 shows the predicted position and the error change process apart from track of each track；
Fig. 8 shows the detection range door change procedure of each track.
Embodiment
With reference to embodiment, accompanying drawing, the invention will be further described.
The present invention method be：The first step：Target scattering point is obtained apart from trace information using tracking before detection；The Two steps：Structure is utilized is fitted fine motion parameter apart from trace information, in turn, is updated and is tracked before detection according to fine motion parameter estimation result During state transfer set function；3rd step：Setting judgement is considered in terms of target detection and fine motion feature extraction two Criterion, to judge whether target detection and fine motion feature extraction tasks complete.
Realize comprising the following steps that for the invention described above method：
The first step：As depicted in figs. 1 and 2,For target centroid,Axle is in electromagnetic wave incident directionWithThe plane that axle is constituted It is interior, and according to screw law set upAxle, target is with angle of precession, precession frequencyAround z-axis precession, radar Ripple incident direction,For average-visual angle, the position of 3 scattering centers of circular cone target is Fig. 1 In,With, bottom radius of circle is, barycenter arrivesPoint distance is designated as, the vertical range that barycenter is justified on earth is designated as；
Assuming that t=0 moment, Projection in plane withThe angle of axle is, then t,On radar line of sight Be projected as
（1）
WhereinFor target radial speed；
Cone baseplane normal direction be
（2）
WithConstitute plane normal direction be
（3）
WithConstitute plane equation be；
WithTwo scattering centers existWithOn intersecting lens of the plane of composition with boring baseplane, thenWithTwo dissipate The vector for hitting the heart is
（4）
Wherein；
By formula（2）-（4）Calculate, can obtainWithBeing projected as on radar line of sight
（5）
As can be seen that vertex of a cone scattering centerShow as sinusoidal form apart from track, and boundary slip scattering centerWith Had deviated from standard sine form apart from track.Target scattering point can effectively be extracted apart from rail using tracking before detection Mark, due to containing the information such as target fine motion parameter and structural parameters in track, and then can be according to target scattering point distance Path implementation target fine motion feature awareness is extracted.
Enter row distance to radar return data first to handle to pulse pressure, vertex of a cone scattering centerApart from track in sinusoidal bent Line form（Such as formula（1）It is shown）, by parameter、、、、WithTogether decide on；In the slip scattering point for boring baseplane The heartWithDeviateed standard sine form apart from track（Such as formula（5）It is shown）, by parameter、、、、、、 Together decide on.
If radarFrame（I.e.Individual pulse）Obtained target one-dimensional range profile byIndividual resolution cell is constituted, the Measuring value on individual resolution cell is；ThenThe measuring value of frame is
（6）
BeforeThe measuring value of frame is represented by
（7）
If stateBefore correspondingFrame energy accumulation value is, then theFrame stateEnergy accumulation value It is represented by
（8）
Wherein,To can be transferred to stateAll possible statesSet, it is true by target range changing rule It is fixed；To accumulationObject function after frameThresholding is set, and the state that thresholding is more than to energy accumulation functional value carries out flight path Backtracking；State involved by trace-back process can be obtained by following formula
（9）
Wherein,For backtracking function, energy accumulated value is set to reach maximum dbjective state for recording each stage；
Dbjective state sequence is the estimate of target range track, and target scattering point effectively obtaining apart from trace information can be achieved Take.
Second step：When targets are present, for scattering center, recall should be formula apart from track（1）It is shown just Chord curve form, for scattering centerWith, recall should be formula apart from track（5）Shown deviation is sinusoidal multiple Miscellaneous form；Recall for each apart from track, respectively according to formula（1）And formula（5）Fit two groups of parameters：、、、、、, and、、、、、、, define according to formula（1）And formula（5）Obtain Parameter fitting error be respectively：
（10）
（11）
Wherein
Compare error of fittingWithIf,, then it is assumed that this is formed apart from track by summit scattering point, can extract Go out fine motion characteristic parameter、、、、、；It is on the contrary, then it is assumed that this is scattered apart from track by the slip for boring baseplane Point is formed, and can extract out fine motion characteristic parameter、、、、、、；
On this basis, it is anti-by setting up information before fine motion characteristic parameter extraction result and target detection between tracking process Feedback, updates dbjective state transfer function；
Dbjective state transfer refers to the time interval by a frame, the location status that target is likely to occur；State transfer collectionDetermination can directly affect detection before tracking and fine motion feature extraction efficiency and performance；Joined according to target fine motion feature Number extracts result, predicts the range information of next frame target scattering point, and the cognitive renewal of state transfer collection can be achieved；
If theThe vertex of a cone scattering point fine motion characteristic parameter extracted during frame is、、、、、；Extract Cone bottom scattering point fine motion characteristic parameter be、、、、、、.Then for, it is special according to fine motion State transfer collection can be determined by levying parameter extraction resultFor
（12）
Wherein
For searching scope adaptive Tuning function, the size of detection range door during being tracked before control detection；FunctionWithFine motion characteristic parameter error of fitting is independent variable during frame, when error of fitting is larger, is illustrated to next frame distance The precision of information prediction is relatively low, and detection range door should suitably increase；When error of fitting is larger, illustrate to next frame range information The precision of prediction is higher, can suitably reduce detection range door to reduce algorithm operation quantity；Therefore,It should be increasing function, It is defined as
（13）
WhereinFor detection range door minimum value.
3rd step：Before traditional detection in track algorithm, by the way that energy accumulation value and threshold value are compared to judge Target whether there is；However, the size of known noise variance and the intensity of backward energy are needed when setting detection threshold, to door Limit setting brings difficulty；
Because the present invention is closely coupled, it is necessary to logical by target detection and fine motion characteristic extraction procedure by closed-loop information feedback loop Cross the judgment criterion that comprehensive analysis provides target presence or absence, and target fine motion feature extraction processing procedure starting and termination Condition；
In the present invention, consider energy accumulation value and fine motion characteristic parameter extraction result, traditional energy is no longer relied on merely Measure accumulated value threshold judgement method to judge target presence or absence, when fine motion feature extraction precision is sufficiently high, it also hold that mesh Mark exists and successfully extracts target fine motion feature, without carrying out energy accumulation again；Two detection thresholds are set first：Low inspection Survey thresholdingWith high detection thresholding；When energy accumulation value is more than high detection thresholdingWhen, judgement target is present, fitting Obtain target fine motion characteristic parameter；When energy accumulation value is less than high detection thresholdingWhen, can be according to fine motion feature extraction precision Judge that target whether there is, specific method is as follows：
WhenFrame energy accumulation valueIt is more thanWhen, according to formula（9）Backtracking obtains estimating apart from track for target scattering point Evaluation, respectively according to vertex of a cone scattering point and cone bottom scattering point apart from track form（Formula（1）And formula（5））Fit two groups it is micro- Dynamic characteristic parameter, compares two kinds of errors of fitting, determines scattering point attribute, so as to obtain theThe fine motion feature of scattering point is carried during frame Result is taken, and error of fitting is designated as, on this basis, shifted using fine motion characteristic parameter extraction result more new state Collection, to theFrame data carry out energy accumulation and simultaneously recall target range track, by curve matching and application condition, obtain theThe fine motion feature extraction result of scattering point during frame, and error of fitting is designated as；If error of fitting very little, and to target The prediction of next frame location information is accurate enough, then it is assumed that fine motion feature extraction precision has reached requirement, while can determine that target In the presence of, it is not necessary to energy accumulation is carried out again；Otherwise it is assumed that accumulated energy is not enough, it is necessary to continue pairFrame data carry out energy Accumulate and repeat the above steps.
In the present invention, defining fine motion feature extraction precision judgment variables is
（14）
WhereinIt is error of fitting thresholding,For according toThe fine motion feature extraction prediction of result of scattering center is obtained during frame Frame scattering center range information,ForObtained target range track is recalled after frame energy accumulation；When Energy accumulation value is more thanOr fine motion feature extraction precision judgment variables are when being 1, think that target is present, and realize to mesh Target fine motion characteristic parameter extraction；
By the above-mentioned first to the 3rd step, the fine motion feature awareness that body of revolution can be achieved is extracted, examined to target Realize that target fine motion feature is effectively extracted while surveying, track.
Example：Body of revolution fine motion Feature Extraction System
Simulation parameter is set：Radar and the geometrical model of target are as shown in figure 1, target is with angle of precession, precession frequencyAround z-axis precession, radar wave average-visual angle, target radial speed.Target is by 3 Scattering center is constituted, respectively vertex of a cone scattering center, and cone bottom scattering centerWith, bottom radius of circle is, matter The heart is arrivedPoint distance is designated as, the vertical range that barycenter is justified on earth is designated as。
Radar emission linear FM signal carrier frequency is 1GHz, and the pulsewidth of pulse signal is, with a width of 3GHz, frequency modulation rate For, radar pulse repetition frequency, signal to noise ratio is。
Parameter for above-mentioned setting is emulated, and Fig. 3 shows that target scattering point distance changes with time rule.Adopt Energy is carried out to the Range Profile of each frame echo data with the cognitive fine motion feature extracting method of the body of revolution proposed to tire out Product, sets low detection threshold value, high detection threshold value, fine motion feature extraction error of fitting thresholding.When running up to 133 frame, energy accumulation value is more than threshold value, Fig. 4 shown according to formula（9）Recall three obtained Bar is designated as track 1, track 2 and track 3 respectively from top to bottom for convenience of description apart from track.
For each track, respectively according to formula（1）And formula（5）, fine motion parameter fitting is carried out using least square method and calculated Error of fittingWith, as a result as shown in Figure 5.From figure 5 it can be seen that track 1It is more than, therefore judge to be somebody's turn to do Track corresponds to vertex of a cone scattering point；Track 2 and 3It is less than, therefore judge that this two tracks correspond to the scattering of cone bottom Point.Least square fitting is respectively adopted and goes out each scattering point fine motion parameter.Now, energy accumulation value is less than limit valueAnd fine motion Feature extraction precision judgment variablesValue is 0, it is therefore desirable to proceed energy accumulation.Obtained using estimation Fine motion characteristic parameter, according to formula（12）Update dbjective state transfer collection.When running up to 185 frame, fine motion feature extraction precision Judgment variablesValue is 1, now thinks that target exists and obtains effective target fine motion parameter extraction result.Three Bar apart from track fine motion parameter fitting error, predicted position with apart from track error and detection range door change procedure As Figure 6-Figure 8.As can be seen that with the accumulation of energy, what backtracking was obtained increases apart from trace information from Fig. 6-Fig. 8, Therefore fine motion feature extraction precision is improved, and predicted position is gradually reduced with the error apart from track, and detection range door is also with pre- Survey the raising of precision and reduce.Table 1 shows the fine motion characteristic parameter extraction result of each scattering point, is sufficiently close to theoretical value, Illustrate the validity of institute's extracting method of the present invention.
The fine motion characteristic parameter extraction result of each scattering point of table 1
Characteristic parameter 1 2.921 0 0.529 0.317 25.969 11.621
Characteristic parameter 2 0.316 0.941 0 0.511 0.321 26.807 11.489
Characteristic parameter 3 0.299 1.106 0 0.514 0.306 26.312 11.358
Scattering point is referred to as " apart from track " apart from versus time curve in the present invention.On this basis, by setting up Target fine motion feature extraction is incorporated into before target detection during tracking by closed loop feedback, that is, utilizes obtained target scattering point Fine motion characteristic parameter is fitted apart from trace information, in turn, is updated to track before detection according to fine motion characteristic parameter extraction result and calculated State transfer set function in method.Finally, by set comprehensive decision rule, to judge that target detection and fine motion feature extraction are appointed Whether business completes, so that target fine motion feature is effectively extracted while target detection, can be to target so as to solve radar Carry out realizing the problem of target fine motion feature extraction is present while detect and track.

Claims (8)

1. a kind of body of revolution fine motion feature awareness extracting method, it is characterised in that：Comprise the following steps：
The first step：Target scattering point is obtained apart from trace information using tracking before detection；
Second step：Fine motion parameter is fitted using apart from trace information, in turn, is updated according to fine motion parameter estimation result before detection State transfer set function during tracking；
3rd step：Setting decision rule is considered in terms of target detection and fine motion feature extraction two, to judge target detection Whether completed with fine motion feature extraction tasks.
2. according to the method described in claim 1, it is characterised in that：The first step is specifically included：
If circular cone target centroid is o, y-axis is in electromagnetic wave incident directionIn the plane constituted with z-axis, built according to screw law Vertical x-axis, target is with angle of precession θ, precession frequency ω rad/s around z-axis precession, radar wave incident direction β is average-visual angle, and target is made up of 3 scattering centers, respectively vertex of a cone scattering center a, and cone bottom scattering center c and d, Bottom radius of circle is r0, barycenter is designated as to a point distances | oa |, the vertical range that barycenter is justified on earth is designated as | ob |；
Assuming that the t=0 moment,The angle of projection and x-axis in xoy planes is φ0, then t,On radar line of sight It is projected as
ra=| oa | (sin β sin θs sin (ω t+ φ0))+vt
Wherein v is target radial speed；
WithBeing projected as on radar line of sight
Wherein
Enter row distance to radar return data to handle to pulse pressure, if the obtained target of radar kth frame (i.e. k-th pulse) it is one-dimensional away from From as by RnIndividual resolution cell is constituted, and the measuring value on i-th of resolution cell is HRRPk(i), then the measuring value of kth frame is
HRRPk={ HRRPk(i) | i=1,2 ..., Rn}
The measuring value of preceding k frames is represented by
Zk={ HRRP1,HRRP2,…,HRRPk}
If state xk-1Corresponding preceding k-1 frames energy accumulation value is I (xk-1), then kth frame state xkEnergy accumulation value can represent For
I ( x k ) = HRRP k ( x k ) + m a x x k - 1 ∈ Γ ( x k ) ( I ( x k - 1 ) )
Wherein, Γ (xk) it is that can be transferred to state xkAll possible state xk-1Set；To the object function I after accumulation k frames (xk) thresholding is set, the state that thresholding is more than to energy accumulation functional value carries out flight path backtracking, the shape involved by trace-back process State can be obtained by following formula
Y k ( x k ) = arg m a x x k - 1 ∈ Γ ( x k ) ( I ( x k - 1 ) )
Wherein, Yk(xk) to recall function, for recording the dbjective state that each stage makes energy accumulated value reach maximum, target Status switch is the estimate of target range track.
3. according to the method described in claim 1, it is characterised in that：The second step is specifically included：
Recall for each apart from track Y (k), respectively according to the vertex of a cone and cone two kinds of different distances of bottom scattering point at any time Between changing rule fit two groups of parameters：|oa|′、φ′0, β ', θ ', ω ', v ', and | ob | ", r "0、φ″0、β″、θ″、 ω ", v ", the error of fitting for defining two groups of parameters is respectively：
err 1 = Σ q = 1 k ( | o a | ′ ( sinβ ′ sinθ ′ sin ( ω ′ t k + φ 0 ′ ) ) + v ′ t k - Y k ( q ) ) k
Wherein
Compare error of fitting err1And err2If, err1≤err2, then it is assumed that this is formed apart from track by summit scattering point, can extract Go out fine motion characteristic parameter | oa | ', φ '0、β′、θ′、ω′、v′；It is on the contrary, then it is assumed that this is dissipated apart from track by the slip for boring baseplane Exit point is formed, and can extract out fine motion characteristic parameter | ob | ", r "0、φ″0、β″、θ″、ω″、v″；
On this basis, it is anti-by setting up information before fine motion characteristic parameter extraction result and target detection between tracking process Feedback, updates dbjective state transfer function；
If the vertex of a cone scattering point fine motion characteristic parameter extracted during kth frame is | oa | 'k、φ′0k、β′k、θ′k、ω′k、v′k；Extract To cone bottom scattering point fine motion characteristic parameter be | ob | "k、r″0k、φ″0k、β″k、θ″k、ω″k、v″k；Then for xk+1=i, according to Fine motion characteristic parameter extraction result can determine state transfer collection Γ (xk+1) be
Wherein
Δr1=| oa | ' (sin β ' sin θs ' sin (ω ' tk+1+φ′0))+v′tk+1-Yk(k)
fR() is searching scope adaptive Tuning function, the size of detection range door during being tracked before control detection；Function fR (), when error of fitting is larger, illustrates to believe next frame distance using fine motion characteristic parameter error of fitting during kth frame as independent variable The precision for ceasing prediction is relatively low, and detection range door should suitably increase；When error of fitting is larger, illustrate pre- to next frame range information The precision of survey is higher, can suitably reduce detection range door to reduce algorithm operation quantity；Therefore, fR() should be increasing function, fR(·) It is defined as
fR(errk)=rmin+λ·errk
Wherein rminFor detection range door minimum value.
4. according to the method described in claim 1, it is characterised in that：3rd step is specifically included：
Two detection thresholds are set：Low detection threshold T α1With high detection thresholding T α2；When energy accumulation value is more than high detection thresholding T α2When, judgement target is present, and fitting obtains target fine motion characteristic parameter；When energy accumulation value is less than high detection thresholding T α2When, root Judge that target whether there is according to fine motion feature extraction precision.
5. method according to claim 4, it is characterised in that：3rd step is specifically included：
As kth frame energy accumulation value I (xk) it is more than T α1When, backtracking obtain target scattering point apart from track estimate Yk, respectively Two groups of fine motion characteristic parameters apart from track form are fitted according to vertex of a cone scattering point and cone bottom scattering point, compare two kinds of fitting mistakes Difference, determines scattering point attribute, so as to obtain the fine motion feature extraction result of scattering point during kth frame, and error of fitting is designated as errk, on this basis, shifted and collected using fine motion characteristic parameter extraction result more new state, energy is carried out to the frame data of kth+1 Accumulate and recall target range track, by curve matching and application condition, the fine motion feature of scattering point is carried when obtaining the frame of kth+1 Result is taken, and error of fitting is designated as errk+1；If error of fitting very little, and it is enough to the prediction of the next frame location information of target Accurately, then it is assumed that fine motion feature extraction precision has reached requirement, while can determine that target is present, it is not necessary to carry out energy again and tire out Product；Otherwise it is assumed that cumlative energy is not enough, it is necessary to continue to carry out k+1 frame data energy accumulation and repeat the above steps；
Fine motion feature extraction precision judgment variables are defined as
Wherein TeIt is error of fitting thresholding,For what is obtained according to the fine motion feature extraction prediction of result of scattering center during kth frame The frame scattering center range information of kth+1, Yk+1To recall obtained target range track after the frame energy accumulation of kth+1；When energy is tired Product value is more than T α2Or fine motion feature extraction precision judgment variables be 1 when, think target exist, realize the fine motion to target Characteristic parameter extraction.
6. method according to claim 2, it is characterised in that：Take θ=π/6, β=π/6, ω=8 π rad/s, r0=1m, | Oa |=3m, | ob |=0.3m, v=12m/s.
7. method according to claim 3, it is characterised in that：Take rmin=20, λ=15.
8. method according to claim 4, it is characterised in that：Take T α1=10, T α2=20, Te=5.
CN201610041784.1A 2016-01-21 2016-01-21 A kind of body of revolution fine motion feature awareness extracting method Active CN106990398B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610041784.1A CN106990398B (en) 2016-01-21 2016-01-21 A kind of body of revolution fine motion feature awareness extracting method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610041784.1A CN106990398B (en) 2016-01-21 2016-01-21 A kind of body of revolution fine motion feature awareness extracting method

Publications (2)

Publication Number Publication Date
CN106990398A true CN106990398A (en) 2017-07-28
CN106990398B CN106990398B (en) 2019-10-15

Family

ID=59413521

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610041784.1A Active CN106990398B (en) 2016-01-21 2016-01-21 A kind of body of revolution fine motion feature awareness extracting method

Country Status (1)

Country Link
CN (1) CN106990398B (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108983190A (en) * 2018-07-19 2018-12-11 中国科学院国家空间科学中心 The method for obtaining rotating object difference scattering center fine motion track based on interferometric phase
CN108983189A (en) * 2018-07-19 2018-12-11 中国科学院国家空间科学中心 A kind of two-dimensional micromotion track estimation method of Vibration Targets
CN109683148A (en) * 2017-10-19 2019-04-26 深圳市新益技术有限公司 Target scattering characteristics test method
CN111443335A (en) * 2020-03-17 2020-07-24 中山大学 Method, system and device for estimating target micro-motion parameters of broadband radar and storage medium
WO2020154962A1 (en) * 2019-01-30 2020-08-06 深圳市大疆创新科技有限公司 Target credibility determination method, target recognition method and system, vehicle, and storage medium
CN113391274A (en) * 2021-06-15 2021-09-14 电子科技大学 Low-altitude target optimal search method for airborne phased array radar
CN114879157A (en) * 2022-05-06 2022-08-09 电子科技大学 High-value target scattering center parameter estimation method based on energy trajectory extraction

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5903217A (en) * 1997-10-21 1999-05-11 Microwave Sensors, Inc. Micro motion sensor
JP2008070227A (en) * 2006-09-14 2008-03-27 Toshiba Corp Target acquisition device
CN102043148A (en) * 2011-01-10 2011-05-04 西安电子科技大学 Ground moving target classification method based on general matched filtering
CN102426360A (en) * 2011-09-01 2012-04-25 西安电子科技大学 Two-dimensional ISRA imaging method of object with micro rotation in air
CN103344947A (en) * 2013-06-04 2013-10-09 四川大学 Micro-motion target characteristic extraction method based on micro-Doppler effect
CN104007430A (en) * 2014-05-29 2014-08-27 西安电子科技大学 Precession target micro-Doppler extracting method based on instant frequency modulation rate estimation

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5903217A (en) * 1997-10-21 1999-05-11 Microwave Sensors, Inc. Micro motion sensor
JP2008070227A (en) * 2006-09-14 2008-03-27 Toshiba Corp Target acquisition device
CN102043148A (en) * 2011-01-10 2011-05-04 西安电子科技大学 Ground moving target classification method based on general matched filtering
CN102426360A (en) * 2011-09-01 2012-04-25 西安电子科技大学 Two-dimensional ISRA imaging method of object with micro rotation in air
CN103344947A (en) * 2013-06-04 2013-10-09 四川大学 Micro-motion target characteristic extraction method based on micro-Doppler effect
CN104007430A (en) * 2014-05-29 2014-08-27 西安电子科技大学 Precession target micro-Doppler extracting method based on instant frequency modulation rate estimation

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
AI XIAOFENG 等: ""FEATURE EXTRACTION OF WOBBLING ROTATIONAL SYMMETRY TARGETS"", 《IGARSS 2015》 *
邹小海 等: ""基于PF-TBD的双基地微多普勒提取"", 《信号处理》 *

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109683148A (en) * 2017-10-19 2019-04-26 深圳市新益技术有限公司 Target scattering characteristics test method
CN108983190A (en) * 2018-07-19 2018-12-11 中国科学院国家空间科学中心 The method for obtaining rotating object difference scattering center fine motion track based on interferometric phase
CN108983189A (en) * 2018-07-19 2018-12-11 中国科学院国家空间科学中心 A kind of two-dimensional micromotion track estimation method of Vibration Targets
CN108983190B (en) * 2018-07-19 2020-07-03 中国科学院国家空间科学中心 Method for acquiring different scattering center micro-motion tracks of rotating target based on interference phase
CN108983189B (en) * 2018-07-19 2020-07-03 中国科学院国家空间科学中心 Two-dimensional micro-motion track estimation method for vibration target
WO2020154962A1 (en) * 2019-01-30 2020-08-06 深圳市大疆创新科技有限公司 Target credibility determination method, target recognition method and system, vehicle, and storage medium
CN111443335A (en) * 2020-03-17 2020-07-24 中山大学 Method, system and device for estimating target micro-motion parameters of broadband radar and storage medium
CN111443335B (en) * 2020-03-17 2022-03-25 中山大学 Method, system and device for estimating target micro-motion parameters of broadband radar and storage medium
CN113391274A (en) * 2021-06-15 2021-09-14 电子科技大学 Low-altitude target optimal search method for airborne phased array radar
CN114879157A (en) * 2022-05-06 2022-08-09 电子科技大学 High-value target scattering center parameter estimation method based on energy trajectory extraction

Also Published As

Publication number Publication date
CN106990398B (en) 2019-10-15

Similar Documents

Publication Publication Date Title
CN106990398A (en) A kind of body of revolution fine motion feature awareness extracting method
CN104297748B (en) One kind is based on tracking before the enhanced Radar Targets'Detection in track
CN102395192B (en) Method and device for locating wireless sensor terminal
CN104021289B (en) Non-Gaussian unsteady-state noise modeling method
CN103235297B (en) Space nutation target parameter estimation method based on broadband radar observation
KR101628154B1 (en) Multiple target tracking method using received signal strengths
CN107066806B (en) Data Association and device
CN106443622A (en) Distributed target tracking method based on improved joint probability data association
CN104502907A (en) Stable ground moving/static target tracking method for airborne radar
CN103729637A (en) Extended target probability hypothesis density filtering method based on cubature Kalman filtering
CN105246153A (en) High-density rapid collection method for indoor fingerprint positioning database
CN105093198A (en) Flight path fusion method for networking detection of distributed external radiation source radars
CN104715154B (en) Core K average Data Associations based on KMDL criterion criterions
CN104331623A (en) Self-adaptive target tracking information filtering algorithm of maneuvering strategies
CN107436427A (en) Space Target Motion Trajectory and radiation signal correlating method
CN106028446A (en) Indoor parking lot location method
CN107356922A (en) Tracking before a kind of more repetition Radar Targets'Detections for expanding table auxiliary law ambiguity solution
CN104881561A (en) Hough transform-based track-before-detect method of multidimensional parameters
CN107886111A (en) A kind of modified part mean decomposition method
Yang et al. Aircraft tracking based on fully conventional network and Kalman filter
CN104407345B (en) Improved dynamic planning method for weak target detection tracking
CN106772299A (en) A kind of PD radar weak target Dynamic Programming detection methods based on distance matching
CN108470189A (en) Multiple domain radiation source information fusion method based on antithesis similarity model
CN107247257A (en) Tracking before a kind of distributed multi-sensor detection approximate based on likelihood function
CN105353353B (en) Multiple search particle probabilities assume the multi-object tracking method of density filtering

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant