CN106772299B - One kind is based on apart from matched PD radar weak target Dynamic Programming detection method - Google Patents
One kind is based on apart from matched PD radar weak target Dynamic Programming detection method Download PDFInfo
- Publication number
- CN106772299B CN106772299B CN201611094528.5A CN201611094528A CN106772299B CN 106772299 B CN106772299 B CN 106772299B CN 201611094528 A CN201611094528 A CN 201611094528A CN 106772299 B CN106772299 B CN 106772299B
- Authority
- CN
- China
- Prior art keywords
- radar
- moment
- target
- measurement
- matching
- 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
- 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)
- 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
The invention discloses one kind based on apart from matched PD radar weak target Dynamic Programming detection method, belongs to radar dim target detection tracking field.PD radar generallys use high medium-PRF, and fuzzy so as to cause distance by radar measurement, especially under the low signal-to-noise ratio environment including clutter and noise, the test problems of PD radar weak target are just become more complicated.The present invention is based on the thoughts tracked before detection, propose a kind of based on apart from matched PD radar weak target Dynamic Programming detection method.Firstly, by restoring the temporal correlation that objective fuzzy measures apart from matching treatment；Then, utilize the characteristic of Dynamic Programming batch processing, realize the non-inherent accumulation to weak target information, avoid PD radar range finding fuzzy problem and weak target low signal-to-noise ratio characteristic and caused by target missing inspection, PD radar can be effectively improved to the detection probability of weak target, Project Realization is easy, and has stronger practicability and application value.
Description
Technical field
The invention belongs to track field before radar dim target detection, it is right under PD radar range finding hazy condition to be suitable for solving
The integration detection problem of weak target.
Background technique
The detecting and tracking of radar weak target is a difficult and most important problem, for winning the following high-tech
War has decisive significance.With the development of stealth technology, the appearance of weak target proposes huge challenge to radar performance,
Since the echo-signal signal-to-noise ratio of weak target is very low, target can not achieve with traditional short time phase-coherent accumulation detection method
Reliable detection；On the other hand, in order to unambiguously measure target velocity, PD radar generallys use high, medium-PRF
Mode, thus cause the range measurement of target be it is fuzzy, this becomes PD radar more to the test problems of weak target
It is difficult.
In order to improve target signal to noise ratio, it is necessary to long-time non-inherent accumulation is carried out to obtain more signal energies, wherein
Most typical method is dynamic programming method.Dynamic Programming is a kind of equivalent exhaustive search calculation based on multistage process decision
Method, a multi-dimensional optimization is searched for problem by multistage classification processing and is divided into several one-dimensional Optimizing Search problems by it, most
The accumulation of multiframe measurement information is realized in the planning process in good path, so as to improve target signal to noise ratio, realizes radar weak target
Reliable detection；But PD radar measurement is when there is ambiguity, target measure in time-space relationship be it is interrupted, can not adopt
Long-time non-inherent accumulation is carried out to the backward energy from the same target with dynamic programming method.
In document, [Zhang Wei, hole enable and saying, a kind of improvement DPA Faint target detection algorithm for HPRF radar of the such as Yang Xiaobo
[J] modern radar .2011,33 (5)] in, author proposes a kind of improved dynamic programming method to solve the above problems,
Basic step is as follows:
1) current time is measured using the predicted state of last moment and carries out jump judgement, judge measure whether cross-module is pasted
Section；
2) according to court verdict, different Dynamic Programming search strategies is selected, realizes the accumulation of target energy；
3) by Threshold detection, the backtracking of objective fuzzy track is obtained.
The above method takes different search strategies by jumping judgement, to realize the non-phase along objective fuzzy track
Ginseng accumulation realizes the detection to objective fuzzy track, but it has the following disadvantages: to improve target signal to noise ratio
1) it needs to carry out jump judgement according to the predicted state of last moment, it is therefore desirable to which the priori of dbjective state model is believed
Breath, when that can not establish dbjective state model or model foundation obtains inaccurate, will lead to judge incorrectly；
2) by being only able to detect the fuzzy track of target after algorithm process, the inspection of radar target can not fundamentally be solved
Survey tracking problem；
If 3) to obtain the true track of target, it is also necessary to carry out ambiguity solution processing, the solution of algorithm by remainder theorem
Fuzzy performance is limited by remainder theorem application conditions.
Summary of the invention
1. technical problems to be solved
The purpose of the present invention is to propose to one kind based on apart from matched PD radar weak target Dynamic Programming detection method, from
Fundamentally solve the problems, such as that dynamic programming method can not be applied to PD radar ambiguity.
2. technical solution
The present invention provides one kind to be based on apart from matched PD radar weak target Dynamic Programming detection method, using technology
Protocol step is as follows:
Step 1: initialization system parameter:
PfaFor initial threshold false-alarm；
RmaxFor maximum radar range；
M is the type of PD radar pulse repetition frequency；
M=1,2 ..., M is the serial number of pulse recurrence frequency；
FmFor m-th of pulse recurrence frequency；
RumFor pulse recurrence frequency FmCorresponding maximum unam；
K is the scanning moment sum for handling data；
K=1,2 ..., K is the scanning moment serial number of data；
NkThe total number measured for the k moment；
N=1,2 ..., NkThe serial number measured for the k moment；
The fuzzy distance measured for n-th of the k moment；
The orientation measured for n-th of the k moment；
AnIt (k) is the echo amplitude of n-th of k moment measurement；
For n-th of measurement unit of k moment；
Z′kFor k moment distance matching measurement sequence；
JkFor the total number of k moment potential track；
J=1,2 ..., JkFor the serial number of k moment potential track；
SkIt (j) is the objective function of j-th of potential track of k moment；
IkIt (j) is the cost function of j-th of potential track of k moment；
H (j) is j-th of the target actual measurements sequence eventually detected；
Step 2: initial threshold processing
Scanning moment k, by the echo data of each range-azimuth unit respectively with an initial threshold η1Compared
Compared with to eliminate partial noise influence, the metric data sequence Z after obtaining PD radar initial detectingk, concrete measure are as follows:
(1) in scanning moment k, by each measurement unit zn(k) respectively with an initial threshold η1It is compared, thus
Eliminating partial noise influences；In order to retain target information to greatest extent, a larger initial false-alarm is selected, initial threshold is set
It is set to:
Wherein,For ZkIn in all measurement units amplitude average value:
(2) initial detecting is carried out to radar measurement data using initial threshold, the radar measurement sequence that obtains that treatedWherein zn(k) value is as follows:
In subsequent processing, the measurement unit that value is all 0 is skipped, in this manner it is possible to the influence of exclusive segment noise, from
And reduce the data volume of computer disposal；
Step 3: apart from matching treatment
PD radar is usually to obscure to measuring for target range, and the measurements of all distance dimensions are all compressed in the 1st and obscure
In section, target is caused to measure discontinuous on space-time, therefore, it is difficult to carry out non-inherent accumulation along targetpath；In order to solve
This problem, to radar measurement sequence ZkIt carries out apart from matching treatment, by distanceEquiprobability is matched to all fuzzy intervals,
Obtain the distance matching measurement sequence Z ' of radark, to restore the temporal correlation of metric data, concrete measure are as follows:
(1) each pulse recurrence frequency F is calculatedmFuzzy interval number Φm
Φm=Int (Rmax/Rum),
Wherein Int () indicates rounding operation, m=1,2 ..., M；
(2) the k moment passes through F for radarmN-th of the measurement z measuredn(k), by its distanceIt is matched to lmA mould
Section is pasted, a matching distance can be obtained
(3) corresponding orientation is addedEcho amplitude An(k), fuzzy interval valued numbers lmEtc. information, construct one distance matching
It measures
(4) fuzzy interval valued numbers l is enabledmIn lm=1,2 ..., ΦmIn continuous value, obtain one group distance matching measure
(5) it enables and measures number n in n=1,2 ..., NkIn continuous value, obtain all measurements of k moment distance matching measure
Sequence Z 'k
In this way, by apart from matching treatment, in Z 'kThe middle temporal correlation for having restored metric data；
(6)Z′kIt is Φ for a line numbermNkSequence, be simplified shown as:
Wherein, i indicates distance matching measurement sequence Z 'kIn measurement line number, i=1,2 ..., ΦmNk, z 'i(k) it indicates
Z′kIn the i-th row distance matching measure vector；
By above-mentioned processing, range-to-go matching measurement sequence Z ' is obtainedk, since treatment process is to all moulds of radar
Paste the equiprobability matching of section row；Therefore necessarily there is distance matching to measure to fall in target realistic blur section；As it can be seen that after processing
Measurement sequence Z 'kIn certainly exist the actual measurements of target, to restore the temporal correlation of metric data.
Step 4: Dynamic Programming processing
All distance matching measurement sequences that K scanning moment is handled constitute a search set Z '1:K=
[Z′1,Z′2,…,Z′K]；After above-mentioned steps 3 are apart from matching treatment, Z 'kIn certainly exist the actual distance at target k moment
It measures, has thus restored the temporal correlation of metric data；And then it is found that the true track of target is inevitable and be uniquely present in
Search for set Z '1:KIn；In the case where not considering error in measurement, the target measurement of different moments different pulse repetition is answered
This is accumulated on its true track, and random noise measurement does not have above-mentioned accumulation then；It therefore, can be by Dynamic Programming side
Method scans for all potential tracks in search set, and recursion seeks objective function Sk(j) and respective value function Ik(j),
Realize that echo amplitude improves radar detedtion probability along the non-inherent accumulation of targetpath so as to improve target signal to noise ratio；
Step 5: Threshold detection
According to the cost function I of accumulationK(j), detection threshold G is set, accumulation is eliminated and is worth lower potential track, retain
Cost function is more than the potential track of detection threshold, is deposited into candidate track set D:
D=j | [IK(j) >=G, j=1,2 ..., Jk],
Wherein, j | [IK(j) >=G, j=1,2 ..., Jk] indicate to extract and all meet IK(j) >=G condition potential track
Serial number j；
Step 6: Objective extraction
For each potential track serial number j ∈ D in candidate track set D, since the k-th moment, backward is carried out
Backtracking, obtains target in k moment corresponding measurement serial number ik:
ik=Sk(j)；
From Z 'KMiddle extraction serial number ikCorresponding target measures
Work as k=1 ..., k ..., when K, obtains the corresponding measurement sequence of target
To realize the extraction to target real trace.
3. beneficial effect
It is compared with background technique, beneficial effects of the present invention explanation:
(1) one kind for using of the present invention is based on apart from matched PD radar weak target Dynamic Programming detection method, can be with
Fundamentally solve the problems, such as dim target detection of the PD radar in the case where measuring ambiguity: first by restoring apart from matching treatment
Then the temporal correlation that objective fuzzy measures is realized to the non-inherent accumulation of weak target information using Dynamic Programming, is avoided
Target missing inspection caused by PD radar range finding fuzzy problem and weak target low signal-to-noise ratio characteristic, can effectively improve PD radar to micro-
The detection probability of weak signal target, Project Realization are easy, and have stronger practicability and application value.
(2) compared with the conventional method, the present invention does not need to carry out jump judgement, but by apart from matching treatment extracted amount
Correlation is surveyed, to realize the recursion accumulation of Dynamic Programming, processing method is simpler flexibly；On the other hand, existing method
It is the blurring trajectorie for detecting target, it is also necessary to which the true rail of target can just be obtained by carrying out ambiguity solution processing by remainder theorem
Mark, the matching of the invention that target actual measurements are realized apart from matching treatment process, therefore final detection result output is
The real trace of target.
Detailed description of the invention
Attached drawing 1 is overall flow figure of the invention；
Attached drawing 2 is initial threshold radar measurement figure before and after the processing in the embodiment of the present invention；
Attached drawing 3 be in the embodiment of the present invention after matching treatment radar measurement figure；
Attached drawing 4 is that track plot is searched in Dynamic Programming in the embodiment of the present invention；
Attached drawing 5 is the cost function accumulation figure that track is respectively searched in the embodiment of the present invention；
Attached drawing 6 is the schematic diagram for carrying out Threshold detection in the embodiment of the present invention to the cost function of accumulation；
Attached drawing 7 is the target trajectory figure finally extracted in the embodiment of the present invention；
Specific embodiment
The present invention is mainly adopted and is experimentally verified, and all steps, conclusion are all verified just on Matlab2010a
Really.
Embodiment condition: it is emulated for a general single goal moving scene.Assuming that target is done in x-y plane
Linear uniform motion, initial position are (51km, 22km), and speed is (- 220m/s, 230m/s), and target is non-fluctuating target,
Its radar cross section 0.1m2.Sensor parameters setting are as follows: two coordinate PD radars are in coordinate origin position, scan period 1s,
Maximum operating range is 70km, transmission power Pt=5kw, carrier wavelength lambda=0.1m, radar antenna gain 34dB, radar away from
High Resolution is 150m, and angular resolution is 0.5 °.Radar is taken turns to operate using the transmitting pulse of 3 difference PRF, and each PRF points
Not Wei 16000Hz, 111400Hz and 119400Hz, corresponding maximum unam is respectively Ru1=9375m, Ru2=13158m,
Ru3=7732m.For target signal to noise ratio be 7dB the case where, emulate 25 scanning the moment data verified, wherein initially
False-alarm is set as Pfa=10-2。
It is based below in conjunction with 1 pair of one kind of the invention of Figure of description apart from matched PD radar weak target Dynamic Programming
Detection method is described in detail.Referring to Figure of description 1, process flow of the invention divides following steps:
Step 1: parameter initialization is carried out according to simulated conditions:
Pfa=0.01, Rmax=70km, M=3；
F1=16000Hz, F2=11400Hz, F3=19400Hz；
Ru1=9375m, Ru2=13158m, Ru3=7732m, K=25；
NkFor the total number of measurement；JkFor the total number of potential track；
An(k)=[], Z 'k=[]；
Sk(j)=[], Ik(j)=[], H (j)=[], j=1,2 ..., Jk；
Step 2: initial threshold processing
(1) in scanning moment k, by each measurement unit zn(k) respectively with an initial threshold η1It is compared, thus
Eliminating partial noise influences；In order to retain target information to greatest extent, a larger initial false-alarm is selected, initial threshold is set
It is set to:
Wherein,For ZkIn in all measurement units amplitude average value:
(2) initial detecting is carried out to radar measurement data using initial threshold, the radar measurement sequence that obtains that treatedWherein zn(k) value is as follows:
In subsequent processing, the measurement unit that value is all 0 is skipped, in this manner it is possible to the influence of exclusive segment noise, from
And reduce the data volume of computer disposal.
Fig. 2 show initial threshold radar measurement comparison diagram before and after the processing, with Fig. 2 (a) comparison as can be seen that in Fig. 2 (b)
Data volume reduces very much, but the data for being derived from target are not filtered out；As it can be seen that the main purpose of initial threshold processing is
By setting a higher false-alarm thresholding, filters out a part of unrelated noise and measure, to reduce data volume, improve algorithm
Efficiency.
Step 3: apart from matching treatment
Figure it is seen that PD radar is usually to obscure to the measurement of target range, the measurements of all distance dimensions are all pressed
Contracting causes target to measure discontinuous on space-time, therefore, it is difficult to carry out non-coherent along targetpath in the 1st fuzzy interval
Accumulation；In order to solve this problem, to radar measurement sequence ZkIt carries out apart from matching treatment, by distanceEquiprobability is matched to
All fuzzy intervals obtain the distance matching measurement sequence Z ' of radark, to restore the temporal correlation of metric data, specifically
Measure are as follows:
(1) each pulse recurrence frequency F is calculatedmFuzzy interval number Φm
Φm=Int (Rmax/Rum),
Wherein Int () indicates rounding operation, m=1,2 ..., M；
(2) the k moment passes through F for radarmN-th of the measurement z measuredn(k), by its distanceIt is matched to lmA mould
Section is pasted, a matching distance can be obtained
(3) corresponding orientation is addedEcho amplitude An(k), fuzzy interval valued numbers lmEtc. information, construct one distance matching
It measures
(4) fuzzy interval valued numbers l is enabledmIn lm=1,2 ..., ΦmIn continuous value, obtain one group distance matching measure
(5) it enables and measures number n in n=1,2 ..., NkIn continuous value, obtain all measurements of k moment distance matching measure
Sequence Z 'k
In this way, by apart from matching treatment, in Z 'kThe middle temporal correlation for having restored metric data；
(6)Z′kIt is Φ for a line numbermNkSequence, be simplified shown as:
Wherein, i indicates distance matching measurement sequence Z 'kIn measurement line number, i=1,2 ..., ΦmNk, z 'i(k) it indicates
Z′kIn the i-th row distance matching measure vector；Due to having been carried out apart from all fuzzy intervals of the matching treatment process to radar etc.
Probability match, therefore necessarily there is distance matching to measure and fall in target realistic blur section；As it can be seen that treated measurement sequence
Z′kIn certainly exist the actual measurements of target, to restore the temporal correlation of metric data；
Fig. 3 show the radar measurement figure after matching treatment, has gone out the reality of target trajectory in figure with white wire collimation mark
Region；By radar in Fig. 3 and Fig. 2 it is original measure comparison as can be seen that the distance range measured after apart from matching treatment by
One not fuzzy distance be matched to radar maximum measure distance range；If not examining error in measurement influence, after apart from matching treatment,
Mixed and disorderly blur measurement is all accumulated near target real trace in Fig. 3 in Fig. 2, this explanation can be effective apart from matching treatment
Extract the temporal and spatial correlations information that target measures.
Step 4: Dynamic Programming processing
Although from figure 3, it can be seen that the correlation of radar measurement is restored after apart from matching treatment；But by
It is relatively low in target signal, if showing target trajectory not over white wire collimation mark, it is difficult to distinguish target from background
Come.In order to solve this problem, next target is measured using dynamic programming method and carries out non-inherent accumulation, so as to improve mesh
Signal-to-noise ratio is marked, target trajectory detected from noise background, the specific method is as follows:
All distance matching measurement sequences that K scanning moment is handled constitute a search set Z '1:K=
[Z′1,Z′2,…,Z′K]；After above-mentioned steps 3 are apart from matching treatment, Z 'kIn certainly exist the actual distance at target k moment
It measures, has thus restored the temporal correlation of metric data；And then it is found that the true track of target is inevitable and be uniquely present in
Search for set Z '1:KIn；In the case where not considering error in measurement, the measurement of different moments different pulse repetition should be accumulated
Tire out on true track.
All potential tracks in search set are scanned for by dynamic programming method, searches in being gathered and owns
Potential track, as shown in Figure 4；Then, recursion seeks the cost function I of each potential trackk(j), realize echo amplitude along mesh
The non-inherent accumulation of track is marked, so as to improve target signal to noise ratio, improves radar detedtion probability, the cost function of each potential track is such as
Shown in Fig. 5.
Step 5: Threshold detection
As shown in fig. 6, according to the cost function I of accumulationK(j), setting normalization detection threshold G=45, eliminates accumulation valence
It is worth lower potential track, reserve value function is more than the potential track of detection threshold, it is deposited into candidate track set D:
D=j | [IK(j) >=G, j=1,2 ..., Jk],
From fig. 6, it can be seen that in embodiments of the present invention, only one element D={ j=30 } in candidate track set D；
Step 6: Objective extraction
For the potential track j=30 in candidate track set D, since the k-th moment, backward backtracking is carried out, is obtained
Target is in k moment corresponding measurement serial number ik:
ik=Sk(30)；
From Z 'KIt is middle to extract corresponding matching measurement
Work as k=1 ..., k ..., when K, obtains the corresponding measurement sequence of target
To realize the extraction to target real trace, as shown in Figure 5；From figure 5 it can be seen that method of the invention can effectively disappear
Except noise Interference Detection to target real trace.
It can be seen that the present invention overcomes weak targets under PD radar range finding hazy condition to measure from embodiment verification result
The problem of data can not be accumulated effectively；By restoring the temporal correlation that objective fuzzy measures apart from matching treatment, then utilize
Dynamic Programming is realized to the non-inherent accumulation of weak target information, and PD radar range finding fuzzy problem and the low letter of weak target are avoided
It makes an uproar target missing inspection caused by ratio characteristic, PD radar can be effectively improved to the detection probability of weak target；Under current simulated conditions,
The present invention has good detection performance to weak target, while can directly extract the real trace of target, no longer needs to carry out
Ambiguity solution processing, Project Realization are easy, and have stronger practicability and application value.
Claims (2)
1. one kind is based on apart from matched PD radar weak target Dynamic Programming detection method, which is characterized in that including following step
It is rapid:
Step 1: initialization system parameter:
PfaFor initial threshold false-alarm；
RmaxFor maximum radar range；
M is the type of PD radar pulse repetition frequency；
M=1,2 ..., M is the serial number of pulse recurrence frequency；
FmFor m-th of pulse recurrence frequency；
RumFor pulse recurrence frequency FmCorresponding maximum unam；
K is the scanning moment sum for handling data；
K=1,2 ..., K is the scanning moment serial number of data；
NkThe total number measured for the k moment；
N=1,2 ..., NkThe serial number measured for the k moment；
The fuzzy distance measured for n-th of the k moment；
The orientation measured for n-th of the k moment；
AnIt (k) is the echo amplitude of n-th of k moment measurement；
For n-th of measurement unit of k moment；
Z′kFor k moment distance matching measurement sequence；
JkFor the total number of k moment potential track；
J=1,2 ..., JkFor the serial number of k moment potential track；
SkIt (j) is the objective function of j-th of potential track of k moment；
IkIt (j) is the cost function of j-th of potential track of k moment；
H (j) is j-th of the target actual measurements sequence eventually detected；
Step 2: initial threshold processing
Scanning moment k, by the echo data of each range-azimuth unit respectively with an initial threshold η1It is compared, from
And partial noise influence is eliminated, the metric data sequence Z after obtaining PD radar initial detectingk；
Step 3: apart from matching treatment
Measurement of the PD radar to target rangeUsually fuzzy, cause target to measure discontinuous on space-time therefore difficult
To carry out non-inherent accumulation along targetpath；In order to solve this problem, to radar measurement sequence ZkIt carries out at matching
Reason, willEquiprobability is matched to all fuzzy intervals, obtains the distance matching measurement sequence Z ' of radark, to restore to measure
The temporal correlation of data；It is as follows apart from matching treatment concrete methods of realizing:
S11: each pulse recurrence frequency F is calculatedmFuzzy interval number Φm
Φm=Int (Rmax/Rum),
Wherein Int () indicates rounding operation, m=1,2 ..., M；
The S12:k moment passes through F for radarmN-th of the measurement z measuredn(k), by its distanceIt is matched to lmIt is a fuzzy
Section, wherein lm=1,2 ..., Φm, a matching distance can be obtained
S13: corresponding orientation is addedEcho amplitude An(k), fuzzy interval valued numbers lmEtc. information, in lmIn a fuzzy interval
It constructs a distance matching and measures vector
S14: fuzzy interval valued numbers l is enabledmIn lm=1,2 ..., ΦmIn continuous value, obtain one group distance matching measurement sequence
S15: enabling and measure serial number n in n=1,2 ..., NkIn continuous value, obtain all measurements of k moment distance matching measure sequence
Arrange Z 'k:
S16: according to the definition of S14 and S15, Z 'kIt is Φ for a line numbermNkSequence, be simplified shown as:
Wherein, z'i(k) Z ' is indicatedkIn the matching of the i-th row distance measure vector, in vector form and step S13Unanimously, i
=1,2 ..., ΦmNk；
Due to having carried out equiprobability matching apart from all fuzzy intervals of the matching treatment process to radar, necessarily there is distance
It is fallen in target realistic blur section with measurement；As it can be seen that treated measurement sequence Z 'kIn certainly exist the substantial amount of target
It surveys, to restore the temporal correlation of metric data；
Step 4: Dynamic Programming processing
All distance matching measurement sequences that K scanning moment is obtained constitute a search set Z '1:K=[Z '1,Z′2,…,
Z′K], all potential tracks in search set are scanned for by dynamic programming method, recursion seeks objective function Sk(j)
With respective value function Ik(j), realize echo amplitude along the non-inherent accumulation of targetpath；
Step 5: Threshold detection
In k=K, according to the cost function I of accumulationK(j), detection threshold G is set, accumulation is eliminated and is worth lower potential track,
Reserve value function is more than the potential track of detection threshold, is deposited into candidate track set D:
D=j | [IK(j) >=G, j=1,2 ..., Jk],
Wherein, j | [IK(j) >=G, j=1,2 ..., Jk] indicate to extract and all meet IK(j) sequence of >=G condition potential track
Number j；
Step 6: Objective extraction
For each potential track serial number j ∈ D in candidate track set D, since the k-th moment, backward backtracking is carried out,
Target is obtained in k moment corresponding measurement serial number ik:
ik=Sk(j)；
From Z 'KMiddle extraction serial number ikCorresponding matching measuresWork as k=1 ..., k ..., when K, obtains the measurement sequence of targetTo realize the extraction to target real trace.
2. one kind according to claim 1 is based on apart from matched PD radar weak target Dynamic Programming detection method,
It is characterized in that initial threshold processing method described in step 2:
S21: in scanning moment k, by each measurement unit zn(k) respectively with an initial threshold η1It is compared, to eliminate
Partial noise influences；In order to retain target information to greatest extent, a larger initial false-alarm, initial threshold setting are selected
Are as follows:
Wherein, natural logrithm is sought in ln () expression,For k moment all measurement unit zn(k) average value of amplitude in, n=1,
2,…,Nk:
Wherein, summation operation is sought in ∑ [] expression；
S22: carrying out Threshold detection to all measurement units using initial threshold, the radar measurement sequence that obtains that treatedWherein zn(k) value is as follows:
In subsequent processing, the measurement unit that value is all 0 is skipped；In this manner it is possible to the influence of exclusive segment noise, to drop
The data volume of low computer disposal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611094528.5A CN106772299B (en) | 2016-12-01 | 2016-12-01 | One kind is based on apart from matched PD radar weak target Dynamic Programming detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611094528.5A CN106772299B (en) | 2016-12-01 | 2016-12-01 | One kind is based on apart from matched PD radar weak target Dynamic Programming detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106772299A CN106772299A (en) | 2017-05-31 |
| CN106772299B true CN106772299B (en) | 2019-04-26 |
Family
ID=58883540
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611094528.5A Active CN106772299B (en) | 2016-12-01 | 2016-12-01 | One kind is based on apart from matched PD radar weak target Dynamic Programming detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106772299B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107561506B (en) * | 2017-07-27 | 2020-08-14 | 中国船舶重工集团公司第七二四研究所 | Multistage false alarm probability joint recursion target detection method |
| CN109471091A (en) * | 2018-11-07 | 2019-03-15 | 中国人民解放军国防科技大学 | Method for simultaneously resolving ambiguity, detecting and tracking intermediate repetition frequency PD radar |
| CN109901154B (en) * | 2019-03-29 | 2023-03-03 | 中国人民解放军海军航空大学 | Self-adaptive adjustment method based on recursive RTHT-TBD |
| CN110726988B (en) * | 2019-10-30 | 2021-08-27 | 中国人民解放军海军航空大学 | Distance and speed fuzzy mutual solution method for detecting hypersonic target by PD radar |
| CN113640752B (en) * | 2021-07-13 | 2023-10-20 | 北京理工大学 | Waveform design method based on inter-pulse phase frequency spectrum double agility |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101825707A (en) * | 2010-03-31 | 2010-09-08 | 北京航空航天大学 | Monopulse angular measurement method based on Keystone transformation and coherent integration |
| CN102323575A (en) * | 2011-07-16 | 2012-01-18 | 西安电子科技大学 | Range migration correction method for pulse Doppler (PD) radar in feeble signal detection process |
-
2016
- 2016-12-01 CN CN201611094528.5A patent/CN106772299B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101825707A (en) * | 2010-03-31 | 2010-09-08 | 北京航空航天大学 | Monopulse angular measurement method based on Keystone transformation and coherent integration |
| CN102323575A (en) * | 2011-07-16 | 2012-01-18 | 西安电子科技大学 | Range migration correction method for pulse Doppler (PD) radar in feeble signal detection process |
Non-Patent Citations (2)
| Title |
|---|
| Joint range ambiguity resolving and multiple maneuvering targets tracking in clutter via MMPHDF-DA;TAN ShunCheng et al.;《SCIENCE CHINA》;20140831;正文第1-12页 |
| 基于抛物线随机 Hough 变换的机载脉冲多普勒雷达机动弱目标检测前跟踪方法;于洪波 等;《兵工学报》;20151031;第36卷(第10期);第1924-1932页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106772299A (en) | 2017-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106772299B (en) | One kind is based on apart from matched PD radar weak target Dynamic Programming detection method | |
| US7132975B2 (en) | Apparatus and method for detecting moving objects | |
| CN107861123B (en) | Method for real-time tracking of multiple moving targets by through-wall radar in complex environment | |
| CN102298142B (en) | HT-STC-TBD (Hough Transform and Successive Target Cancellation Based Track-Before-Detect) detection method for multiple weak targets under radar-ranging fuzzy condition | |
| CN107942342B (en) | Data processing method, device, system and the storage medium of anemometry laser radar | |
| CN103439697B (en) | Target detection method based on dynamic programming | |
| CN104714225B (en) | Dynamic programming tracking-before-detection method based on generalized likelihood ratios | |
| CN106772352B (en) | It is a kind of that Weak target detecting method is extended based on the PD radar of Hough and particle filter | |
| CN107561518A (en) | Three-dimensional radar Plot coherence method based on two-dimentional sliding window local extremum | |
| CN105785340B (en) | The angle measurement and recognition methods that monopulse radar interferes target in main lobe and bait | |
| CN104166135A (en) | Method for processing original point trace condensation of broadband radar target | |
| CN103885057A (en) | Self-adaptation variable-sliding-window multi-target tracking method | |
| CN110058222B (en) | Double-layer particle filter tracking-before-detection method based on sensor selection | |
| CN106569193B (en) | The small targets detection in sea clutter method filtered based on anterior-posterior to income reference particle | |
| CN104459661B (en) | Method for detecting rapid artillery type dim target | |
| CN113221631B (en) | Sequence pulse anti-interference target detection method based on convolutional neural network | |
| CN107255814A (en) | A kind of radar target detection method based on LFMSK waveforms | |
| CN106468771A (en) | A kind of multi-target detection and tracking method under the high clutter conditions of low Observable | |
| CN107942324A (en) | Multi-frame joint Small object double check method based on Doppler's guiding | |
| CN109061625B (en) | Distance and speed matched moving target detection method | |
| CN108872977B (en) | Life body double-station cooperative detection method based on single-channel ultra-wideband radar | |
| CN110488273A (en) | A kind of vehicle tracking detection method and device based on radar | |
| CN208060701U (en) | A kind of multiple target categorizing system based on microwave | |
| CN102508271B (en) | Navigation satellite signal capturing method based on peak value location comparison | |
| JP5036392B2 (en) | Target detection device |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20190321 Address after: 264001 scientific research office, two Road 188, Zhifu District, Yantai, Shandong Applicant after: Naval Aviation University of PLA Address before: 264001 scientific research department, two Road 188, Yantai, Shandong Applicant before: Naval Aeronautical Engineering Institute PLA |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |