CN105259545B - Distributed true and false bullet recognition methods based on hard decision - Google Patents
Distributed true and false bullet recognition methods based on hard decision Download PDFInfo
- Publication number
- CN105259545B CN105259545B CN201510230515.5A CN201510230515A CN105259545B CN 105259545 B CN105259545 B CN 105259545B CN 201510230515 A CN201510230515 A CN 201510230515A CN 105259545 B CN105259545 B CN 105259545B
- Authority
- CN
- China
- Prior art keywords
- mrow
- msubsup
- msub
- msup
- mfrac
- 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.)
- Expired - Fee Related
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
Abstract
The invention discloses a kind of distributed true and false bullet recognition methods based on hard decision, belong to Radar Technology field.According to the modeling analysis of the bullet target kinetic characteristic to midcourse, establish the state-space model based on distributed treatment, true and false bullet fine motion difference is characterized with the parameter matrix in state-space model, bullet identification problem is converted into binary hypothesis test problem, each local processor observes data input LOUD detectors by local, export local hard decision result 0/1, each local decisions result is admitted to fusion center, global decision result is provided using K/G fusion criterions, so as to complete the differentiation of true and false bullet.This method need not estimate unknown state matrix, employ distributed processing structure, while the data transfer and loss of communications of radar system is reduced, ensure that higher detection probability, have the effect that data processing time is short, recognition accuracy is high.
Description
Technical field
The invention belongs to Radar Technology field, is a kind of distributed method for being used to identify true and false bullet, based on LOUD
(Locally Optimal Unknown Direction) detector and hard decision detection algorithm.
Background technology
Ballistic missile with its far firing range, power is big, precision is high and survival ability is strong the advantages that " being killed as modern war
One of hand mace " weapon.With the development of ballistic missile technology, ballistic missile defense system is arisen at the historic moment.True and false bullet target is known
It is not the key and difficult point in ballistic missile anti-missile system, has obtained extensive concern both domestic and external in recent years.
The trajectory of ballistic missile is divided into motors in boost phase penetration, stage casing (free segment) and reentry stage.In boost phase defense ballistic missile, need
System of defense is wanted to carry out preposition deployment, this is relatively difficult to achieve；Ballistic missile is defendd in reentry stage, leaves the reaction of system of defense for
Time is extremely short, and it is relatively low to intercept chance of success；And it is relative under, free flight phasel account for the 80%~90% of the whole trajectory time with
On, time length can be intercepted and defend area big, be preferred identification and the interception stage of system of defense.It is true and false in midcourse
Bullet target centroid does the difference very little of track motion, it is difficult to for identifying.Decoy warhead have the profile similar to true bullet and
Surfacing, thus the Electromagnetic Scattering Characteristics of decoy warhead, kinetic characteristic etc. are many-sided close with true bullet, which increase true and false bullet
Head target identification and the difficulty intercepted.
Bullet will carry out gesture stability in trajectory free segment, and the stability flown during being maintained at atmospheric reentry improves
Bullet accuracy at target.Spin is the conventional gesture stability mode of bullet free segment, if decoy warhead does not take gesture stability, can be gone out
The now random motion mode such as rolling.Caused Horizonal Disturbing makes bullet class target produce precession, i.e. mesh during decoy warhead release
Target spin axis is around space orientation symmetry axis rotation.From dynamic analysis, the speed of target precession and nutational angle it is big
It is small to be determined by the quality distribution of target.Thus, the difference of decoy warhead and the distribution of true warhead mass will result directly in precession ginseng
Several difference.Difference of the true and false bullet in fine motion feature identifies that problem provides theoretical foundation for true and false bullet.
The transmitter and receiver for splitting antenna MIMO radar system is separated, and distance meets far field bar between bay
Part so that the signal of each array element transmitting is independent, space diversity gain is derived from, from multiple angle object observings, so as to carry
Take the information that target is trickleer.MIMO radar launches separate and orthogonal signal, passes through matched filter in receiver end
Each road waveform signal is isolated, so as to introduce more observation passages and the free degree, compared with traditional phased-array radar greatly
Ground improves the overall performance of radar.Splitting antenna MIMO radar has excellent target detection capabilities and anti-stealthy, anti-destruction
Characteristic, available for detecting remote Weak target and smaller difference.Bullet target is observed using MIMO radar, Neng Goujian
The fine accurately difference in fine motion feature is measured, is the feasible method for solving true and false bullet identification problem.
State-space model includes state transition equation and observational equation, describes internal system change to system output
Influence：
xk+1=Fk+1xk+Gk+1uk+wk+1 (1)
zk=Hxk+ek (2)
Wherein xkIt is the state vector of k moment bullet targets, including the information such as position, speed, ukBe the k moment system it is defeated
Incoming vector, system mode noise wk+1Gaussian random vector is modeled as, meets distributionFk+1It is state transfer
Matrix, Gk+1It is system input vector ukCoefficient matrix；zkIt is k moment MIMO radar observation vectors, ekIt is the systematic perspective at k moment
Survey noise, it is assumed that its Gaussian distributed, i.e.,H is observing matrix.
From the above-mentioned analysis to true and false bullet motion feature, the difference of true and false bullet fine motion feature shows that precession is joined
Among number etc., that is, Fk+1With Gk+1Among.Here it is G in state transition model by problem reductionk+1Identical and Fk+1Different
Situation.Thus true and false bullet identification problem is modeled as binary hypothesis test problem：Midcourse true and false bullet with same fortune
Dynamic state flight, and its state-transition matrix keeps constant, i.e. F before the separation of true and false bulletk+1=F0, wherein F0Represent just
The state-transition matrix at moment beginning；In the true and false bullet separation of n moment, because true bullet has attitude control system, its state turns
Move matrix Fk+1It is still F to keep constant0, corresponding to H0Assuming that；And decoy warhead is due to lacking such control device, or quality and quality
Distribution will not be identical with true bullet, and its corresponding parameter matrix changes, Fk+1=Fc≠F0, corresponding to H1Assuming that：
If F0And FcAll it is known, likelihood ratio test can be carried out using Newman-Pearson came detector；It is but actual
F in situationcIt is matrix of unknown parameters, this, which allows for (3), turns into composite hypothesis check problem.
Somewhat like with Newman-Pearson came detector, local optimum unknown direction (LOUD) detector is in false-alarm probability PFA
Make detection probability maximum while≤α (α is a normal number), its detection criteria is when parameter vector change direction is unknown, is incited somebody to action
Detection probability corresponding to all possible change direction is averaging, and is obtained the average detected probability of maximum, is a kind of local optimum
Detector, it can be used for solving the problems, such as true and false bullet target identification.The corresponding composite hypothesis check problem as shown in (3), obtain
It is as follows to LOUD detector forms：
Wherein Fk+1(i, j), i, j=1,2 ..., L ' isElement, p (zk+1|zk；Fk+1) it is in zkIt is known
When zk+1Conditional probability density function, ηLOUDIt is detection threshold, is determined by false-alarm probability.
The content of the invention
Background technology uses centralized processing structure, and the observation data of each reception antenna of MIMO radar are all sent into fusion
Center complex processing, transmitted data amount is larger, cost is higher and easily causes data transmission loss.The present invention is for background skill
In art the defects of centralized processing, a kind of distributed true and false bullet recognition methods of MIMO radar is proposed, can not only avoid concentrating
Mass data is transmitted in formula processing, while can ensure higher detection identification probability.Here relational language is described below first：
Distributions spatial model：Distributions spatial model is to be based on each local processor in distributed frame
Establish.In the present invention, the reception array element of MIMO radar is divided into G sub-block (local processor), each sub-block is also
Antenna arrangement is split, similar to centralized processing, each sub-block carries out Combined Treatment to local observation signal, uses
To represent the local observation vector of each sub-block, distributed observational equation form is as follows：
Formula (1) forms the state-space model for describing each sub-block system with (5).
Hard decision Distributed Detection：In Distributed Detection structure, each local processor makees hard decision detection, obtains this
Ground court verdict 0/1 (0 or 1), then these comprehensive local court verdicts are made the final overall situation according to certain fusion criterion and sentenced
Certainly 0/1.
Fusion criterion：In distributed detection system, each local processor by court verdict be sent to fusion center it
Afterwards, fusion criterion determines fusion center carrys out each local decisions information of aggregation process with what rule.From the reason of data fusion
By analysis, the selection of fusion criterion must is fulfilled for " monotone increasing " principle.In the Distributed Detection application of reality, conventional fusion
Criterion has AND, OR and K/G criterion.AND criterions are when all local decisions are H1Global decision is H1, otherwise judgement is H0；OR
Criterion is when all local decisions are H0Global decision is H0, otherwise judgement is H1；K/G criterions are to work as G local decisions result
In, at least K judgement is H1Global decision is H1, otherwise judgement is H0。
Kalman filtering：One kind utilizes linear system state equation, data is observed by system input and output, to system shape
State carries out the algorithm of optimal estimation.Because observation data include the noise in system and the influence of interference, so can also regard as
It is filtering.State estimation is the important component of Kalman filtering.The state space based on determined by formula (1) and (5)
Model, define the correlated variables of kalman filter method：
Cramér-Rao lower bound：Cramer-Rao Bound, abridge CRB.For Parameter Estimation Problem, CRB is unbiased estimator
Estimate variance lower bound, i.e., can not possibly try to achieve estimate variance be less than the lower bound unbiased estimator.
The present invention is by analyzing motion feature of the true and false bullet target in midcourse, it is established that based on distributed treatment
State-space model, true and false bullet identification problem is changed into composite hypothesis check problem, each sub-block is utilized respectively LOUD inspections
Survey device and make hard decision detection, local decisions result 0/1 is sent to fusion center, finally makes the overall situation using K/G fusion criterions
Judgement, completes the identification of true and false bullet.
Distributed true and false bullet recognition methods step based on hard decision is as follows：
MIMO radar is received array element and is divided into G by step 1 receives sub-block, and the reception antenna in each sub-block is combined
Observation, obtain k moment part observation vectors
Step 2 is calculated as follows the parameter in the shown state-space model based on distributed treatment：
xk+1=Fk+1xk+Gk+1uk+wk+1
Step 2-1 obtains state-noise w according to engineering experience and MIMO radar estimation performancekAverage and covariance matrix
Σw；
Step 2-2 tries to achieve system input vector u according to bullet target motion analysisk；
Step 2-3 is pre-processed to obtain F by system0With Gk；
Step 2-4 is according to observation vectorDetermine observing matrix H；
Step 2-5 calculates local observation noiseAverage and covariance matrix
Step 2-5-1 part observation noisesIt is modeled as white Gaussian noise；
Step 2-5-2 is setAverage is 0；
Step 2-5-3 is setIt is relevant with receiving sub-block g, and it is unrelated with observation moment k, and its diagonal entry is by state
Estimate variance CRB is calculated；
Step 2-6 is according to bullet target initial state vector x0, MIMO radar transmitted waveform and transmitting, reception antenna position
Put, calculate state estimation variance CRB to be asked in step 2-5-3；
Step 3 calculates local LOUD detectors,
Step 3-1 gives Kalman filtering initial value x0|0With
Step 3-2 calculates the single-step iteration result of Kalman filtering process；
Step 3-2-1xk+1|k=F0xk|k+Gk+1uk；
Step 3-2-2
Step 3-2-3
Step 3-2-4
If step 3-3 k ＜ n-2, make k=k+1, repeat step 3-2；If k=n-2, stop iteration；
X is obtained after step 3-4 iteration stoppingsn-1|n-1With
Step 3-5 design conditions probability density functions
Step 3-5-1 makes Fn=F0；
Step 3-5-2xn|n-1=Fnxn-1|n-1+Gnun-1；
Step 3-5-3
Step 3-5-4 is obtainedConditional mean
Step 3-5-5 is obtainedConditional covariance matrix
Step 3-5-6 byTry to achieve its probability density function：
The sum term of step 3-6 detector molecular moieties
Step 3-6-1 makes i=1, j=1；
Step 3-6-2 makes Fn=F0+ Δ F δ (i, j), wherein δ (i, j) are except the element of the i-th row jth row is 1 other yuan
Element is 0 L' dimension square formations；
Step 3-6-3 is obtained using step 3-5-2 to 3-5-6
Step 3-6-4 makes Fn=F0-ΔFδ(i,j)；
Step 3-6-5 is obtained using step 3-5-2 to 3-5-6
Step 3-6-6 defines the sum term for obtaining detector molecule part with local derviation
If step 3-6-7 j ＜ L', make j=j+1, repeat step 3-6-2 to 3-6-6；If j=L' and i ＜ L', make i=i+
1, make j=1, repeat step 3-6-2 to 3-6-6；If j=L' and i=L', into next step；
Step 3-7 calculates the LOUD test statistics of local processor,
Step 4 tries to achieve false-alarm probability PfaDetection threshold under=α
Step 5 by the test statistics of local processor compared with thresholding, ifThen it is determined as vacation
Bullet, i.e. local decisions result ugFor 1, on the contrary ugFor 0；
Step 6 recycling step 3~5 obtain the court verdict u of G local processorg, g=1 ..., G；
Step 7 fusion center is collected into each local decisions result, and global decision is carried out using K/G criterions,
If step 8 global decision uf=1, then it is determined as decoy warhead；If uf=0, it is determined as true bullet.
Brief description of the drawings
Fig. 1 is antenna distribution and the partition schematic diagram for splitting antenna MIMO radar；
Fig. 2 is the LOUD based on hard decision, preferable likelihood ratio (ILR), the detection probability of mismatch likelihood ratio (MLR) detector
PDWith Sudden Changing Rate Δ fsThe curve map of change.
Embodiment
Distributed Detection method specific steps based on hard decision and LOUD detectors are as described in the content of the invention, and 10000
The simulation result that secondary Monte Carlo experiment obtains is as shown in Fig. 2 the parameter wherein used is as follows：
Consider the motion conditions of bullet target multidimensional parameter, bullet state vector includes its position, speed and fine motion position
Move, set its initial value here as x0=[100,200,1000,100,50,10,0.05,0.06,0.08]T.During system incipient stability
State matrix is F0, true bullet state matrix keeps constant after true and false bullet separation, and the state matrix of decoy warhead becomes Fc.Root
Analyzed according to bullet target motion modeling, F can be obtainedk+1With Gk+1Concrete form.
Wherein observation interval T=0.1s, fine motion vector are [ωx,ωy,ωz]=[0.6,0.5,0.6245],
Fine motion rotates angular frequencys=2 π fs, wherein initial rotation frequency fs=1Hz, with Δ fsRepresent that decoy warhead rotates frequency
Rate fsVariable quantity, also just characterize the difference between true and false bullet jogging motion.Due to only considering F herek+1Change, institute
To make Gk+1In ωsKeep constant.Make system input vector (vector acceleration) uk=[0,0,0,0,2,9.6,0,0,0]T, shape
State noise covariance matrix is Σw=0.2I9。
Observing matrix H=[I6,06×3], that is, position and the velocity amplitude of bullet target are observed, without believing including its micromotion
Breath.Using one 4 × 8MIMO radar systems, it launches as shown in Figure 1 with reception antenna distribution.The low-pass equivalent of transmission signal is
The Gaussian monopulse signal of frequency expansionPulse width Tt=0.1, Δ f must take
Be wide enough so that mutually orthogonal between transmission signal, take Δ f=0.7/T heret.Transmitting antenna relative coordinate in Fig. 1 is former
The angle of point is respectively { 80 °, 150 °, 225 °, 325 ° }, and distance is respectively { 6000,8000,6000,10000 } m；Reception antenna
Three sub-blocks are divided into according to aerial position, the angle of reception antenna is { 30 °, 44 °, 45 ° } in sub-block 1, distance for 10000,
12000,12000 } m, the angle of reception antenna is { 135 °, 150 °, 160 ° } in sub-block 2, and distance is all 6000m, is connect in sub-block 3
The angle for receiving antenna is { 320 °, 325 ° }, and distance is { 10000,8000 } m.Each sub-block pair can be obtained according to above- mentioned information
Target location and the CRB of velocity estimation variance, obtain local observation noise covariance matrixThe observation used in emulation to
AmountProduced by the state-space model based on distributed treatment.
In these three detection methods shown in Fig. 2, each sub-block is drawn local hard with LOUD, ILR, MLR detector respectively
Court verdict, last fusion center obtain global decision using K/G criterions, and G=3, takes K=2 here.Set the void of each sub-block
Alarm probability is identical and global false-alarm probability is 0.001.ILR detectors use preferable Fc, and MLR detectors are using mismatch
mis_Fc.As can be seen from Figure 2 Δ fsWhen changing between 0 to 3, the detection probability of LOUD detectors is slightly below ILR detections
Device, but significantly greater than MLR detectors, and work as Δ fsWhen getting 3, the detection probability of LOUD detectors can reach 1, this explanation
The validity and feasibility identified based on hard decision and LOUD the Distributed Detection algorithm examined to true and false bullet.
Claims (1)
1. a kind of distributed true and false bullet recognition methods based on hard decision, it is characterised in that comprise the following steps：
MIMO radar is received array element and is divided into G by step 1 receives sub-block, and the reception antenna in each sub-block carries out joint sight
Survey, obtain k moment part observation vectors
Step 2 is calculated as follows the parameter in the shown state-space model based on distributed treatment：
xk+1=Fk+1xk+Gk+1uk+wk+1
<mrow>
<msubsup>
<mi>z</mi>
<mi>k</mi>
<mi>g</mi>
</msubsup>
<mo>=</mo>
<msub>
<mi>Hx</mi>
<mi>k</mi>
</msub>
<mo>+</mo>
<msubsup>
<mi>e</mi>
<mi>k</mi>
<mi>g</mi>
</msubsup>
<mo>,</mo>
<mi>g</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>G</mi>
</mrow>
Step 2-1 obtains state-noise w according to engineering experience and MIMO radar estimation performancekAverage and covariance matrix Σw；
Step 2-2 tries to achieve system input vector u according to bullet target motion analysisk；
Step 2-3 is pre-processed to obtain F by system0With Gk；
Step 2-4 is according to observation vectorDetermine observing matrix H；
Step 2-5 calculates local observation noiseAverage and covariance matrix
Step 2-5-1 part observation noisesIt is modeled as white Gaussian noise；
Step 2-5-2 is setAverage is 0；
Step 2-5-3 is setIt is relevant with receiving sub-block g, and it is unrelated with observation moment k, and its diagonal entry is by state estimation side
Poor CRB is calculated；
Step 2-6 is according to bullet target initial state vector x0, MIMO radar transmitted waveform and transmitting, reception antenna position, meter
Calculate state estimation variance CRB to be asked in step 2-5-3；
Step 3 calculates local LOUD detectors,
Wherein n represents the judgement moment of LOUD detectors, and L ' is state-transition matrix Fk+1Dimension；H0And HlTrue bullet is corresponded to respectively
Two head, decoy warhead hypothesis；
Step 3-1 gives Kalman filtering initial value x0|0With
Step 3-2 calculates the single-step iteration result of Kalman filtering process；
Step 3-2-1 xk+1|k=F0xk|k+Gk+1uk；
Step 3-2-2
Step 3-2-3
Step 3-2-4
If step 3-3 k ＜ n-2, make k=k+1, repeat step 3-2；If k=n-2, stop iteration；
X is obtained after step 3-4 iteration stoppingsn-1|n-1With
Step 3-5 design conditions probability density functions
Step 3-5-1 makes Fn=F0；
Step 3-5-2 xn|n-1=Fnxn-1|n-1+Gnun-1；
Step 3-5-3
Step 3-5-4 is obtainedConditional mean
Step 3-5-5 is obtainedConditional covariance matrix
Step 3-5-6 byTry to achieve its probability density function：
<mrow>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>z</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mo>|</mo>
<msubsup>
<mi>Z</mi>
<mrow>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mo>;</mo>
<msub>
<mi>F</mi>
<mi>n</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>exp</mi>
<mo>{</mo>
<mo>-</mo>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<msup>
<mrow>
<mo>(</mo>
<msubsup>
<mi>z</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>z</mi>
<mrow>
<mi>n</mi>
<mo>|</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mo>)</mo>
</mrow>
<mi>T</mi>
</msup>
<msup>
<mrow>
<mo>(</mo>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>n</mi>
<mo>|</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mo>)</mo>
</mrow>
<mrow>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<msubsup>
<mi>z</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>z</mi>
<mrow>
<mi>n</mi>
<mo>|</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mo>)</mo>
</mrow>
<mo>}</mo>
</mrow>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mi>&pi;</mi>
<mo>)</mo>
</mrow>
<mfrac>
<mi>L</mi>
<mn>2</mn>
</mfrac>
</msup>
<mo>|</mo>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>n</mi>
<mo>|</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<msup>
<mo>|</mo>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
</msup>
</mrow>
</mfrac>
</mrow>
The sum term of step 3-6 detector molecular moieties
Step 3-6-1 makes i=1, j=1；
Step 3-6-2 makes Fn=F0+ Δ F δ (i, j), wherein δ (i, j) are except the element of the i-th row jth row is equal for 1 other elements
Square formation is tieed up for 0 L'；
Step 3-6-3 is obtained using step 3-5-2 to 3-5-6
Step 3-6-4 makes Fn=F0-ΔFδ(i,j)；
Step 3-6-5 is obtained using step 3-5-2 to 3-5-6
Step 3-6-6 defines the sum term for obtaining detector molecule part with local derviation
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<msubsup>
<mi>&Pi;</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mo>=</mo>
<mfrac>
<mrow>
<msup>
<mo>&part;</mo>
<mn>2</mn>
</msup>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>z</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mo>|</mo>
<msubsup>
<mi>Z</mi>
<mrow>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mo>;</mo>
<msub>
<mi>F</mi>
<mi>n</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mo>&part;</mo>
<msubsup>
<mi>F</mi>
<mi>n</mi>
<mn>2</mn>
</msubsup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<msub>
<mo>|</mo>
<mrow>
<msub>
<mi>F</mi>
<mi>n</mi>
</msub>
<mo>=</mo>
<msub>
<mi>F</mi>
<mn>0</mn>
</msub>
</mrow>
</msub>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>P</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>z</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mo>|</mo>
<msubsup>
<mi>Z</mi>
<mrow>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mo>;</mo>
<msub>
<mi>F</mi>
<mn>0</mn>
</msub>
<mo>+</mo>
<mi>&Delta;</mi>
<mi>F</mi>
<mi>&delta;</mi>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>)</mo>
<mo>+</mo>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>z</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mo>|</mo>
<msubsup>
<mi>Z</mi>
<mrow>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mo>;</mo>
<msub>
<mi>F</mi>
<mn>0</mn>
</msub>
<mo>-</mo>
<mi>&Delta;</mi>
<mi>F</mi>
<mi>&delta;</mi>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>)</mo>
<mo>-</mo>
<mn>2</mn>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>z</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mo>|</mo>
<msubsup>
<mi>Z</mi>
<mrow>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mo>;</mo>
<msub>
<mi>F</mi>
<mn>0</mn>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<msup>
<mrow>
<mo>(</mo>
<mi>&Delta;</mi>
<mi>F</mi>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mfrac>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
If step 3-6-7 j ＜ L', make j=j+1, repeat step 3-6-2 to 3-6-6；If j=L' and i ＜ L', make i=i+1, order
J=1, repeat step 3-6-2 to 3-6-6；If j=L' and i=L', into next step；
Step 3-7 calculates the LOUD test statistics of local processor,
<mrow>
<msubsup>
<mi>&Gamma;</mi>
<mrow>
<mi>L</mi>
<mi>O</mi>
<mi>U</mi>
<mi>D</mi>
</mrow>
<mi>g</mi>
</msubsup>
<mrow>
<mo>(</mo>
<msubsup>
<mi>z</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msup>
<mi>L</mi>
<mo>&prime;</mo>
</msup>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msup>
<mi>L</mi>
<mo>&prime;</mo>
</msup>
</munderover>
<msubsup>
<mi>&Pi;</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>z</mi>
<mi>n</mi>
<mi>g</mi>
</msubsup>
<mo>|</mo>
<msubsup>
<mi>Z</mi>
<mrow>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mi>g</mi>
</msubsup>
<mo>;</mo>
<msub>
<mi>F</mi>
<mn>0</mn>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
Step 4 tries to achieve false-alarm probability PfaDetection threshold under=α
Step 5 by the test statistics of local processor compared with thresholding, ifThen it is determined as decoy warhead,
That is local decisions result ugFor 1, on the contrary ugFor 0；
Step 6 recycling step 3~5 obtain the court verdict u of G local processorg, g=1 ..., G；
Step 7 fusion center is collected into each local decisions result, and global decision is carried out using K/G criterions,
If step 8 global decision uf=1, then it is determined as decoy warhead；If uf=0, it is determined as true bullet.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510230515.5A CN105259545B (en) | 2015-05-07 | 2015-05-07 | Distributed true and false bullet recognition methods based on hard decision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510230515.5A CN105259545B (en) | 2015-05-07 | 2015-05-07 | Distributed true and false bullet recognition methods based on hard decision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105259545A CN105259545A (en) | 2016-01-20 |
| CN105259545B true CN105259545B (en) | 2018-04-06 |
Family
ID=55099306
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510230515.5A Expired - Fee Related CN105259545B (en) | 2015-05-07 | 2015-05-07 | Distributed true and false bullet recognition methods based on hard decision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105259545B (en) |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9562968B2 (en) * | 2012-10-22 | 2017-02-07 | Saab-Sensis Corporation | Sensor system and method for determining target location using sparsity-based processing |
| CN103487801B (en) * | 2013-09-02 | 2015-07-22 | 电子科技大学 | Method of radar for recognizing true and false warheads |
| CN104267393B (en) * | 2014-09-02 | 2016-08-24 | 电子科技大学 | A kind of MIMO radar true and false bullet Sequential Detection |
-
2015
- 2015-05-07 CN CN201510230515.5A patent/CN105259545B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN105259545A (en) | 2016-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103487801B (en) | Method of radar for recognizing true and false warheads | |
| CN101832738A (en) | Remote air-to-air missile multi-platform cooperative guidance system and realization method thereof | |
| CN104459662B (en) | Micro-motion target characteristic extraction method and system based on wavelet multi-scale analysis | |
| CN103728598B (en) | The method of track spoofing interference is suppressed with the active radar and passive radar net of other place configure | |
| CN106680806A (en) | Multi-radar plot fusion method | |
| CN103728599B (en) | The method of false targets interference is suppressed with the active radar and passive radar net of other place configure | |
| CN107167785A (en) | A kind of sane big array MIMO radar target transmitting-receiving angle combined estimation method | |
| CN107255924A (en) | Method for extracting guidance information of strapdown seeker through volume Kalman filtering based on dimension expansion model | |
| CN109933847A (en) | A kind of improved boost phase trajectory algorithm for estimating | |
| CN108107423B (en) | Distributed networked radar target detection and tracking integrated processing method | |
| Ma et al. | Target tracking system for multi-sensor data fusion | |
| CN104680002B (en) | A kind of distributed fusion method theoretical based on random set | |
| CN107220409A (en) | A kind of radar network cloth station method that people's group hunting is immunized based on particle mechanism | |
| CN103675808B (en) | A kind of monopulse radar seeker can not differentiate multi-target detection method | |
| CN109598946A (en) | A kind of multilane speed-measuring method based on radar system | |
| CN103247057A (en) | Road target multi-hypothesis tracking algorithm under target-echo-road network data association | |
| CN113702940B (en) | Spatial cluster target resolution method based on multi-element characteristic information hierarchical fusion and application | |
| CN103809161A (en) | Radar net distance deception + SOJ (stand-off jamming) resisting method | |
| CN110363219A (en) | Midcourse target fine motion form recognition methods based on convolutional neural networks | |
| CN105259545B (en) | Distributed true and false bullet recognition methods based on hard decision | |
| CN107133555A (en) | Recognize 8 word movement locus mesh calibration methods | |
| CN108535704A (en) | A kind of signal Pre-sorting method based on self-adaption two-dimensional cluster | |
| CN106291585A (en) | Terahertz high-resolution imaging method under low sampling number | |
| CN104931945B (en) | Distributed true and false bullet recognition methods based on common recognition | |
| Li-hua et al. | Precession period extraction of ballistic missile based on radar measurement |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180406 Termination date: 20210507 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |