CN104714226A - Phase based dynamic planning before-detection tracking method - Google Patents

Abstract

The invention provides a phase based dynamic planning before-detection tracking method, belongs to the technical field of radar target detection tracking and particularly relates to the technical field of fluctuating target detection tracking under a complex Gaussian noise background. When radar echo data are received, phase information is utilized to calculate complex likelihood ratios, the complex likelihood ratios are accumulated to search all possible tracks and estimate a target state. Compared with an existing dynamic planning before-detection tracking algorithm, the difference of a target and noise can be well embodied by utilizing the phase information and the detection tracking performance of a fluctuating target can be improved by means of the phase based dynamic planning before-detection tracking method.

Description

Tracking before a kind of dynamic programming based on phase place detects

Technical field

The present invention relates to Radar Targets'Detection tracking technique, particularly rise and fall under multiple Gaussian Background weak target detections of radar tracking technique.

Background technology

Along with reaching its maturity of stealth technology, the radar cross section of the target such as aircraft, guided missile reduces one to two orders of magnitude, comprise helicopter simultaneously, a large amount of Military application of the Small object at a slow speed of the low-latitude flyings such as SUAV (small unmanned aerial vehicle), low-speed operations device, make the problem all facing dim targets detection difficulty in each fields such as airborne fire control, airborne early warning, ground information, shipborne radars.Therefore, the technical matters that radar is badly in need of solution is become to the detection and tracking of weak target.

Before dynamic programming detects, tracking is a kind of effective dim target detection tracking, and it carries out Combined Treatment to multiframe radar return data, is searched for all possible flight path by accumulation targetpath value function, thus the time of day of estimating target.Before existing dynamic programming detects, tracking targetpath value function has two kinds: (1) amplitude; (2) envelope log-likelihood ratio.These two kinds of targetpath value functions all do not utilize phase information, and not utilizing of phase information makes target detection tracking performance decline, and is unfavorable for the detecting and tracking of weak target.For improving the detecting and tracking performance to constant amplitude weak target, the people such as Davey propose tracking before the detection based on multiple likelihood ratio, see document " S. ~ J.Davey; M. ~ G.Rutten; and B. ~ Cheung; ``Using phase toimprove track-before-detect; " IEEE Trans.Aerosp.Electron.Syst., vol.48, no.1, pp.832-849, Jan.2012 ", result display phase information effectively can improve the detecting and tracking performance to constant amplitude target.

But the radar cross section of real goal can not with a simple constant.Even if usually for simple target, radar cross section is also the complicated function of the angle of sight, frequency, plan.Swerling 3 fluctuating target model is conventional fluctuating target model, and it characterizes radar cross section probability density function is 4 ° of χ ²the target of distribution is the approximate solution of a lot of little scatterer and a strong scatterer, and wherein the radar cross section of strong scatterer equals little scatterer radar cross section sum doubly.

Summary of the invention

The problem that the detecting and tracking probability that the object of the invention is to exist when following the tracks of Swerling 3 fluctuating target for tracking before existing dynamic programming detection is low, tracking before a kind of dynamic programming based on phase place of Curve guide impeller detects, utilizes phase information to improve and to rise and fall the detecting and tracking performance of weak target and the object of system stability to Swerling 3.

The invention provides a kind of dynamic programming based on phase place and detect front tracking, the method comprises:

Step 1: initializes system parameters comprises: datum plane size N _x× N _y, process frame number K, state transfer number q, utilizes the thresholding V that Monte Carlo simulation experimental calculation goes out _t, impact point spread function h (x _k), noise average power σ _n, target average power σ _t;

Step 2: targetpath value function assignment is carried out to each resolution element of each frame: the targetpath value function that each resolution element of each frame is corresponding is the multiple likelihood ratio that this resolution element is corresponding;

Step 3, by dynamic programming algorithm, all possible targetpath in front K frame to be searched for, and record the positional information of each targetpath at resolution element corresponding to every frame, targetpath value function corresponding for the resolution element belonging to same targetpath in front K frame is superposed, obtains the target discrimination value of this target at K frame;

Step 4, by the thresholding V of each for K frame target discrimination value and setting _tcompare, if higher than thresholding, then assert in the corresponding resolution element of this decision content and have target to exist; If all target discrimination values of K frame are all lower than thresholding, then announce that target does not exist;

If step 5 step 4 assertive goal exists, K frame location information before this target utilizing step 3 to record, recover the flight path of target;

Step 6: false track is deleted: contrasted by all targetpaths recovering in step 5 to obtain, if there are many flight paths to have same position at a certain frame, then using there is the flight path of maximum target decision content as real goal flight path in these flight paths, delete all the other flight paths;

Step 7, output targetpath.

The concrete steps of described step 2 are:

Step 2.1, calculate multiple likelihood ratio corresponding to each resolution element of each frame: to resolution element (i, j), suppose wherein to comprise target, then dbjective state is x _k=[i, v _x, j, v _y], wherein the position of target is (i, j), and the speed of target is (v _x, v _y), then the multiple likelihood ratio that this resolution element is corresponding is

$\frac{4}{{\mathrm{\σ}}_t^2}{\mathrm{\β}}^{-1}\exp \left(\frac{{\mathrm{\β}}^2}{2\mathrm{\α}}\right){\mathrm{\α}}^{-1.5}{M}_{-\mathrm{1.5,0}}\left(\frac{{\mathrm{\β}}^2}{\mathrm{\α}}\right)$

Wherein

$\mathrm{\β}=\frac{\left|h^H\left(x_k\right)Z_k\right|}{{\mathrm{\σ}}_n}$

Z _k(i, j) represents resolution element (i, j) correspondence

Measuring value; Z _k=[z _k(1,1) ..., z _k(1, N _y), z _k(2,1) ..., z _k(1, N _y)] ^trepresent the set of kth frame amount measured value; () ^trepresent transposition; $h\left(x_k\right)={[h^{\left(\mathrm{1,1}\right)}\left(x_k\right),...,h^{(1,N_y)}\left(x_k\right),h^{\left(\mathrm{2,1}\right)}\left(x_k\right),...,h^{(1,N_y)}\left(x_k\right)]}^T$ Represent impact point spread function, wherein h ^{(m, n)}(x _k); 1≤m≤N _x, 1≤n≤N _yrepresent that target energy is to the contribution of resolution element (m, n); N _xrepresent the line number of datum plane, N _yrepresent the columns of datum plane, h ^h(x _k) represent h (x _k) conjugate transpose, M _-1.5,0represent that parameter is the Whittaker function of-1.5,0.

Step 2.2, be the multiple likelihood ratio that this resolution element of calculating in step 2.1 is corresponding by targetpath value function assignment corresponding for individual for kth frame (i, j) resolution element.

The concrete steps of described step 3 are:

Step 3.1, to the 1st frame each Range resolution unit object decision content assignment: the target discrimination value that the 1st each Range resolution unit of frame is corresponding is the targetpath value function that this Range resolution unit calculates in step 2.3;

Step 3.2, to the 2nd frame each Range resolution unit object decision content assignment: assuming that in each resolution element of the 2nd frame exist a target, judge the region that each target of the 2nd frame may exist in the 1st frame, find out resolution element that in this region, target discrimination value is maximum and record the position of this unit, the target discrimination value of the corresponding resolution element of the 2nd frame be the targetpath value function that the maximum target decision content found out in the 1st frame is corresponding with the 2nd this resolution element of frame and;

Step 3.3, the employing method identical with step 3.2 calculate the target discrimination value of the 3rd frame to K frame.

In described step 3, the value of K is 3 ~ 20.

The concrete steps of described step 6 are:

Step 6.1, will through judgement all targetpaths in, the targetpath at K frame with maximum target decision content is judged to be real goal flight path;

Step 6.2, the targetpath each being passed through judgement and real goal flight path compare, if this flight path and real goal flight path have same position at certain frame, then this flight path are judged to be false track;

If step 6.3 still has the targetpath not through judging, then repeat step 6.1, step 6.2 until all targetpaths all pass through judgement.

Tracking before a kind of dynamic programming based on phase place of the present invention detects, utilizes phase information to calculate multiple likelihood ratio corresponding to each resolution element of each frame, chooses multiple likelihood ratio and carry out the accumulation of dynamic programming value as targetpath value function.Utilize phase information effectively can improve detecting and tracking performance to Swerling 3 fluctuating target, reduce undetected.

Accompanying drawing explanation

Fig. 1 is FB(flow block) of the present invention.

Fig. 2 is the present invention (N _x=30, N _x=30, q=9, K=5, σ _n=1, if m=i-1, i, i+1 be then h ^{(m, n)}(x _k)=exp (-(m-i) ², otherwise h/8) ^{(m, n)}(x _k)=0) (do not utilize phase place, N with traditional single frame detection _x=30, N _x=30, q=9, K=1, σ _n=1, if m=i-1, i, i+1 be then h ^{(m, n)}(x _k)=exp (-(m-i) ², otherwise h/8) ^{(m, n)}(x _k)=0) detection perform comparison diagram.Wherein be with the solid line of " o " to represent performance of the present invention, the solid line of band " " represents traditional single frame detection performance.

Fig. 3 is the present invention (N _x=30, N _x=30, q=9, K=5, σ _n=1, if m=i-1, i, i+1 be then h ^{(m, n)}(x _k)=exp (-(m-i) ², otherwise h/8) ^{(m, n)}(x _k)=0) plan that detecting front tracking (does not utilize phase place, N with conventional dynamic _x=30, N _x=30, q=9, K=5, σ _n=1, if m=i-1, i, i+1 be then h ^{(m, n)}(x _k)=exp (-(m-i) ², otherwise h/8) ^{(m, n)}(x _k)=0) detection perform comparison diagram.Wherein be with the solid line of " o " to represent performance of the present invention, the solid line of band " " represents that conventional dynamic planning detects front tracking detection perform.

Embodiment

The present invention chooses multiple likelihood ratio as targetpath value function, and choosing of this targetpath value function takes full advantage of target amplitude statistical property, noise statistics and target phase information.Amplitude target flight path value function three kinds of information that before conventional dynamic planning detects, tracking adopts all do not utilize, and envelope log-likelihood ratio make use of target amplitude statistical property and noise statistics, does not but utilize phase information.The present invention utilizes phase information can improve the detecting and tracking performance of the front tracking of dynamic programming detection further.

The present invention mainly adopts the method for l-G simulation test to verify, institute in steps, conclusion all on MATLAB R2012b checking correct.Concrete implementation step is as follows:

Step 1: initializes system parameters comprises: datum plane size 30 × 30 (i.e. N _x=N _y=30), process frame number K=5, state transfer number q=9, utilizes the thresholding V that Monte Carlo simulation experimental calculation goes out _t=[-0.0124,0.6802,1.2511,1.7538,1.9351,2.3633,1.9839,1.6518,1.4338,0.2654 ,-0.5999 ,-1.8316], impact point spread function h (x _k): if m=i-1, i, i+1 be then h ^{(m, n)}(x _k)=exp (-(m-i) ², otherwise h/8) ^{(m, n)}(x _k)=0, noise average power σ _n=1, target average power σ _t=[1.0000,1.2589,1.5849,1.9953,2.5119,3.1623,3.9811,5.0119,6.3096,7.9433,10.0000,12.5893] (corresponding signal to noise ratio (S/N ratio) is 0-11dB).

Step 2: targetpath value function assignment is carried out to each resolution element of each frame: the targetpath value function that each resolution element of each frame is corresponding is the multiple likelihood ratio that this resolution element is corresponding;

Step 2.1, calculate multiple likelihood ratio corresponding to each resolution element of each frame: to resolution element (i, j), (then dbjective state is x to suppose wherein to comprise target _k=[i, v _x, j, v _y], wherein the position of target is (i, j), and the speed of target is (v _x, v _y)), then the multiple likelihood ratio that this resolution element is corresponding is

$\frac{4}{{\mathrm{\σ}}_t^2}{\mathrm{\β}}^{-1}\exp \left(\frac{{\mathrm{\β}}^2}{2\mathrm{\α}}\right){\mathrm{\α}}^{-1.5}{M}_{-\mathrm{1.5,0}}\left(\frac{{\mathrm{\β}}^2}{\mathrm{\α}}\right)$

Wherein

$\mathrm{\α}=\frac{h^H\left(x_k\right)h\left(x_k\right)}{{\mathrm{\σ}}_n}+\frac{2}{{\mathrm{\σ}}_t}$

$\mathrm{\β}=\frac{\left|h^H\left(x_k\right)Z_k\right|}{{\mathrm{\σ}}_n}$

Z _k(i, j) represents the measuring value that resolution element (i, j) is corresponding; Z _k=[z _k(1,1) ..., z _k(1, N _y), z _k(2,1) ..., z _k(1, N _y)] ^trepresent the set of kth frame amount measured value; () ^trepresent transposition; $h\left(x_k\right)={[h^{\left(\mathrm{1,1}\right)}\left(x_k\right),...,h^{(1,N_y)}\left(x_k\right),h^{\left(\mathrm{2,1}\right)}\left(x_k\right),...,h^{(1,N_y)}\left(x_k\right)]}^T$ Represent impact point spread function, wherein h ^{(m, n)}(x _k); 1≤m≤N _x, 1≤n≤N _yrepresent that target energy is to the contribution of resolution element (m, n); N _xrepresent the line number of datum plane, N _yrepresent the columns of datum plane, h ^h(x _k) represent h (x _k) conjugate transpose, M _-1.5,0represent that parameter is the Whittaker function of-1.5,0.

Step 2.2, be the multiple likelihood ratio that this resolution element of calculating in step 2.1 is corresponding by targetpath value function assignment corresponding for individual for kth frame (i, j) resolution element.

Step 3, by dynamic programming algorithm, all possible targetpath in front K frame to be searched for, and record the positional information of each targetpath at resolution element corresponding to every frame, targetpath value function corresponding for the resolution element belonging to same targetpath in front K frame is superposed, obtains the target discrimination value of this target at K frame;

Step 3.1, to the 1st frame each Range resolution unit object decision content assignment: the target discrimination value that the 1st each Range resolution unit of frame is corresponding is the targetpath value function that this Range resolution unit calculates in step 2.3;

Step 3.2, to the 2nd frame each Range resolution unit object decision content assignment: assuming that in each resolution element of the 2nd frame exist a target, judge the region that each target of the 2nd frame may exist in the 1st frame, find out resolution element that in this region, target discrimination value is maximum and record the position of this unit, the target discrimination value of the corresponding resolution element of the 2nd frame be the targetpath value function that the maximum target decision content found out in the 1st frame is corresponding with the 2nd this resolution element of frame and; Concrete processing mode is as follows: the individual Range resolution unit of the 2nd frame (i, j), supposes wherein there is a target, then the region that this target may exist in former frame (the 1st frame) is (i-1-v _x, j-1-v _y), (i-1-v _x, j-v _y), (i-1-v _x, j+1-v _y), (i-v _x, j-1-v _y), (i-v _x, j-v _y), (i-v _x, j+1-v _y), (i+1-v _x, j-1-v _y), (i+1-v _x, j-v _y), (i+1-v _x, j+1-v _y);

Step 3.3, the employing method identical with step 3.2 calculate the target discrimination value of the 3rd frame to K frame.

Step 4, by the thresholding V of each for K frame target discrimination value and setting _tcompare, if higher than thresholding, then assert in the corresponding resolution element of this decision content and have target to exist; If all target discrimination values of K frame are all lower than thresholding, then announce that target does not exist;

If step 5 step 4 assertive goal exists, K frame location information before this target utilizing step 3 to record, recover the flight path of target;

Step 6: false track is deleted: contrasted by all targetpaths recovering in step 5 to obtain, if there are many flight paths to have same position at a certain frame, then using there is the flight path of maximum target decision content as real goal flight path in these flight paths, delete all the other flight paths;

Step 6.1, will through judgement all targetpaths in, the targetpath at K frame with maximum target decision content is judged to be real goal flight path;

Step 6.2, the targetpath each being passed through judgement and real goal flight path compare, if this flight path and real goal flight path have same position at certain frame, then this flight path are judged to be false track;

If step 6.3 still has the targetpath not through judging, then repeat step 6.1, step 6.2 until all targetpaths all pass through judgement.

Step 7, output targetpath.

Can find out that the method that the present invention proposes has greatly improved than the front tracking of conventional dynamic planning detection and the equal tool of traditional single frame detection method detection perform, effectively can carry out detecting and tracking to faint Swerling 3 fluctuating target.

Claims (5)

1. the dynamic programming based on phase place detects a front tracking, and the method comprises:

Step 1: initializes system parameters comprises: datum plane size N _x× N _y, process frame number K, state transfer number q, utilizes the thresholding V that Monte Carlo simulation experimental calculation goes out _t, impact point spread function h (x _k), noise average power σ _n, target average power σ _t;

Step 2: targetpath value function assignment is carried out to each resolution element of each frame: the targetpath value function that each resolution element of each frame is corresponding is the multiple likelihood ratio that this resolution element is corresponding;

Step 3, by dynamic programming algorithm, all possible targetpath in front K frame to be searched for, and record the positional information of each targetpath at resolution element corresponding to every frame, targetpath value function corresponding for the resolution element belonging to same targetpath in front K frame is superposed, obtains the target discrimination value of this target at K frame;

Step 4, by the thresholding V of each for K frame target discrimination value and setting _tcompare, if higher than thresholding, then assert in the corresponding resolution element of this decision content and have target to exist; If all target discrimination values of K frame are all lower than thresholding, then announce that target does not exist;

If step 5 step 4 assertive goal exists, K frame location information before this target utilizing step 3 to record, recover the flight path of target;

Step 6: false track is deleted: contrasted by all targetpaths recovering in step 5 to obtain, if there are many flight paths to have same position at a certain frame, then using there is the flight path of maximum target decision content as real goal flight path in these flight paths, delete all the other flight paths;

Step 7, output targetpath.

2. a kind of dynamic programming based on phase place detects front tracking as claimed in claim 1, it is characterized in that the concrete steps of described step 2 are:

Step 2.1, calculate multiple likelihood ratio corresponding to each resolution element of each frame: to resolution element (i, j), suppose wherein to comprise target, then dbjective state is x _k=[i, v _x, j, v _y], wherein the position of target is (i, j), and the speed of target is (v _x, v _y), then the multiple likelihood ratio that this resolution element is corresponding is

$\frac{4}{{\mathrm{\σ}}_t^2}{\mathrm{\β}}^{-1}\mathrm{esp}\left(\frac{{\mathrm{\β}}^2}{2\mathrm{\α}}\right){\mathrm{\α}}^{-1.5}{M}_{-\mathrm{1.5,0}}\left(\frac{{\mathrm{\β}}^2}{\mathrm{\α}}\right)$

Wherein

$\mathrm{\β}=\frac{\left|h^H\left(x_k\right)Z_k\right|}{{\mathrm{\σ}}_n}$

Z _k(i, j) represents the measuring value that resolution element (i, j) is corresponding; Z _k=[z _k(1,1) ..., z _k(1, N _y), z _k(2,1) ..., z _k(1, N _y)] ^trepresent the set of kth frame amount measured value; () ^trepresent transposition; $h\left(x_k\right)={[h^{\left(\mathrm{1,1}\right)}\left(x_k\right),...,h^{(1,N_y)}\left(x_k\right),h^{\left(\mathrm{2,1}\right)}\left(x_k\right),...,h^{\left({1,N}_y\right)}\left(x_k\right)]}^T$ Represent impact point spread function, wherein h ^{(m, n)}(x _k); 1≤m≤N _x, 1≤n≤N _yrepresent that target energy is to the contribution of resolution element (m, n); N _xrepresent the line number of datum plane, N _yrepresent the columns of datum plane, h ^h(x _k) represent h (x _k) conjugate transpose, M _-1.5,0represent that parameter is the Whittaker function of-1.5,0.

Step 2.2, be the multiple likelihood ratio that this resolution element of calculating in step 2.1 is corresponding by targetpath value function assignment corresponding for individual for kth frame (i, j) resolution element.

3. a kind of dynamic programming based on phase place detects front tracking as claimed in claim 1, it is characterized in that the concrete steps of described step 3 are:

Step 3.1, to the 1st frame each Range resolution unit object decision content assignment: the target discrimination value that the 1st each Range resolution unit of frame is corresponding is the targetpath value function that this Range resolution unit calculates in step 2.3;

Step 3.2, to the 2nd frame each Range resolution unit object decision content assignment: assuming that in each resolution element of the 2nd frame exist a target, judge the region that each target of the 2nd frame may exist in the 1st frame, find out resolution element that in this region, target discrimination value is maximum and record the position of this unit, the target discrimination value of the corresponding resolution element of the 2nd frame be the targetpath value function that the maximum target decision content found out in the 1st frame is corresponding with the 2nd this resolution element of frame and;

Step 3.3, the employing method identical with step 3.2 calculate the target discrimination value of the 3rd frame to K frame.

4. a kind of dynamic programming based on phase place detects front tracking as claimed in claim 3, it is characterized in that the value of K in described step 3 is 3 ~ 20.

5. a kind of dynamic programming based on phase place detects front tracking as claimed in claim 1, it is characterized in that the concrete steps of described step 6 are:

Step 6.1, will through judgement all targetpaths in, the targetpath at K frame with maximum target decision content is judged to be real goal flight path;

Step 6.2, the targetpath each being passed through judgement and real goal flight path compare, if this flight path and real goal flight path have same position at certain frame, then this flight path are judged to be false track;

If step 6.3 still has the targetpath not through judging, then repeat step 6.1, step 6.2 until all targetpaths all pass through judgement.