CN1327242C - Method for compensating relative motion of mobile multiple objective for reverse synthetic aperture radar - Google Patents

Abstract

The present invention provides a compensation method for ISAR power-driven multi-target relative movements. Movement compensation is carried out for sub echo by extracting the sub echo of a target of a relative movement from echo data after power-driven multi-target primary compensation. Thus, the influence of Doppler diffusion caused by the relative movement on target imaging is overcome. Then, secondary movement compensation is carried out for the total echo so as to eliminate compensating errors caused by the primary compensation. Thus, a two-dimensional distribution diagram of multi-target clear space of a high-speed movement can be obtained. Complex targets of high-speed movements, a plurality of targets, relative movements, etc. can be imaged by adopting the method of the present invention. The compensation method for ISAR power-driven multi-target relative movements provides a tried and true method for high-speed movement and multi-target two-dimensional imaging.

Description

Inverse synthetic aperture radar (ISAR) multiple-moving target relative motion compensation method

Technical field

The invention belongs to the technical field of synthetic aperture radar image-forming, it is particularly related to the motion compensation technique of the maneuvering target that has relative motion.

Background technology

Inverse synthetic aperture radar (ISAR) (is called for short: ISAR) can carry out round-the-clock, round-the-clock, remote imaging to moving targets such as guided missile, satellite, naval vessel, celestial bodies from fixing or motion platform, all have important use to be worth in strategic defensive, anti-satellite, tactical weapon and radar astronomy.Inverse synthetic aperture radar (ISAR) is based on distance-multispectral and reins in the principle imaging, the key of imaging is motion compensation, by motion compensation, on the target certain is a bit become " fixed point ", then the motion of target is equivalent to rotate around " fixed point ", the ISAR imaging is equivalent to the turntable target imaging, and the latter then is easy to realize.

The method of motion compensation at present is a lot, and in the data that the distance compression is finished, a polynomial expression be finished and be fitted to range-aligned can by the historical time (resembling the peak value or the centre of moment) of envelope simple crosscorrelation or track reference point.Yet, laterally the requirement of Phase Tracking than the range calibration strictness many, and distance error must be controlled in the part of radar wavelength.The existing horizontal phase estimation algorithms of great majority are that the phase history by the strong scattering body of following the tracks of the good target of single isolation obtains horizontal phase error, as strong scattering body algorithm (DSA), multiple scattering body algorithm (MSA) (the statistics scattering centre of moment (SSC) or Doppler's centre of moment).Better performances when DSA has isolated strong scattering body on target, however because the effect of target glint and shade, effect is undesirable as a rule.Algorithm based on the centre of moment is more sane than DSA in some sense, but need carry out phase average, if phase unwrapping is incorrect, this averaging process adds harmful all the better.Document has proposed the autofocus algorithm of a kind of AUTOCLEAN of being called, he is different from other multiple scattering body algorithm, be based on a kind of parametric algorithm of utmost point flexible data model, in two dimension (2-D) image area, select (need not isolate or very strong) a plurality of scatterers automatically, and adopt the best approach that its phase place and RCS information is in addition comprehensive, thereby the phase unwrapping step that avoids trouble.

Distance compression back radar echo signal can be expressed as

$r(m,n)=\left\{\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{\mathrm{\α}}_k{e}^{j4\mathrm{\π}{f}_m[x_k\cos \left(\mathrm{nt}\right)+y_k\sin \left(\mathrm{nt}\right)]/c}\right\}\×e^{j[4\mathrm{\π}{f}_m\mathrm{\Δ}{R}_0\left(\mathrm{nt}\right)}+e(m,n)$

$\begin{array}{cc}0\≤m\≤M-1,& 0\≤n\≤N-1\end{array}$

Form; Here K represents the number of scattering point, α _k, x _k, y _kComplex magnitude, horizontal ordinate and the ordinate of representing K scattering point respectively; R ₀(nt) expression distance moves, and it is poor between n pulse institute's tracking position of object and the radargrammetry, and (m n) represents clutter and noise to e, and c is the light velocity, and M is a sampling number, and N is apart from the umber of pulse in the window; f _mBe the frequency that disperses, its usable samples time t _mBe expressed as follows:

$f_m=f_0+\frac{f_R}{\mathrm{\π}}{t}_m$

F wherein ₀The expression carrier frequency, f _RBe the linear frequency modulation rate.(m n) can be simplified to following form to r

Here frequency is right

Coordinate (x with k scattering point _k, y _k) correspondence,

Corresponding to range migration amount Δ R ₀(nt), { ψ (n) } _N=0 ^N-1The horizontal arbitrarily phase error of expression.See document for details: JIANLI, RENBIAOWU, VICTOR C.CHEN.Robust Autofocus Algorithm for ISAR Imaging of Moving Targets[J] .IEEE Transactions on Aerospace and electronic systems.2001,37 (3): 1056-1069.

Existing all imaging algorithms all are at targets such as aircraft, naval vessels, though the scattering point on these targets is a lot, do not have relative motion between scattering point, do not have variable in distance, can be considered as the single rigid body of a plurality of scattering points.Above-mentioned algorithm is to the good result that arrives of this type games target energy.

Yet all do not solve the problem that has relative motion between multiple goals such as multiple warhead, air formation that resembles in these algorithms; Especially inapplicable especially to the target imaging that relative motion is big between this class high-speed motion of guided missile, spin, target, the distance and the phase place that can not cause the relative motion between target compensate.For the multiple goal echo, when being that reference point is when carrying out motion compensation with wherein any target, even compensation precision is very high, but owing to have relative motion between target, the doppler frequency that relative motion produces can make horizontal generation Doppler frequency-shift, thereby makes the object space distribution that two-dimensional imaging is fuzzy or generation is false (see (b), (c) of accompanying drawing 5; Fig. 6,7,8 (a) figure).Therefore, must the relative motion between target be compensated, eliminate the influence of relative motion imaging results.

Summary of the invention

The purpose of this invention is to provide a kind of ISAR multiple-moving target relative motion compensation method, according to method provided by the invention relative motion between multiple goal is compensated, can eliminate the influence of relative motion to imaging results, make it to satisfy the requirement of high-speed motion, multiple goal ISAR imaging, obtain extraterrestrial target two-dimensional distribution clearly really.

Content of the present invention is eliminated the influence of relative motion to imaging results for target echo being carried out the relative motion compensation.The specific implementation process is as follows:

In order to describe content of the present invention easily, at first do terminological interpretation:

1, about the nonlinear least square method of optimizing parameter is carried out method of estimation

Detailed process is as follows:

1) echo of adjusting the distance after compressing with the envelope cross-correlation method tentatively compensates, and obtains initial

2) feature extraction;

1) in provided

Estimated value

, then can get kinematic parameter by nonlinear least square method

Feasible estimated value , computing formula is

Wherein

3) estimation; By 2) middle estimation

, utilize the nonlinear least square method can be initial Estimated value

Computing formula

Wherein

4) repeat above-mentioned steps, the cost function between two adjacent iteration

Relative variation be lower than certain predetermined thresholding, the parameter value of then last iteration is the final estimated value of parameter.

5) (m n) carries out motion compensation, and computing formula is to r utilize to estimate the parameter value that obtains

The result who obtains is the echo data after the compensation.

For above-mentioned apart from compressed echo signal r (m, n) nonlinear least square method of utilize optimizing is estimated parameter;

Algorithm flow chart is seen accompanying drawing 2, detailed content is seen document: JIAN LI, RENBIAO WU, VICTOR C.CHEN.Robust Autofocus Algorithm for ISAR Imaging of Moving Targets[J] .IEEETransactions on Aerospace and electronic systems.2001,37 (3): 1056-1069.

2, about the time-the frequency analysis method

Utilize STFT (Fourier in short-term) conversion in the time frequency analysis method in the present invention, its expression formula is:

STFT(t，ω)＝∫s(τ)g(τ-t)exp{-jωτ}dτ

Wherein, s (τ) is a signal, and (τ-t) is a time window to g.By selecting a suitable window, just can be in the very narrow time characteristic on the observation signal frequency domain.Consider the characteristic of various windows, in implementation procedure, selected Hamming window type for use.Because it is better that this window type has frequency resolution, the advantage of the rate of decay that secondary lobe is lower and bigger.

The algorithm flow chart that existing time frequency analysis method is carried out lateral processes is seen accompanying drawing 9, and detailed content is seen document: L. Koln work, Bai Juxian translates, time-frequency analysis: theoretical and application, publishing house of Xi'an Communications University.

A kind of inverse synthetic aperture radar (ISAR) multiple-moving target relative motion compensation method, it comprises following step: step 1, distance compression

If the radar emission single pulse signal is p (t), the radar image data is base band echoed signal s _b(t), T _pBe the time of triggering collection gate signal, ω _cBe the radar carrier wave;

At first adopt the frequency domain matched filtering to do digital signal processing, concrete steps are as follows:

1) the echoed signal s that the ISAR radar is gathered _b(t) be the frequency spectrum S that Digital Fourier Transform obtains echoed signal _b(ω);

2) utilize formula S _Ob(t)=p (t-T _p) exp (j ω _cT), calculate with reference to base band echoed signal S _Ob(t);

3) be S _Ob(t) Digital Fourier Transform is to obtain the frequency spectrum S of base band reference signal _Ob(ω);

4) calculate S _b(ω) S _Ob ^*Contrary Fourier conversion (ω) obtains base band matched filtering S as a result _Mb(t), the result who obtains is the echo data after the distance compression;

Step 2, echo is carried out distance and phase place compensates for the first time

(m n) utilizes the nonlinear least square method of optimizing that parameter is estimated apart from compressed echo signal r by step 1;

It is characterized in that it also comprises following step:

The sub-echo of step 3, extraction relative motion target

1) to carrying out amplitude normalization, each normalization echo distance images of imaging is asked on average, obtained more stable mean distance picture through the echo data after the first compensation of step 2;

2) set an amplitude thresholding, from the mean distance picture that obtains, find out the position of stronger scattering point by the amplitude thresholding;

3) use the scattering point positional information that obtains from the overall echo of imaging, to extract the sub-echo information of corresponding relative motion target;

Step 4, the sub-echo that extracts is carried out distance and phase compensation

Utilize the method in the step 2 that the sub-echo that extracts in the step 3 is carried out distance and phase compensation, and the overall echo data after the first compensation of sub-echo data renewal after the compensation; Obtain the overall echo after relative motion compensates;

Step 5, overall echo is carried out quadratic phase compensation

Utilize the method in the step 2 that step 4 is carried out motion compensation once more through the echo data that relative motion compensates; Obtain the echo data after the required motion compensation of imaging;

Step 6, when utilizing-the frequency analysis method is to overall echo two-dimensional imaging

When the overall echo after the compensation that step 5 is obtained utilizes-the frequency analysis method carries out lateral processes, obtains the two-dimensional distribution of target.

Through after the above step, just can realize the relative motion compensation of ISAR multiple-moving target, obtain the two-dimensional distribution of target.

A kind of ISAR multiple-moving target relative motion compensation method process flow diagram of the present invention as shown in Figure 4.Accompanying drawing 5 to 8 is relative motion compensating image example.

Need to prove:

The purpose of above-mentioned steps 2 mainly is to be that reference point is carried out motion compensation with a certain scattering point in the target, after motion compensation, target echo is that reference point is alignd with certain scattering point, but because other targets exist the relative motion with respect to reference target, other targets of adjacent echo still may exist translation component, and possibility cross over distance resolution element, seriously influence image quality relative motion target the influence of imaging is not eliminated.Therefore also must carry out motion compensation to the relative motion target; Its process flow diagram as shown in Figure 2.

The purpose of above-mentioned steps 3 is by drawing the echo position of relative motion target in the distance images after the first compensation, extract the sub-echo of relative motion target from overall echo; To the echo distance images in the same imaging time relatively, the significant maximum point of change in location is found out in the position of calculating maximum point in the echo distance images respectively, and then the pairing target of this extreme point is the target that has relative motion.If the distance of this target and other scattering point is far away, then can therefrom extract independently echo of this scattering point, shown in Fig. 2 (b).

The purpose of step 4 mainly is that the relative motion target echo is compensated, and eliminates the influence of the Doppler frequency of relative motion generation to imaging;

Because the existence of relative motion makes to make that in the compensation process of step 2 other scattering point has produced bigger error in compensation process, the fundamental purpose of step 5 is to reduce this error, makes it to satisfy the requirement of imaging;

Handle by step 4,5, can eliminate when in step 2, carrying out motion compensation because of the influence of the Doppler frequency of target caused by relative motion to phase of echo.Thereby make whole echo obtain motion compensation completely.

Innovation of the present invention is: extract the sub-echo of relative motion target from the echo data after the first compensation of multiple-moving target, the antithetical phrase echo carries out motion compensation, thereby overcomes the influence of Doppler's diffusion couple target imaging that relative motion causes; Again overall echo is carried out the secondary motion compensation, eliminate the compensating error that first compensation causes; Thereby can obtain the multiobject clear space two-dimensional distribution plan of high-speed motion; For the multiobject two-dimensional imaging of high-speed motion provides a kind of efficient ways.

Principle Analysis of the present invention

As Fig. 1, establish radar and be positioned at initial point L, A, B are two scatterers of imaging space, the distance of initial time A and radar is R ₀, the distance of A and B is r ₀, the movement velocity of target A is

With

Angle be β, the relative velocity of A and B is v ₀, and

With speed

Direction identical, initial time B is R to the distance of radar _b=R ₀+ r ₀Cos β.T target A, B move to A ', B ' respectively constantly.This moment, target A to the distance of radar was

$R_0\left(t\right)=\sqrt{{({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+\mathrm{vt}\cos \mathrm{\&beta;})}^2+{\left(\mathrm{vt}\sin \mathrm{\&beta;}\right)}^2}$

Work as R ₀During＞＞vt, $R_0\left(t\right)\&cong;{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+\mathrm{vt}\cos \mathrm{\&beta;}+\frac{{\left(\mathrm{vt}\sin \mathrm{\&beta;}\right)}^2}{2({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+\mathrm{vt}\cos \mathrm{\&beta;})}---\left(1\right)$

And constantly t target B to the distance of radar L is

$R_b\left(t\right)\&cong;{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+\left[r_0(v+v_0)t\right]\cos \mathrm{\&beta;}+\frac{{\{{[\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0+(v+v_0)t]\sin \mathrm{\&beta;}\}}^2}{2\{{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+[{\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0+(v+v_0)t\left]\cos \mathrm{\&beta;}\right\}}---\left(2\right)$

With A is that reference point is carried out motion compensation, promptly is equivalent to A is moved to R ₀On the circumference for radius, be about at A ' and move to A ", then B ' moves to B ", when not having relative motion between target, t target B moves to a C constantly, and compensation back C to the distance of radar is

$R_c^{\text{'}}\left(t\right)=\sqrt{{({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+{\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0\cos (\mathrm{\&beta;}-\mathrm{\&theta;}))}^2+{\left({\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0\sin (\mathrm{\&beta;}-\mathrm{\&theta;})\right)}^2}$

Wherein θ be radar line of sight corner and $\mathrm{\&theta;}\&ap;\frac{\mathrm{vt}\sin \mathrm{\&beta;}}{{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}$ , then

$R_c^{\&prime;}\left(t\right)\&ap;{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+{\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0\cos \mathrm{\&beta;}+\frac{{\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0\mathrm{vt}{\sin }^2\mathrm{\&beta;}}{{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}+\frac{{\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0{\sin }^2\mathrm{\&beta;}}{2{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}$

At this moment, C with respect to the range difference of radar is

$\mathrm{\&Delta;}{R}_c\left(t\right)=R_c^{\&prime;}\left(t\right)-R_b==\frac{{\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0\mathrm{vt}{\sin }^2\mathrm{\&beta;}}{{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}+\frac{{\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0{\sin }^2\mathrm{\&beta;}}{2{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}$

So the doppler frequency that the motion of B point produces when not having relative motion is

$f_d^{\&prime;}=\frac{2\mathrm{<figure-callout\; id="0"\; label="f\; c\; r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">f</figure-callout>}}{c}\frac{r_{}}{}\frac{{}_0{\sin }^2\mathrm{\&beta;}}{R_0}---\left(3\right)$

And B " distance to radar is

$R_n^{\&prime;}\left(t\right)=\sqrt{{[{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+({\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0+v_{0}t)\cos (\mathrm{\&beta;}-\mathrm{\&theta;})]}^2+{\left[({\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0+v_{0}t)\sin (\mathrm{\&beta;}-\mathrm{\&theta;})\right]}^2}$

$\&ap;{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+({\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0+v_{0}t)\cos \mathrm{\&beta;}+({\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0+v_{0}t)\frac{\mathrm{vt}{\sin }^2\mathrm{\&beta;}}{{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}+\frac{{\left(({\mathrm{<figure-callout\; id="0"\; label="r"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">r</figure-callout>}}_0+v_{0}t)\sin \mathrm{\&beta;}\right)}^2}{2{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}$

Then the variable in distance with respect to radar that causes of relative velocity is

$\mathrm{\&Delta;R}\left(t\right)=R_b^{\&prime;}\left(t\right)-R_c^{\&prime;}=v_{0}t\cos \mathrm{\&beta;}+\frac{{\mathrm{<figure-callout\; id="0"\; label="v"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">v</figure-callout>}}_0{\mathrm{vt}}^2{\sin }^2\mathrm{\&beta;}}{{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}+\frac{[{\left(v_{0}t\right)}^2+{2b}_0{v}_{0}t]{\sin }^2\mathrm{\&beta;}}{2{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}$

So the doppler frequency that causes because of relative motion is

$f_d=\frac{2f}{c}\frac{\mathrm{d\&Delta;R}}{\mathrm{dt}}=\frac{2f}{c}({\mathrm{<figure-callout\; id="0"\; label="v"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">v</figure-callout>}}_0\cos \mathrm{\&beta;}+\frac{2{\mathrm{<figure-callout\; id="0"\; label="v"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">v</figure-callout>}}_0\mathrm{vt}{\sin }^2\mathrm{\&beta;}}{{\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">R</figure-callout>}}_0}+\frac{[{2\mathrm{<figure-callout\; id="0"\; label="v"\; filenames="C20041004030400161.png,C20041004030400191.png"\; state="\{\{state\}\}">v</figure-callout>}}_0^{2}t+{2r}_0{v}_0]{\sin }^2\mathrm{\&beta;}}{2R_0})---\left(4\right)$

F ' wherein _dBe target when not having relative motion, target Equivalent is the doppler frequency that the turntable rotation is produced, and is that radar imagery needs; And f _dBe the doppler frequency that the motion of target relative reference point is produced, work as f _dGreater than laterally multispectral when reining in resolution, will laterally produce multispectral rein in fuzzy.Must compensate relative motion.For the moving target that has acceleration, in same imaging time, can regard uniformly accelrated rectilinear motion as.At this moment as long as in the above influence that adds acceleration in various.

Adopt method of the present invention then can eliminate of the influence of the Doppler shift of relative motion generation by compensating that speed of related movement is caused apart from translation to imaging.

A kind of ISAR multiple-moving target relative motion compensation method provided by the invention in sum, it is by eliminating the horizontal generation Doppler frequency-shift that relative motion produces, thereby eliminated the false target fuzzy and that generate because of the relative motion influence of the lateral separation in the two-dimensional imaging, obtained the correct two dimensional image of ISAR multiple-moving target.Adopt method of the present invention and to have the complex target imaging of relative motion etc. to high-speed motion, multiple goal, it provides a kind of efficient ways for the multiobject two-dimensional imaging of high-speed motion.

Description of drawings

Fig. 1 is an ISAR multiple-moving target relative motion compensation synoptic diagram

Wherein L is the position at radar place, and A, B are the position of initial time two targets, and the distance of initial time A and radar is R ₀, the distance of A and B is r ₀, the movement velocity of target A is

With Angle be β, the relative velocity of A and B is v ₀T target A, B move to A ', B ' respectively constantly; With A is that reference point is carried out motion compensation, promptly is equivalent to A is moved to R ₀On the circumference for radius, be about at A ' and move to A ", then B ' moves to B "; Because B exists relatively and the relative motion of target A, the distance of B this moment " to A " also is not equal to the distance of B to A, promptly is that target B still existed apart from translation after reference point was carried out motion compensation with A, if relative motion is not compensated, imaging results must exist doppler ambiguity and diffusion.Therefore must be to compensating that relative motion produces apart from translation.

Fig. 2 carries out distance and the first backoff algorithm process flow diagram of phase place for existing to echo

Wherein,

The number of expression scattering point, For

Intermediate variable, For Estimated value.

Fig. 3 extracts the synoptic diagram of the sub-echo of relative motion target from overall echo for step 3 of the present invention

Wherein, Fig. 3 (a) is for can not extract sub-echo information synoptic diagram;

In Fig. 3 (a), the echo of the relative motion target in the square frame and the echo of adjacent target are overlapped, and can not extract the sub-echo of relative motion target this moment, therefore can't compensate relative motion;

Fig. 3 (b) is for extracting sub-echo information synoptic diagram

In Fig. 3 (b), the distance of other scattering point of relative motion target is far away, then can therefrom extract independently echo of this scattering point; The data of extracting are carried out the influence that motion compensation then can be eliminated relative motion; M, n, m+p represent which sampled point of echo.

Fig. 4 is a system flowchart of the present invention

Wherein, the auto-clean algorithm is the algorithm of step 2.

Fig. 5 is the imaging simulation of simulated data of the present invention figure as a result

Wherein figure (a) is the two-dimensional image of target when not having relative motion between three targets

Figure (b) adds speed of related movement v respectively for the intermediate objective to figure (a) ₀Behind=the 10m/s, do not carry out the two-dimensional image of relative motion compensation

Figure (c) adds speed of related movement v respectively for the intermediate objective to figure (a) ₀Behind=the 20m/s, do not carry out the two-dimensional image of relative motion compensation

Because the Doppler shift that relative motion causes makes the relative motion target among figure (b), (c) depart from actual position, and laterally producing certain doppler ambiguity;

Figure (d) is the two-dimensional image after using compensation method of the present invention to (b), (c) compensation

Figure (d) is consistent with the imaging results of figure (a).This result shows that method of the present invention has compensated the influence of relative motion to two-dimensional imaging, is effective to the relative motion compensation;

Fig. 6,7,8 is the imaging results figure of certain radar measured data

(a) figure among these several figure is the imaging results before the relative motion compensation, (b) is the imaging results after the relative motion compensation.By these several figure as can be seen, relative motion make target two-dimensional image laterally producing serious fuzzy, the number that can not resolution target from image and the true distribution situation of two dimension of target.Through relative motion compensation, eliminated the influence of Doppler shift that relative motion causes to the target two-dimensional image, therefore obtained clear, the real two-dimensional distribution of target, shown in (b) figure among Fig. 6,7,8;

The program flow diagram that Fig. 9 carries out for time frequency analysis method (STFT)

Wherein g (τ) is a time window, and s (τ) is the echo data after the motion compensation, and FFT is a fast fourier transform, and (t τ) is the product of g (τ) and s (τ) to f;

Embodiment:

We adopt method of the present invention to carry out the relative motion compensation for a plurality of multiple-moving targets of L-113 radar collection, adopt the method for VC programming, can obtain the result shown in Fig. 6～8;

From Fig. 5,6,7,8 as can be seen, algorithm of the present invention can well between target because of compensating that relative motion causes apart from translation, eliminate the influence of relative motion to the quality of imaging, obtain truly, object space Two dimensional Distribution clearly, therefore algorithm of the present invention is very effective to the multi-target imaging of forms of motion complexity, high-speed motion.

Claims (1)

1, a kind of inverse synthetic aperture radar (ISAR) multiple-moving target relative motion compensation method, it comprises following step:

Step 1, distance compression

If the radar emission single pulse signal is p (t), the radar image data is base band echoed signal s _b(t), T _pBe the time of triggering collection gate signal, ω _cBe the radar carrier wave;

At first adopt the frequency domain matched filtering to do digital signal processing, concrete steps are as follows:

1) the echoed signal s that the ISAR radar is gathered _b(t) be the frequency spectrum S that Digital Fourier Transform obtains echoed signal _b(ω);

2) utilize formula S _Ob(t)=p (t-T _p) exp (j ω _cT), calculate with reference to base band echoed signal S _Ob(t);

3) be S _Ob(t) Digital Fourier Transform is to obtain the frequency spectrum S of base band reference signal _Ob(ω);

4) calculate S _b(ω) S _Ob ^*Contrary Fourier conversion (ω) obtains base band matched filtering S as a result _Mb(t), the result who obtains is the echo data after the distance compression;

Step 2, echo is carried out distance and phase place compensates for the first time

By the result that the distance of the first step is compressed, echoed signal can be expressed as

$r(m,n)=\left\{\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_k{e}^{j4\mathrm{\&pi;}{f}_m\&lsqb;x_k\cos \left(\mathrm{nt}\right)+y_k\sin \left(\mathrm{nt}\right)]/c}\right\}\&times;e^{j\&lsqb;4\mathrm{\&pi;}{f}_m\mathrm{\&Delta;}{R}_0\left(\mathrm{nt}\right)}+e(m,n)$

0≤m≤M-1， 0≤n≤N-1

Form; Here K represents the number of scattering point, α _k, x _k, y _kComplex magnitude, horizontal ordinate and the ordinate of representing K scattering point respectively; R ₀(nt) expression distance moves, and it is poor between n pulse institute's tracking position of object and the radargrammetry, and (m n) represents clutter and noise to e, and c is the light velocity, and M is a sampling number, and N is apart from the umber of pulse in the window; f _mBe the frequency that disperses, its usable samples time t _mBe expressed as follows:

$f_m=f_0+\frac{f_R}{\mathrm{\&pi;}}{t}_m$

F wherein ₀The expression carrier frequency, f _RBe the linear frequency modulation rate.

(m n) can be simplified to following form to r

Here frequency to ( ) with the coordinate (x of k scattering point _k, y _k) correspondence, Corresponding to range migration amount Δ R ₀(nt), { ψ (n) } _N=0 ^N-1The horizontal arbitrarily phase error of expression.

For above-mentioned apart from compressed echo signal r (m, n) nonlinear least square method of utilize optimizing is estimated parameter;

It is characterized in that it also comprises following step

The sub-echo of step 3, extraction relative motion target

1) to carrying out amplitude normalization, each normalization echo distance images of imaging is asked on average, obtained more stable mean distance picture through the echo data after the first compensation of step 2;

2) set an amplitude thresholding, from the mean distance picture that obtains, find out the position of stronger scattering point by the amplitude thresholding:

3) use the scattering point positional information that obtains from the overall echo of imaging, to extract the sub-echo information of corresponding relative motion target;

Step 4, the sub-echo that extracts is carried out distance and phase compensation

Utilize the method in the step 2 that the sub-echo that extracts in the step 3 is carried out distance and phase compensation, and the overall echo data after the first compensation of sub-echo data renewal after the compensation; Obtain the overall echo after relative motion compensates;

Step 5, overall echo is carried out quadratic phase compensation

Utilize the method in the step 2 that step 4 is carried out motion compensation once more through the echo data that relative motion compensates; Obtain the echo data after the required motion compensation of imaging;

Step 6, when utilizing-the frequency analysis method is to overall echo two-dimensional imaging

When the overall echo after the compensation that step 5 is obtained utilizes-the frequency analysis method carries out lateral processes, obtains the two-dimensional distribution of target.

Images