CN102998672B - Step frequency inverse synthetic aperture radar (ISAR) imaging method based on coherent processing - Google Patents

Abstract

The invention relates to a step frequency inverse synthetic aperture radar (ISAR) imaging method based on coherent processing. The method mainly solves the problems in a traditional parameter estimation method of being low in calculating efficiency and having large effects of phase error on frequency synthesis after envelope compensation. The method includes the steps: building a step frequency ISAR echo signal model; conducting motion parameter estimation according to the echo signal model; constructing a penalty function and compensating a sub-pulse envelope of echo signals; conducting frequency synthesis based on coherent processing on the compensated echo signals; conducting inverse fast Fourier transformation (IFFT) on the synthesized echo signals, and obtaining a high-resolution range profile; and conducting direction imaging through an RD algorithm or an RID algorithm to obtain a directional high-resolution image. Experiments prove that the method has the advantages of being high in phase error compensation precision and calculation efficiency and can be used for achieving two-dimensional high-resolution ISAR imaging of a target.

Description

Step frequency ISAR formation method based on phase drying and other treatment

Technical field

The invention belongs to Radar Technology field, particularly inverse synthetic aperture radar (ISAR) high-resolution imaging technology, can be used for the high-resolution two-dimensional imaging of realize target.

Background technology

High-resolution ISAR imaging radar, due to two-dimensional imaging that can realize target, obtains abundanter target scattering information, is conducive to identification and the classification of target, thereby is widely used in modern radar system.Target is carried out high-resolution ISAR imaging and need to be improved the vertical and horizontal resolution of radar.Lateral resolution is to realize relative to rotating of radar by target, and longitudinal frame is to realize by the large-signal bandwidth of radar emission.

Linearly modulated stepped frequency is because the longitudinal frame that can synthesize large signal bandwidth raising radar is widely used in high-resolution imaging radar system.Linearly modulated stepped frequency is synthetic generally has three kinds of methods apart from high resolution processing: synthetic apart from envelope method, time domain synthetic method and frequency domain synthetic method.Be subject to more attention and application because frequency domain synthetic method operation efficiency is high, can not produce false pseudo-peak.

The kinematic parameter that focuses on moving target of step frequency high-resolution ISAR imaging algorithm is estimated and kinematic error compensates, and for the synthesis of High Range Resolution, and finally obtains the high-resolution two dimensional image of target.Utilize Minimum entropy method to carry out parameter estimation to moving target, and compensate the kinematic error in arteries and veins group in echo data.Carry out the kinematic parameter of matching target by the amount of movement of echo envelope.But this several method has a lot of deficiencies in practical application:

1. Minimum entropy method need to utilize the method for search to estimate kinematic parameter, and its search stepping amount size is difficult to control, and operation efficiency is lower.

2. the kinematic parameter that carrys out matching target by the amount of movement of echo envelope, its evaluated error is larger, and residual phase error is also large, has a strong impact on the coherence between each subpulse in arteries and veins group, makes target after synthetic look like to occur the phenomenon of main lobe broadening and secondary lobe variation.

Summary of the invention

The object of the invention is to the deficiency for above-mentioned prior art, a kind of step frequency ISAR formation method based on phase drying and other treatment has been proposed, efficiently, accurately to compensate in step frequency signal relative envelope and the phase error between each subpulse in arteries and veins group, obtain better the synthetic high-resolution one-dimensional range profile of target.

Realizing the object of the invention technical thought is: by neighboring and correlative method, echo data is carried out to dimension-reduction treatment, utilize the correlation energy peak value of Fourier pair echo data to accumulate; Utilize peak value after accumulation distance to orientation to projected position, estimate rapidly speed and the acceleration of moving target, the interpulse envelope migration of structure penalty function syndrome; Utilize phase drying technique to proofread and correct the orientation phase error of each subpulse in arteries and veins group, obtain the high-resolution one-dimensional range profile of target by frequency domain synthetic method; Utilize RD algorithm or RID algorithm to carry out orientation imaging, obtain the full resolution pricture of target azimuth.Implementation step comprises as follows:

(1) set up step frequency ISAR echo signal model;

(2) carry out kinematic parameter estimation according to echo signal model:

2a) first subpulse data of different arteries and veins groups are taken out respectively, obtain echoed signal neighboring and correlative expression formula and be:

$R(\mathrm{\τ},m{T}_a)={\mathrm{\σ}}_P\sin c\left(A(\mathrm{\τ}+\frac{2M}{c})\right)\×\exp \left(j\frac{4\mathrm{\π}}{\mathrm{\λ}}{M}_1\right)\×\exp \left(j\frac{4\mathrm{\π}}{\mathrm{\λ}}{M}_2\right)$

Wherein, τ is correlation time, and m is arteries and veins group number, 0≤m≤M-1, mT _afor the discrete representation of orientation time, σ _pfor target echo backscattering coefficient, A is relevant matches amount, and c is the light velocity, and λ is signal wavelength, M=M ₁+ M ₂, ${\mathrm{<figure-callout\; id="1"\; label="M"\; filenames="HDA00002477120200031.png"\; state="\{\{state\}\}">M</figure-callout>}}_1=(v_r-x_P\mathrm{\&omega;})T_a+\frac{1}{2}{a}_r{T}_a^2,$ M ₂＝T _aa _rmT _a

Wherein, v _rthe radial velocity of the relative radar motion of target, x _pbe the lateral attitude information of scattering point, ω is the rotational angular velocity of the relative radar motion of target, T _afor the repetition period between arteries and veins group, a _rit is the radial acceleration of the relative radar motion of target.

To echoed signal neighboring and correlative expression formula along orientation to doing FFT accumulation, obtaining orientation frequency domain echo signal neighboring and correlative expression formula be 2b):

$R(\mathrm{\&tau;},f_a)={\mathrm{\&sigma;}}_P\sin c\left(A(\mathrm{\&tau;}+\frac{2M}{c})\right)\&times;\sin c\left(\mathrm{\&pi;}{T}_m(f_a-\frac{2T_a}{\mathrm{\&lambda;}}{a}_r)\right)\&times;\exp \left(j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}{M}_1\right)$

Wherein, T _mfor subpulse repetition period, f _afor orientation is to frequency,

it is the correlation peak location of echoed signal neighboring and correlative expression formula;

Ignore the envelope variation that in M, adjacent twice return ω causes, correlation peak location can be expressed as again

2c) by correlation peak location distance to orientation to coordinate be made as respectively

with

calculate the radial acceleration a of target _rand speed v _r:

$\left\{\begin{array}{cc}{a}_r=\frac{\mathrm{\&lambda;prf}[\tilde{n}-(\frac{\tilde{N}}{2}+1)]}{2N{T}_a}& \tilde{n}=1,.....,\tilde{N}\\ v_r=-\frac{c[\tilde{m}-(\frac{\tilde{M}}{2}+1)]}{2f_s{T}_a}-\frac{\mathrm{\&lambda;}[\tilde{n}-(\frac{\tilde{n}}{2}+1)]}{4\tilde{N}{T}_a}& \tilde{m}=1,.....,\tilde{M}\end{array}\right.$

Wherein,

with

be respectively distance to orientation to discrete counting, prf be orientation to sample frequency, f _sfor distance is to sample frequency, T _rfor the pulse repetition time;

(3) according to radial acceleration a _rand speed v _r, utilize the phase compensation function s of following envelope cancellation _srefsubpulse envelope to echoed signal compensates,

$s_{\mathrm{sref}}=\exp (-j\frac{4\mathrm{\&pi;}}{c}(v_r{f}_r{t}_m+\frac{1}{2}{a}_r{f}_r{t}_m^2))$

Wherein, f _rfor distance is to frequency, t _mfor orientation is to the time;

(4) echoed signal that completes subpulse envelope cancellation is carried out to the frequency synthesis based on phase drying and other treatment, the echoed signal frequency spectrum s (f after being synthesized _r, T _r);

(5) to the echoed signal frequency spectrum s (f after synthetic _r, T _r) carry out contrary Fourier IFFT conversion, obtain high-resolution one-dimensional range profile;

(6) utilize RD algorithm or RID algorithm to carry out orientation imaging to high-resolution one-dimensional range profile, obtain the full resolution pricture of target azimuth.

The present invention compared with prior art has the following advantages

The first, the present invention's employing is first constructed wave filter and is realized range pulse compression, intercepts out backward energy data, utilizes motion compensation and phase coherenceization to realize the method synthetic apart from high-resolution, with respect to traditional frequency domain synthetic method, reduces handled data volume;

The second, the present invention utilizes the neighboring and correlative of adjacent twice return, obtains the position of main peak value energy accumulation along orientation accumulation, estimates the value of speed and acceleration, makes estimated accuracy, estimated efficiency higher.Even than also applying in situation, little on estimated accuracy impact at low signal;

The 3rd, because the precision of the required compensation of phase place will be higher than the precision of envelope cancellation, the parameters of target motion precision that the conventional method such as minimum entropy and data envelopment fitting is estimated is not high enough again, and often synthetic effect is not obvious.The present invention adopts the method for phase drying and other treatment to compensate phase error between subpulse, and synthetic effect is obvious.

Accompanying drawing explanation

Fig. 1 is realization flow figure of the present invention;

Fig. 2 is the sub-process figure that in the present invention, the parameters of target motion are estimated;

Fig. 3 is the schematic diagram of adjacent subpulse frequency spectrum shift in the present invention;

Fig. 4 is simulation objectives high-resolution ISAR imaging results figure of the present invention;

Fig. 5 is that the present invention surveys Ship Target step frequency high-resolution ISAR imaging results figure.

Embodiment

Below in conjunction with accompanying drawing, the present invention will be further described.

With reference to Fig. 1, specific embodiment of the invention step is as follows:

Step 1. is set up step frequency ISAR echo signal model.

The waveform of supposing radar emission linearly modulated stepped frequency is:

U(t)＝u ₁(t)exp(j2π(f ₀+nΔf)t) 0≤n≤N-1

In formula, u ₁(t)=rect (t/T ₁) exp (j π γ t ²) be linear frequency modulation subpulse, exp is the exponential function truth of a matter, j is imaginary number, f ₀for fundamental frequency, n is frequency modulation stepping subpulse number, and Δ f is number of frequency steps, f ₀+ n Δ f is the carrier frequency of n frequency modulation stepping subpulse, and t is the time, and N is the stepping frequency modulation number of subpulse in each arteries and veins group, T ₁for subpulse width, γ is subpulse frequency modulation rate;

The distance table that during by radar emission n sub-pulse signal, moving target is taken up an official post between meaning one scattering point P and radar is shown:

$R_P\left(t_m\right)\&ap;(R_0+y_P)+(v_r-\mathrm{\&omega;}{x}_P)t_m+\frac{1}{2}{a}_{\mathrm{<figure-callout\; id="2"\; label="r\; t\; m"\; filenames="HDA00002477120200031.png"\; state="\{\{state\}\}">r</figure-callout>}}{t}_m^{}{}_2^{}$

In formula, t _mfor orientation time, R ₀for target is to the initial action distance of radar, x _pand y _prespectively the horizontal and vertical positional information of scattering point P, v _rbe the radial velocity of the relative radar motion of target, ω is the rotational angular velocity of the relative radar motion of target, a _rit is the radial acceleration of the relative radar motion of target;

Utilize U (t) and R _p(t _m), the echoed signal that obtains scattering point P is:

$s_n\left(t\right)={\mathrm{\&sigma;}}_P\mathrm{rect}\left(\frac{t-2R_P\left(t_m\right)/c}{T_1}\right)\exp \left(\mathrm{j\&pi;\&gamma;}{(t-\frac{2R_P\left(t_m\right)}{c})}^2\right)\exp (-j4\mathrm{\&pi;}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+\mathrm{n\&Delta;f})\frac{R_P\left(t_m\right)}{c})$

In formula, rect is rectangular function, and c is the light velocity, σ _pfor target echo backscattering coefficient;

Step 2. is carried out kinematic parameter estimation according to echo signal model.

With reference to Fig. 2, being implemented as follows of this step:

2a) to echoed signal s _n(t) carry out the pulse compression apart from matched filtering, the echoed signal obtaining after pulse compression is:

$s_n^{\&prime;}\left(t\right)={\mathrm{\&sigma;}}_P\sin c\left(B(t-\frac{2R_P\left(t_m\right)}{c})\right)\exp (-j\frac{4\mathrm{\&pi;}{R}_P\left(t_m\right)}{c}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+\mathrm{n\&Delta;f}))$

In formula, B is transmitting subpulse signal bandwidth;

2b) taking out respectively first subpulse echoed signal of different arteries and veins groups after pulse compression is:

$s_{f_0}\left(t\right)={\mathrm{\&sigma;}}_P\sin c\left(B(t-\frac{2R_P\left(t_m\right)}{c})\right)\exp (-j\frac{4\mathrm{\&pi;}{f}_0}{c}{R}_P\left(t_m\right))$

The neighboring and correlative expression formula that obtains echoed signal is:

$R(\mathrm{\&tau;},m{T}_a)={\mathrm{\&sigma;}}_P\sin c\left(A(\mathrm{\&tau;}+\frac{2M}{c})\right)\&times;\exp \left(j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}{M}_1\right)\&times;\exp \left(j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}{M}_2\right)$

Wherein, τ is correlation time, and m is arteries and veins group number, 0≤m≤M-1, mT _afor the discrete representation of orientation time, A is relevant matches amount, and λ is signal wavelength, M=M ₁+ M ₂,

m ₂=T _aa _rmT _a, in formula, T _afor the repetition period between arteries and veins group;

To echoed signal neighboring and correlative expression formula along orientation to doing Fourier transform FFT accumulation, obtaining orientation frequency domain echo signal neighboring and correlative expression formula be 2c):

$R(\mathrm{\&tau;},f_a)={\mathrm{\&sigma;}}_P\sin c\left(A(\mathrm{\&tau;}+\frac{2M}{c})\right)\&times;\sin c\left(\mathrm{\&pi;}{T}_m(f_a-\frac{2T_a}{\mathrm{\&lambda;}}{a}_r)\right)\&times;\exp \left(j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}{M}_1\right)$

Wherein, T _mfor subpulse repetition period, f _afor orientation is to frequency,

it is the correlation peak location of echoed signal neighboring and correlative expression formula;

Ignore the envelope variation that in M, adjacent twice return ω causes, correlation peak location can be expressed as again:

2d) by correlation peak location distance to orientation to coordinate be made as respectively with

, the radial acceleration a of calculating target _rand speed v _r:

$\left\{\begin{array}{cc}{a}_r=\frac{\mathrm{\&lambda;prf}[\tilde{n}-(\frac{\tilde{N}}{2}+1)]}{2N{T}_a}& \tilde{n}=1,.....,\tilde{N}\\ v_r=-\frac{c[\tilde{m}-(\frac{\tilde{M}}{2}+1)]}{2f_s{T}_a}-\frac{\mathrm{\&lambda;}[\tilde{n}-(\frac{\tilde{n}}{2}+1)]}{4\tilde{N}{T}_a}& \tilde{m}=1,.....,\tilde{M}\end{array}\right.$

Wherein,

with

be respectively distance to orientation to discrete counting, prf be orientation to sample frequency, f _sfor distance is to sample frequency, T _rfor the pulse repetition time.

Step 3. is configured to the phase compensation function of envelope cancellation.

According to radial acceleration a _rand speed v _r, utilize the phase compensation function s of following envelope cancellation _srefsubpulse envelope to echoed signal compensates,

$s_{\mathrm{sref}}=\exp (-j\frac{4\mathrm{\&pi;}}{c}(v_r{f}_r{t}_m+\frac{1}{2}{a}_r{f}_r{t}_m^2)),$

Wherein, f _rfor distance is to frequency, t _mfor orientation is to the time.

Step 4. is carried out the frequency synthesis based on phase drying and other treatment.

4a) utilize phase compensation function to proofread and correct the envelope of subpulse, obtain the echoed signal after envelope is proofreaied and correct:

$s_n^{\&prime;\&prime;}\left(t\right)={\mathrm{\&sigma;}}_P\sin c\left(B(t-\frac{2({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P)}{c})\right)\exp (-j\frac{4\mathrm{\&pi;}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+\mathrm{n\&Delta;f})}{c}({(\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P)+(v_r-\mathrm{\&omega;}{x}_P)t_m+\frac{1}{2}{a}_r{t}_m^2))$

Echoed signal s after envelope is proofreaied and correct " _n(t) transform to apart from frequency domain s _n(f _r, nT _r)

$s_n(f_r,n{T}_r)=\widetilde{\mathrm{\&sigma;}}\mathrm{rect}\left(\frac{f_r}{B}\right)\&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0}{c}({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P))\&times;\exp (-j\frac{4\mathrm{\&pi;}(f_r+\mathrm{n\&Delta;f})}{c}({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P))$

$\&times;\exp (-j\frac{4\mathrm{\&pi;}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+\mathrm{n\&Delta;f})(v_r-\mathrm{\&omega;}{x}_P)n{T}_r+2\mathrm{\&pi;}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+\mathrm{n\&Delta;f})a_r{\left(n{T}_r\right)}^2}{c})$

Wherein, f _rfor distance is to frequency, nT _rrepresent the orientation time,

for range coefficient, B is transmitting subpulse signal bandwidth;

4b) adjacent subpulse frequency domain echo signal location and phase place are changed, variable quantity is Δ f/2,

A n pulse frequency domain echoed signal before position and phase place change is as Fig. 3 (a),

A n+1 pulse frequency domain echoed signal before position and phase place change is as Fig. 3 (b),

$s_n(f_r+\mathrm{\&Delta;f}/2,n{T}_r)=\widetilde{\mathrm{\&sigma;}}\mathrm{rect}\left(\frac{f_r+\mathrm{\&Delta;f}/2}{B}\right)\&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0}{c}({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P))$

$\&times;\exp (-j\frac{4\mathrm{\&pi;}(f_r+\mathrm{n\&Delta;f}+\mathrm{\&Delta;f}/2)}{c}({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P))$

$\&times;\exp (-j\frac{4\mathrm{\&pi;}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+\mathrm{n\&Delta;f})(v_r-\mathrm{\&omega;}{x}_P)n{T}_r+2\mathrm{\&pi;}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+\mathrm{n\&Delta;f})a_r{\left(n{T}_r\right)}^2}{c})$

$s_{n+1}(f_r-\mathrm{\&Delta;f}/2,(n+1)T_r)=\widetilde{\mathrm{\&sigma;}}\mathrm{rect}\left(\frac{f_r-\mathrm{\&Delta;f}/2}{B}\right)\&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0}{c}({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P))$

$\&times;\exp (-j\frac{4\mathrm{\&pi;}(f_r+\mathrm{n\&Delta;f}+\mathrm{\&Delta;f}/2)}{c}({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P))$

$\&times;\exp (-j\frac{4\mathrm{\&pi;}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+(n+1)\mathrm{\&Delta;f})(v_r-\mathrm{\&omega;}{x}_P)(n+1)T_r+2\mathrm{\&pi;}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+(n+1)\mathrm{\&Delta;f})a_r{\left((n+1)T_r\right)}^2}{c})$

Wherein, s _n(f _r+ Δ f/2, nT _r) be the subpulse frequency domain echo signal after n position and phase place change, s _n+1(f _r-Δ f/2, (n+1) T _r) be the subpulse frequency domain echo signal after n+1 position and phase place change, f _rscope be [B _c/ 2～B _c/ 2], B _cfor shared signal bandwidth, nT _rrepresent the orientation time, the adjacent subpulse frequency domain echo signal after position and phase place change is as Fig. 3 (c);

4c) obtain the poor ΔΦ of conjugate phase that changes frequency domain echo signal between rear adjacent subpulse _n:

$\mathrm{\&Delta;}{\mathrm{\&Phi;}}_n=\frac{4\mathrm{\&pi;}{T}_r({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+\mathrm{n\&Delta;f})(v_r-\mathrm{\&omega;}{x}_P)+\mathrm{\&Delta;f}(v_r-\mathrm{\&omega;}{x}_P)(n+1)T_r+2\mathrm{\&pi;}({\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">f</figure-callout>}}_0+(n+1)\mathrm{\&Delta;f})a_r(2n+1)T_r^2}{c};$

4d) take the phase place of first subpulse in echoed signal arteries and veins group as reference, the poor ΔΦ of antithetical phrase impulse compensation conjugate phase successively _n, obtain the echoed signal frequency spectrum s (f after frequency synthesis _r, T _r):

$s_n(f_r,T_r)=\widetilde{\mathrm{\&sigma;}}\mathrm{rect}\left(\frac{f_r}{B_{\mathrm{\&Delta;}}}\right)\&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0}{c}({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P))\&times;\exp (-j\frac{4\mathrm{\&pi;}(f_r+\mathrm{\&Delta;f}/2)}{c}({\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002477120200031.png,HDA00002477120200041.png"\; state="\{\{state\}\}">R</figure-callout>}}_0+y_P))$

$\&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0(v_r-\mathrm{\&omega;}{x}_P)T_r+2\mathrm{\&pi;}{f}_0{a}_r{\left(T_r\right)}^2}{c}),$

Wherein, B _Δ=N Δ f is the signal bandwidth after synthesizing.

Step 5. is obtained target high-resolution one-dimensional range profile.

To the echoed signal frequency spectrum s (f after frequency synthesis _r, T _r) carry out contrary Fourier IFFT conversion, obtain high-resolution one-dimensional range profile.

Step 6. is obtained target azimuth full resolution pricture.

Utilize RD algorithm or RID algorithm to carry out orientation imaging to high-resolution one-dimensional range profile, obtain the full resolution pricture of target azimuth.

Effect of the present invention can be illustrated by following emulation experiment:

1. emulated data

Aircraft Targets data to emulation are carried out imaging, and its simulation parameter is as shown in table 1.

The work of table 1 radar and the parameters of target motion

Radar horizon	30km	Sample frequency	250MHz
				Arteries and veins group number	128	Pulse width	10us
Subpulse number
		5	Signal bandwidth	200MHz
Pulse repetition rate					1000Hz	Target radial acceleration	6m/s
	Initial carrier frequency	10GHz	Target radial speed	120m/s
Step frequency					180MHz	Target rotational angular velocity	0.07rad/s

2. pair emulated data imaging

Emulation Aircraft Targets, Aircraft Targets is made up of 59 effective scattering points altogether, obtains the original image of Aircraft Targets, as Fig. 4 (a);

In conjunction with radar work and the parameters of target motion, the Aircraft Targets of emulation is carried out to ISAR imaging, obtain the front ISAR image of frequency synthesis of emulation Aircraft Targets, as Fig. 4 (b);

In conjunction with radar work and the parameters of target motion, the Aircraft Targets of emulation is adopted to the synthetic method ISAR imaging of step frequency of the present invention, obtain the ISAR image of emulation Aircraft Targets, as Fig. 4 (c).

3. pair measured data imaging

Each arteries and veins group of known actual measurement Ship Target data is made up of the subpulse of 31 frequency step;

The optical imagery of known Ship Target, as Fig. 5 (a);

Utilize actual measurement Ship Target data to carry out ISAR imaging, obtain the front ISAR image of frequency synthesis of Ship Target, as Fig. 5 (b);

By the method that step frequency of the present invention is synthetic, actual measurement Ship Target data are carried out to ISAR imaging, obtain surveying the frequency synthesis ISAR image of Ship Target, as Fig. 5 (c);

Adopt as seen the synthetic method of step frequency of the present invention than directly carrying out ISAR imaging by Fig. 5 (b), 5 (c), can obtain more high-resolution target ISAR image, further prove high efficiency and the validity of method proposed by the invention.

Claims (3)

1. the step frequency ISAR formation method based on phase drying and other treatment, comprises the steps:

(1) set up step frequency ISAR echo signal model;

(2) carry out kinematic parameter estimation according to echo signal model:

2a) first subpulse data of different arteries and veins groups are taken out respectively, obtain echoed signal neighboring and correlative expression formula and be:

$R(\mathrm{\&tau;},m{T}_a)={\mathrm{\&sigma;}}_P\sin c\left(A(\mathrm{\&tau;}+\frac{2M}{c})\right)\&times;\exp \left(j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}{M}_1\right)\&times;\exp \left(j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}{M}_2\right)$

Wherein, τ is correlation time, and m is arteries and veins group number, 0≤m≤M-1, mT _afor the discrete representation of orientation time, σ _pfor target echo backscattering coefficient, A is relevant matches amount, and c is the light velocity, and λ is signal wavelength, M=M ₁+ M ₂, $M_1=(v_r-x_P\mathrm{\&omega;})T_a+\frac{1}{2}{a}_r{T}_a^2,$ M ₂＝T _aa _rmT _a

Wherein, v _rthe radial velocity of the relative radar motion of target, x _pbe the lateral attitude information of scattering point, ω is the rotational angular velocity of the relative radar motion of target, T _afor the repetition period between arteries and veins group, a _rit is the radial acceleration of the relative radar motion of target;

To echoed signal neighboring and correlative expression formula along orientation to doing FFT accumulation, obtaining orientation frequency domain echo signal neighboring and correlative expression formula be 2b):

$R(\mathrm{\&tau;},f_a)={\mathrm{\&sigma;}}_P\sin c\left(A(\mathrm{\&tau;}+\frac{2M}{c})\right)\&times;\sin c\left(\mathrm{\&pi;}{T}_m(f_a-\frac{{2T}_a}{\mathrm{\&lambda;}}{a}_r)\right)\&times;\exp \left(j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}{M}_1\right)$

Wherein, T _mfor subpulse repetition period, f _afor orientation is to frequency,

it is the correlation peak location of echoed signal neighboring and correlative expression formula;

Ignore the envelope variation that in M, adjacent twice return ω causes, correlation peak location can be expressed as again $(-\frac{{2v}_r{T}_a+a_r{T}_a^2}{c},\frac{{2T}_a{a}_r}{\mathrm{\&lambda;}});$

2c) by correlation peak location distance to orientation to coordinate be made as respectively

with

calculate the radial acceleration a of target _rand speed v _r:

$\left\{\begin{array}{cc}{a}_r=\frac{\mathrm{\&lambda;prf}[\tilde{n}-(\frac{\tilde{N}}{2}+1)]}{{2\mathrm{NT}}_a}& \tilde{n}=1,.....,\tilde{N}\\ v_r=-\frac{c[\tilde{m}-(\frac{\tilde{M}}{2}+1)]}{{2f}_s{T}_a}-\frac{\mathrm{\&lambda;}[\tilde{n}-(\frac{\tilde{N}}{2}+1)]}{4\tilde{N}{T}_a}& \tilde{m}=1,.....,\tilde{M}\end{array}\right.$

Wherein,

with

be respectively distance to orientation to discrete counting, prf be orientation to sample frequency, f _sfor distance is to sample frequency, T _rfor the pulse repetition time;

(3) according to radial acceleration a _rand speed v _r, utilize the phase compensation function s of following envelope cancellation _srefsubpulse envelope to echoed signal compensates,

$s_{\mathrm{sref}}=\exp (-j\frac{4\mathrm{\&pi;}}{c}(v_r{f}_r{t}_m+\frac{1}{2}{a}_r{f}_r{t}_m^2))$

Wherein, f _rfor distance is to frequency, t _mfor orientation is to the time;

(4) echoed signal that completes subpulse envelope cancellation is carried out to the frequency synthesis based on phase drying and other treatment, the echoed signal frequency spectrum s (f after being synthesized _r, T _r);

(5) to the echoed signal frequency spectrum s (f after synthetic _r, T _r) carry out contrary Fourier IFFT conversion, obtain high-resolution one-dimensional range profile;

(6) utilize RD algorithm or RID algorithm to carry out orientation imaging to high-resolution one-dimensional range profile, obtain the full resolution pricture of target azimuth.

2. the step frequency ISAR formation method based on phase drying and other treatment according to claim 1, wherein the described echoed signal to completing subpulse envelope cancellation of step (4) is carried out the frequency synthesis based on phase drying and other treatment, carries out as follows:

4a) echoed signal that completes subpulse envelope cancellation is transformed to apart from frequency domain s _n(f _r, nT _r);

$\begin{array}{c}{s}_n(f_r,n{T}_r)=\widetilde{\mathrm{\&sigma;}}\mathrm{rect}\left(\frac{f_r}{B}\right)\&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0}{c}(R_0+y_P))\&times;\exp (-j\frac{4\mathrm{\&pi;}(f_r+\mathrm{n\&Delta;f})}{c}(R_0+y_P))\\ \&times;\exp (-j\frac{4\mathrm{\&pi;}(f_0+\mathrm{n\&Delta;f})(v_r-{\mathrm{\&omega;x}}_P){\mathrm{nT}}_r+2\mathrm{\&pi;}(f_0+\mathrm{n\&Delta;f})a_r{\left({\mathrm{nT}}_r\right)}^2}{c})\end{array}$

Wherein, f _rfor distance is to frequency, nT _rrepresent the orientation time, for range coefficient, B is transmitting subpulse signal bandwidth, and c is the light velocity, R ₀for target is to the initial action distance of radar, y _pthe longitudinal position information of scattering point P, f ₀+ n Δ f is the carrier frequency of n frequency modulation stepping subpulse, and Δ f is number of frequency steps;

4b) adjacent subpulse frequency domain echo signal location and phase place are changed, variable quantity is Δ f/2, obtains the poor ΔΦ of conjugate phase that changes frequency domain echo signal between rear adjacent subpulse _n:

${\mathrm{\&Delta;\&Phi;}}_n=\frac{4{\mathrm{\&pi;T}}_r(f_0+\mathrm{n\&Delta;f})(v_r-\mathrm{\&omega;}{x}_P)+\mathrm{\&Delta;f}(v_r-{\mathrm{\&omega;x}}_P)(n+1)T_r+2\mathrm{\&pi;}(f_0+(n+1)\mathrm{\&Delta;f})a_r(2n+1)T_r^2}{c}$

Wherein, T _rfor the pulse repetition time, v _rbe the radial velocity of the relative radar motion of target, ω is the rotational angular velocity of the relative radar motion of target, x _pthe lateral attitude information of scattering point, a _rit is the radial acceleration of the relative radar motion of target;

4c) take the phase place of first subpulse in echoed signal arteries and veins group as reference, the poor ΔΦ of antithetical phrase impulse compensation conjugate phase successively _n, obtain the echoed signal frequency spectrum s (f after frequency synthesis _r, T _r):

$\begin{array}{c}s(f_r,T_r)=\widetilde{\mathrm{\&sigma;}}\mathrm{rect}\left(\frac{f_r}{B_{\mathrm{\&Delta;}}}\right)\&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0}{c}(R_0+y_P))\&times;\exp (-j\frac{4\mathrm{\&pi;}(f_r+\mathrm{\&Delta;f}/2)}{c}(R_0+y_P))\\ \&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0(v_r-{\mathrm{\&omega;x}}_P)T_r+2\mathrm{\&pi;}{f}_0{a}_r{\left(T_r\right)}^2}{c})\end{array}$

Wherein, B _Δ=N Δ f is the signal bandwidth after synthesizing.

3. the step frequency ISAR formation method based on phase drying and other treatment according to claim 2, wherein step

4b) described subpulse frequency domain echo signal location and phase place to adjacent changes, and undertaken by following formula:

$\begin{array}{c}{s}_n(f_r+\mathrm{\&Delta;f}/2,n{T}_r)=\widetilde{\mathrm{\&sigma;}}\mathrm{rect}\left(\frac{f_r+\mathrm{\&Delta;f}/2}{B}\right)\&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0}{c}(R_0+y_P))\\ \&times;\exp (-j\frac{4\mathrm{\&pi;}(f_r+\mathrm{n\&Delta;f}+\mathrm{\&Delta;f}/2)}{c}(R_0+y_P))\\ \&times;\exp (-j\frac{4\mathrm{\&pi;}(f_0+\mathrm{n\&Delta;f})(v_r-{\mathrm{\&omega;x}}_P){\mathrm{nT}}_r+2\mathrm{\&pi;}(f_0+\mathrm{n\&Delta;f})a_r{\left({\mathrm{nT}}_r\right)}^2}{c})\end{array}$

$\begin{array}{c}{s}_{n+1}(f_r-\mathrm{\&Delta;f}/2,(n+1)T_r)=\widetilde{\mathrm{\&sigma;}}\mathrm{rect}\left(\frac{f_r-\mathrm{\&Delta;f}/2}{B}\right)\&times;\exp (-j\frac{4\mathrm{\&pi;}{f}_0}{c}(R_0+y_P))\\ \&times;\exp (-j\frac{4\mathrm{\&pi;}(f_r+\mathrm{n\&Delta;f}+\mathrm{\&Delta;f}/2)}{c}(R_0+y_P))\\ \&times;\exp (-j\frac{4\mathrm{\&pi;}(f_0+(n+1)\mathrm{\&Delta;f})(v_r-{\mathrm{\&omega;x}}_P)(n+1)T_r+2\mathrm{\&pi;}(f_0+(n+1)\mathrm{\&Delta;f})a_r{\left((n+1)T_r\right)}^2}{c})\end{array}$

Wherein, s _n(f _r+ Δ f/2, nT _r) be the subpulse frequency domain echo signal after n position and phase place change, s _n+1(f _r-Δ f/2, (n+1) T _r) be the subpulse frequency domain echo signal after n+1 position and phase place change, f _rscope be [B _c/ 2～B _c/ 2], B _cfor shared signal bandwidth, nT _rrepresent the orientation time, f ₀for fundamental frequency, exp is the exponential function truth of a matter, and rect is rectangular function.

Images