CN104502906B - Spatial ultrahigh-speed maneuvered target detection method based on RMDCFT (Radon-Modified Discrete Chirp-Fourier Transform) - Google Patents

Abstract

The invention relates to a spatial ultrahigh-speed maneuvered target detection method based on RMDCFT (Radon-Modified Discrete Chirp-Fourier Transform). According to the method, firstly, modification is carried out based on the conventional Radon-Fourier transform method by using prior information of spatial target motion parameters, searching is carried out in a known target motion parameter range, and thus, target echo signals are extracted along the motion trajectory of a target; meanwhile, modification is also carried out based on the conventional MDCFT method, transform processing is carried out on the extracted target echo signals, and thus, the estimation on the frequency modulation rate of the target can be obtained; secondary phase of echoes is compensated based on the estimation, coherent accumulation is carried out on the target echo signals, and target detection is realized. The method has no need of predicting or estimating the fuzzy number of speed of the target and is applicable to the circumstances of limited CPI (Coherent Processing Time) and low signal-to-noise ratio; simulation results are compared with processing results of MTD, Radon-Fourier transform, FRFT (Fractional Fourier Transform) and RFRFT (Radon-FRFT) detection methods, experimental results show that the method provided by the invention can be used for effectively detecting the target when the CPI is limited and the signal-to-noise ratio is relatively low, and thus, the effectiveness of the method provided by the invention is proved.

Description

Space ultrahigh speed maneuvering target detection method based on rmdcft

Technical field

The invention belongs to Radar Signal Processing and maneu-vering target detection technical field, more particularly to one kind are based on radon- Revise the space ultrahigh speed maneuvering target detection method of Discrete Linear frequency modulation Fourier transformation (rmdcft).

Background technology

Extraterrestrial target includes satellite in space, guided missile, space junk etc., recently as extraterrestrial target quantity drastically Increase, more serious threat is all caused to the transmitting of aircraft, in-orbit safe operation and Homeland air defense, so to space mesh Mark is detected, thus finding the extraterrestrial target to attack early, and provides longer time just to become particularly important for early warning.My god Base radar, using satellite as platform, can provide possible, it has for the realization of the function such as Space Object Detection and deep space imaging Monitor that scope is wide, real-time, operation effectiveness and survival ability high the features such as, and existing radar system can be detected no The deep and broad scope that method reaches, is not limited by landform and area, so space based radar will become following surveillance radar development Inexorable trend.Only the U.S. and Russia have the space surveillance network of comparatively perfect at present, visit with to extraterrestrial target Survey task increasingly urgent, the space surveillance network of building national independence will become further important.

The extraterrestrial target that space based radar detects generally has higher motion speed and acceleration, and such as ballistic missile is in powered phase Under the huge thrust of rocket, speed can reach 7000m/s, and space junk speed then can reach 10km/s；Short distance is absolutely empty to lead The maximum overload of bullet has exceeded 50g, and medium-range air-to-air missile acceleration can reach 20g.The fortune of the extraterrestrial target by being detected Dynamic feature, the echo-signal of target can produce to walk across Range cell migration and Doppler within the radar Coherent processing time (cpi) Dynamic problem, simultaneously because target is in ultrahigh speed kinestate, and radar is relatively limited for the detection time of target, institute In low signal-to-noise ratio and limited cpi, how to be effectively realized to the target echo that there is range walk and Doppler walks about Signal carries out energy accumulation and has become space based radar space exploration target key issue to be solved.

There are spectral peak tracking, minimum entropy method, envelope correlation alignment schemes currently used for the method that correction distance is walked about；By In these three methods, the signal to noise ratio of target echo signal is required higher, so the weak echo signal receiving in space based radar In the case of be difficult to obtain preferable envelope alignment effect.And when target has acceleration, keystone alternative approach then cannot Correct the range curvature phenomenon brought due to acceleration.

And radon- Fourier transformation method passes through detection range and speed, target echo can be taken out along target trajectory Data simultaneously does coherent accumulation, backward energy effectively can be accumulated when target echo has range walk phenomenon.The party Method is equally applicable in the case that signal to noise ratio is relatively low, but conventional radon- Fourier transformation method is only used for detecting uniform motion Target.

For solving the method such as wigner-ville location mode that Doppler walks about, because it is a non-linear fortune Calculate, and there is the interference of cross term, be difficult to obtain therefore under conditions of low signal-to-noise ratio and preferably detect performance.In addition, point Number rank Fourier transformations (frft) and radon-frft (rfrft) be two kinds of phenomenons of walking about for compensating for doppler method, but When having range walk, accumulation umber of pulse limited situation, to compensate Performance comparision limited, so frft and rfrft can not have for frft Extraterrestrial target is detected to effect.

And revise Discrete Linear frequency modulation Fourier transformation (mdcft) method and can eliminate conventional discrete linear frequency modulation Fourier Conversion (dcft) method is improved for the restriction of sampling number and chirp parameter, therefore practicality, but in extraterrestrial target speed When degree is higher, the mid frequency of echo-signal can exceed mdcft mid frequency not fuzzy ranges, so being unfavorable for subsequently to target The integration detection of energy, thus cannot be directly used to the detection of space ultrahigh speed maneuvering target.And often not yet occur at present It is modified on the basis of the radon- Fourier transformation of rule and mdcft method and the outstanding advantage connected applications of two methods exist The correlational study achievement of space ultrahigh speed maneu-vering target detection technical elements.

Content of the invention

In order to solve the above problems, it is an object of the invention to provide a kind of motor-driven mesh of space ultrahigh speed based on rmdcft Mark detection method.

In order to achieve the above object, the space ultrahigh speed maneuvering target detection method bag based on rmdcft that the present invention provides Include the following step carrying out in order:

1) total echo-signal that space based radar is received is demodulated and process of pulse-compression；

2) using rmdcft, the target velocity obtaining in priori, distance are converted to the signal after above-mentioned demodulation pulse compression Obtain maneuvering target echo-signal with scanning for target echo signal in the range of frequency modulation rate, then in above-mentioned basis of signals On carry out corresponding filtering transformation process, and estimate to obtain the initial velocity of target according to peak in distance-speed domain, with When obtain target range and the estimated value of frequency modulation rate；

3) in step 2) on the basis of determine the scope of maneuvering target echo-signal mid frequency, target range and frequency modulation rate, And with rmdcft conversion, target echo signal is scanned in the scope of above-mentioned determination, that is, in step 2) the motor-driven mesh that obtains It is filtered conversion process again on the basis of mark echo data, in domain, mesh is obtained according to peak in frequency modulation rate-mid frequency Target frequency modulation rate essence estimated value；

4) utilize step 3) in estimate that the smart estimated value of target frequency modulation rate obtaining is carried out to the quadratic phase of echo-signal Compensate, and then carry out coherent accumulation along target trajectory on the basis of the just conventional radon- Fourier transformation method of repairing, so that Echo-signal energy is effectively accumulated, and is eventually used for follow-up CA-CFAR detection.

In step 2) in, described converts, using rmdcft, the mesh obtaining in priori to the signal after demodulation pulse compression In the range of mark speed, distance and frequency modulation rate, target echo signal is scanned for and obtain maneuvering target echo-signal, Ran Hou Corresponding filtering transformation process is carried out on above-mentioned basis of signals, and estimates to obtain target according to peak in distance-speed domain Initial velocity, the method simultaneously obtaining the estimated value of target range and frequency modulation rate is:

Total echo-signal that space based radar receives is:

$s_r(\hat{r},t_m)=a_r\sin c\[\frac{\mathrm{\π}(\hat{r}-r_0-v_0{t}_m-\frac{1}{2}{a}_0{t}_m^2)}{{\mathrm{\ρ}}_r}\]\exp (-j\frac{4\mathrm{\π}}{\mathrm{\λ}}{r}_0)\exp \{-j\frac{2\mathrm{\π}}{m}\[\left(\frac{l_0}{m}\right)m^2+k_{0}m\]\}---\left(1\right)$

A in formula (1)_rIt is the amplitude of echo-signal, t'=t-mt_rRepresent the fast time,t_rWhen representing that pulse repeats Between, t_m=mt_rRepresent the slow time, m=1 ..., m, wherein m represent the umber of pulse in the Coherent processing time, f_cRepresent carrier frequency Rate,For signal wavelength, b_sFor signal bandwidth,r₀Represent the initial distance of target, v₀Represent the initial of target Speed, a₀Represent the acceleration of target；Mid frequency for echo-signal；It is echo-signal Frequency modulation rate；

Echo-signal in range gate-slow time domain for the target is taken out one by one, then carries out on this basis for the first time The Matched Fourier Transform revised, correspondingly mates speed and the frequency modulation rate of target, according to peak value position in distance-speed domain Put the estimated value obtaining target initial velocity, and obtain target range and the estimated value of frequency modulation rate simultaneously.

In step 3) in, described in step 2) on the basis of determine maneuvering target echo-signal mid frequency, target range With the scope of frequency modulation rate, and in the scope of above-mentioned determination with rmdcft conversion target echo signal is scanned for, that is, in step It is filtered conversion process again, root in frequency modulation rate-mid frequency domain on the basis of the maneuvering target echo data that rapid 2) obtain The method obtaining the frequency modulation rate essence estimated value of target according to peak is:

Using step 2) in obtain target initial velocity, target range and adjust frequency estimation, determine in target echo The scope of frequency of heart, distance and frequency modulation rate, then in this scope, target echo data is carried out mid frequency, frequency modulation rate and away from From smart search process, do the Matched Fourier Transform of second correction on the basis of search obtains target echo data, this When Matched Fourier Transform mated be target mid frequency and frequency modulation rate, according to peak in mid frequency-tune frequency domain Value position obtains the frequency modulation rate essence estimated value of target.

In step 4) in, described utilization step 3) in estimate the smart estimated value of the target frequency modulation rate obtaining to echo-signal Quadratic phase compensate, and then be concerned with along target trajectory on the basis of the just conventional radon- Fourier transformation method of repairing Accumulation, so that echo-signal energy is effectively accumulated, is eventually used for the method that follow-up CA-CFAR detects It is:

With step 3) in obtain target echo signal frequency modulation rate essence estimated result compensate echo-signal quadratic phase, so After carry out coherent accumulation process, carry out CA-CFAR detection after coherent accumulation target energy, thus judging that target has No, complete to detect.

The present invention makes full use of the prior information of maneuvering target kinematic parameter in space, becomes in conventional radon- Fourier Change with mdcft method on be modified, by the outstanding advantage connected applications of two methods it is proposed that a kind of sky based on rmdcft Between ultrahigh speed maneuvering target detection method；First with the echo data after demodulated and process of pulse-compression, in target Under kinematic parameter prior information, application rmdcft conversion is processed, and reaches accurate knockout machine transient echo data and estimates mesh The purpose of mark initial velocity, target range and frequency modulation rate；Then maneuvering target echo-signal mid frequency, distance and frequency modulation are determined The scope of rate, and obtain target tune Frequency Estimation using rmdcft conversion process within this range；Finally compensate target echo letter Number quadratic phase, carries out coherent accumulation along target trajectory on the basis of the just conventional radon- Fourier transformation method of repairing, is concerned with long-pending Tired result is used for the average CFAR detection of subsequent cell.By simulation result and mtd, radon- Fourier transformation, frft, rfrft The testing result of method is compared, and it is less and can when signal to noise ratio is relatively low that this method accumulates umber of pulse in space based radar cpi Obtain good detection performance, thus demonstrating the effectiveness of this method.

Brief description

The space ultrahigh speed maneuvering target detection method flow chart based on rmdcft that Fig. 1 provides for the present invention.

Fig. 2 is the space based radar receiving data scattergram after demodulated and process of pulse-compression.

Fig. 3 is umber of pulse m=64, and signal to noise ratio is the result using mtd method, receiving data being processed during -10db Figure.

Fig. 4 is umber of pulse m=64, and signal to noise ratio is using radon- Fourier transformation method, receiving data to be entered during -10db The result figure that row is processed.

Fig. 5 is umber of pulse m=64, and signal to noise ratio is for utilizing the space superelevation based on rmdcft that the present invention provides during -10db The result figure that fast maneuvering target detection method is processed to receiving data.

Fig. 6 be false-alarm probability be 10^-4、10^-6And umber of pulse is when being 64, to mtd, radon- Fourier transformation, Each method list that result after the process of frft, rfrft and rmdcft method obtains after doing 2000 Monte Carlo experiments respectively Unit's average constant false alarm rate detection performance curve.

Specific embodiment

The space ultrahigh speed maneuvering target based on rmdcft with specific embodiment, the present invention being provided below in conjunction with the accompanying drawings Detection method is described in detail.

Set up maneuvering target echo data model:

If the ultrahigh speed maneuvering target echo-signal expression formula that space based radar receives is:

$s_r(\hat{r},t_m)=a_r\sin c\[\frac{\mathrm{\π}(\hat{r}-r_0-v_0{t}_m-\frac{1}{2}{a}_0{t}_m^2)}{{\mathrm{\ρ}}_r}\]\exp (-j\frac{4\mathrm{\π}}{\mathrm{\λ}}{r}_0)\exp \{-j\frac{2\mathrm{\π}}{m}\[\left(\frac{l_0}{m}\right)m^2+k_{0}m\]\}---\left(1\right)$

A in formula (1)_rRepresent the amplitude of echo-signal, t'=t-mt_rRepresent the fast time,t_rWhen representing that pulse repeats Between, t_m=mt_rRepresent slow time, m=1 ..., m (m represents the umber of pulse in the Coherent processing time), f_cRepresent carrier frequency,For signal wavelength, b_sFor signal bandwidth, Wherein r₀Represent the first of target Beginning distance, v₀Represent the initial velocity of target, a₀Represent the acceleration of target；Center frequency for echo-signal Rate；It is the frequency modulation rate of echo-signal.

The echo data model of ultrahigh speed maneuvering target may be considered linear FM signal, and permissible from formula (1) Find out, due to the uniformly accelerated motion of target, the echo signal envelope after process of pulse-compression not only produces range walk and also can Produce range curvature phenomenon.In space based radar cpi accumulation umber of pulse is less and low signal-to-noise ratio in the case of, because target energy is divided It is dispersed in different distance unit and doppler cells, therefore bring very big difficulty to target is detected；Based on the problems referred to above this Bright propose a kind of space ultrahigh speed maneuvering target detection method based on rmdcft.

Application rmdcft method is processed to echo data:

Method proposed by the present invention can make full use of the prior information of the parameters of target motion and rmdcft conversion to returning Wave number, according to being processed, is further compensate for echo secondary on the basis of obtaining the estimation to target initial velocity and tune frequency parameter Phase place, then carries out coherent accumulation along target trajectory on the basis of the just conventional radon- Fourier transformation method of repairing, is eventually used for CA-CFAR detects.

As shown in figure 1, the present invention provide included by suitable based on the space ultrahigh speed maneuvering target detection method of rmdcft The following step that sequence is carried out:

1) total echo-signal that space based radar is received is demodulated and process of pulse-compression；

The expression formula of echo-signal above-mentioned set up maneuvering target echo data model when made respective explanations；

2) using rmdcft, the target velocity obtaining in priori, distance are converted to the signal after above-mentioned demodulation pulse compression Obtain maneuvering target echo-signal with scanning for target echo signal in the range of frequency modulation rate, then in above-mentioned basis of signals On carry out corresponding filtering transformation process, and estimate to obtain the initial velocity of target, note according to peak in distance-speed domain ForThe estimated value of target range and frequency modulation rate can also be obtained simultaneously；

The typical kinematic parameter of known spatial target is: velocity interval typically between 1.4km/s～7km/s, acceleration Scope is typically in 19.6m/s²～190m/s²Between, therefore the velocity interval setting target is [v_min,v_max], distance range is [- r_max, r_max], it is [l according to the scope that acceleration range determines frequency modulation rate_min,l_max], the step-size in search of speed is △ v=λ/2t, distance Step-size in search be △ r=2b/c, frequency modulation rate step-size in search can be determined according to its precision, is usually chosen to 0.01；So it is right The speed search number of target data isRange search scope is In formula, round represents and rounds process；And then have a following search step-by-step movement:

$\begin{array}{c}v\left(q\right)=v_{\min }+q\mathrm{\δ}v,q=0,\mathrm{...},n_v\\ r\left(i\right)=-r_{\max }+i\mathrm{\δ}r,i=0,\mathrm{...},n_r\\ l=l_{\min }+{\mathrm{\δ}}_l,{\mathrm{\δ}}_l=0.01\end{array}---\left(2\right)$

This method is first modified, by target data on conventional radon- Fourier transformation target data searching method Search expression is written as:

$s_r\{m,round\[\frac{r\left(j\right)+v\left(q\right){\mathrm{mt}}_r+\frac{1}{2}a{\left({\mathrm{mt}}_r\right)}^2}{{\mathrm{\δ}}_r}\]\}---\left(3\right)$

A represents acceleration, due toRepresent target echo signal frequency modulation rate, then formula (3) can be rewritten as:

$s_r\{m,round\[\frac{r\left(i\right)+v\left(q\right){\mathrm{mt}}_r+\frac{1}{2}\frac{l\mathrm{\λ}}{t_r^2}{\left({\mathrm{mt}}_r\right)}^2}{{\mathrm{\δ}}_r}\]\}---\left(4\right)$

According to formula (4), by the search to target range r (i), speed v (q) and frequency modulation rate l, thus by target in distance Echo data in door-slow time domain takes out one by one.In addition, quadratic phase item occurs in the phase place of echo-signalAccording toK ＞ 0, at this moment the phase place in echo-signal formula (1) can be written as:

$\exp \left\{-j\frac{2\mathrm{\π}}{m}\[\left(\frac{l_0}{m}\right)m^2+k_{0}m\]\right\}=\exp \[-j\frac{2\mathrm{\π}}{m}\left(\frac{l_0}{m}\right)m^2\]\exp \left(-j\frac{4{\mathrm{\πv}}_0}{\mathrm{\λ}}{\mathrm{mt}}_r\right)---\left(5\right)$

According to formula (5), the matched filtering item in conjunction with conventional mdcft is modified, and is written as:

$h_{l,v\left(q\right)}^{\′}\left(m\right)=\exp \[j\frac{2\mathrm{\π}}{m}\left(\frac{l}{m}\right)m^2\]\exp \[j\frac{4\mathrm{\π}v\left(q\right)}{\mathrm{\λ}}{\mathrm{mt}}_r\]---\left(6\right)$

By the transform that formula (1), (4) and (6) defines rmdcft can be then:

$g_{rmdcft-a}\[(i,q),l\]=\underset{m=0}{\overset{m-1}{\σ}}{s}_r\{m,round\[\frac{r\left(j\right)+v\left(q\right){\mathrm{mt}}_r+\frac{1}{2}\frac{l\mathrm{\λ}}{t_r^2}{\left({\mathrm{mt}}_r\right)}^2}{{\mathrm{\δ}}_r}\]\}{h}_{l,v\left(q\right)}^{\′}\left(m\right)---\left(7\right)$

In step 2) in make full use of the kinematic parameter prior information of target, in conventional radon- Fourier transformation and It is modified on the basis of mdcft method；In the distance-speed of target and tune frequency search range, (i.e. formula (2)) use formula (7) Respective handling is carried out to echo-signal；Realize accurately taking out the echo data of maneuvering target and carrying out the mesh of matched filtering process 's；Obtain the estimated value of target initial velocity according to peak in distance-speed domain, be designated asAnd mesh can be obtained simultaneously Target distance and the estimated value of frequency modulation rate.

3) in step 2) on the basis of determine the scope of maneuvering target echo-signal mid frequency, target range and frequency modulation rate, And with rmdcft conversion, target echo signal is scanned in the scope of above-mentioned determination, that is, in step 2) the motor-driven mesh that obtains It is filtered conversion process again on the basis of mark echo data, in domain, mesh is obtained according to peak in frequency modulation rate-mid frequency Target frequency modulation rate essence estimated value；

In step 3) in, make full use of step 2) in obtain target initial velocity, target range and adjust frequency estimation, Determine the scope of target echo signal mid frequency, target range and frequency modulation rate, obtain target center frequency, frequency modulation rate above-mentioned On the basis of distance range, in order to accurately estimate the tune frequency values of echo signal further, so in target will being carried out Frequency of heart, frequency modulation rate and distance carry out smart search process, and now data search expression formula is:

$s_r\{m,round\[\frac{r\left(j\right)+\frac{k\mathrm{\λ}}{2m}m+\frac{1}{2}\frac{l\mathrm{\λ}}{t_r^2}{\left({\mathrm{mt}}_r\right)}^2}{{\mathrm{\δ}}_r}\]\}---\left(8\right)$

Now also conventional mdcft matched filtering item is modified according to phase representation in echo-signal formula (1), and writes For:

$h_{l,k}^{\′\′}\left(m\right)=\exp \{j\frac{2\mathrm{\π}}{m}\[\left(\frac{l}{m}\right)m^2+km\]\}---\left(9\right)$

Then rmdcft transform can be defined according to formula (1), (8) and (9) is:

$g_{rmdcft-b}\[(l,k),i\]=\underset{m=0}{\overset{m-1}{\σ}}{s}_r\{m,round\[\frac{r\left(i\right)+\frac{k\mathrm{\λ}}{2m}m+\frac{1}{2}\frac{l\mathrm{\λ}}{t_r^2}{\left({\mathrm{mt}}_r\right)}^2}{{\mathrm{\δ}}_r}\]\}{h}_{l,k}^{\′\′}\left(m\right)---\left(10\right)$

Definition (10) is the frequency modulation rate-mid frequency to target and distance scans for, in step 3) the middle mesh determining In the smaller range of subject distance, mid frequency and frequency modulation rate, matched filtering process is carried out to echo data, matching echo letter During number parameter, in frequency modulation rate-mid frequency plane, peak value occurs, such that it is able to estimate to obtain the tune of target according to peak Frequency.

4) utilize step 3) in estimate that the smart estimated value of target frequency modulation rate obtaining is carried out to the quadratic phase of echo-signal Compensate, and then carry out coherent accumulation along target trajectory on the basis of the just conventional radon- Fourier transformation method of repairing, so that Echo-signal energy is effectively accumulated, and is eventually used for follow-up CA-CFAR detection.

The coherent energy accumulation method that conventional radon- Fourier transformation is revised in definition is:

$g_{modified\_rft}\[(i,q),l\]=\underset{m=0}{\overset{m-1}{\σ}}{s}_r^{\′}\{m,round\[\frac{r\left(j\right)+v\left(q\right){\mathrm{mt}}_r+\frac{1}{2}\frac{l\mathrm{\λ}}{t_r^2}{\left({\mathrm{mt}}_r\right)}^2}{{\mathrm{\δ}}_r}\]\}{q}_{v\left(q\right)}\left(m\right)---\left(11\right)$

Wherein Represent the echo data after compensating already for echo quadratic phase.According to formula (11), along the target compensating echo quadratic phase Echo data is taken out by track again, is then multiplied by q_v(q)(m) sue for peace to do coherent accumulation, finally carry out cell-average permanent empty Alert detection, thus judging that target has or not, completes to detect.

Simulation result and analysis:

The effect of the space ultrahigh speed maneuvering target detection method based on rmdcft that the present invention provides can be passed through following Emulation experiment further illustrates.Simulation parameter is arranged: space based radar transmission signal wavelength is 0.3m, and pulse recurrence frequency is 1500hz, accumulation umber of pulse is 64, and signal bandwidth is 30mhz, and carrier frequency is 1ghz.Target component is set to: initial bit set In the 400th range cell, initial velocity v₀=3000m/s, acceleration a₀=180m/s².

1st, echo data scattergram

Fig. 2 is the space based radar receiving data scattergram after demodulated and process of pulse-compression.Figure it is seen that Echo data after demodulated and process of pulse-compression, because target has larger initial velocity and acceleration hence it is evident that producing Give birth to across Range cell migration phenomenon, and also with range curvature phenomenon.Because backward energy is dispersed in different distance lists In unit and doppler cells, so being difficult in the case of low signal-to-noise ratio effectively find from noise circumstance target.

2nd, application mtd method is processed to echo data

It is -10db that Fig. 3 gives in signal to noise ratio, and accumulation umber of pulse is to process after echo data with mtd method when 64 Result.From figure 3, it can be seen that because target speed is higher, and having very high acceleration, target energy is dispersed in difference Range cell and doppler cells in, the conventional mtd method of application cannot effectively accumulate target energy, in low signal-to-noise ratio situation It is difficult to down tell target from noise, therefore, it is difficult to obtaining good detection performance.

3rd, application radon- Fourier transformation method is processed to echo data

It is -10db that Fig. 4 gives in signal to noise ratio, and accumulation umber of pulse is with radon- Fourier transformation method when 64 Result after reason echo data.From fig. 4, it can be seen that because target has larger acceleration, target echo signal can produce Range curvature, backward energy is dispersed in different doppler cells；Conventional radon- Fourier transformation method can not solve Test problems with acceleration target, in the case of having range curvature problem, radon- Fourier transformation can not be exactly Target echo data taking-up is processed, thus the energy of target echo signal can not effectively be accumulated, therefore detection performance can be subject to Large effect.

4th, application rmdcft method is processed to echo data

It is -10db that Fig. 5 gives in signal to noise ratio, and accumulation umber of pulse is to process echo data with rmdcft method when 64 Result afterwards.From fig. 5, it can be seen that because rmdcft alternative approach is to conventional radon- Fourier transformation and to revise discrete lines Property frequency modulation Fourier transformation method be modified, the outstanding advantage connected applications of two methods take out band therefore, it is possible to effective There is the echo data of range curvature, thus overcome conventional radon- Fourier transformation can not solve to carry acceleration target detection Problem, simultaneously to take out echo data mate, echo-signal can be compensated on the basis of obtaining target frequency modulation rate Quadratic phase, carries out coherent accumulation to echo data further；So as to the situation in less accumulation umber of pulse and low signal-to-noise ratio Under effectively accumulate the energy of echo-signal, therefore good detection can be obtained in follow-up CA-CFAR detection Energy.

5th, CA-CFAR detection performance

Result unit after Fig. 6 gives mtd, radon- Fourier transformation, frft, rfrft, rmdcft method is processed is put down All CFAR detection performance charts.The false-alarm probability of Fig. 6 (a) is 10^-4, the false-alarm probability of Fig. 6 (b) is 10^-6.From Fig. 6 (a) As can be seen that due to the motion of target ultrahigh speed, and with acceleration so that the backward energy of target is dispersed in different distances In unit and doppler cells, conventional mtd method cannot effectively accumulate the energy of target, leads to not obtain good detection Performance, mtd method detection probability p_dWhen=0.9, signal to noise ratio is about in 4db；Conventional radon- Fourier transformation method is not Can effective process carry range curvature target echo data, thus nor effectively accumulate target energy, reduce inspection Survey performance, radon- Fourier transformation method detection probability p_dWhen=0.9, signal to noise ratio is about about -2db；Fractional order Fourier Conversion is to be processed for the data in some range cell, because target energy is dispersed in different range cells and many In general Le unit, the backward energy ratio in single range cell is relatively low, and in space based radar cpi, accumulation umber of pulse is limited, connects Receive is the relatively low echo-signal of signal to noise ratio, and such frft can not effectively estimate signal frequency modulation rate it is impossible to reach ideal Detection performance, frft method detection probability p_dWhen=0.9, signal to noise ratio is about about -5db；Although rfrft method can take Go out target echo data, but also due to Fourier Transform of Fractional Order is affected by accumulation umber of pulse, result can cause not Preferably detection performance, and the output signal-to-noise ratio of rfrft is low compared with the output signal-to-noise ratio that rmdcft converts, so that The detection performance of rfrft will be less than rmdcft；Rfrft method detection probability p_dWhen=0.9, signal to noise ratio is about about -8db；And The rmdcft method that the present invention provides is directed to range walk and range curvature problem, effectively can carry out echo signal relevant long-pending Tired, target, detection probability p can be effectively detected in the case that low signal-to-noise ratio, accumulation umber of pulse are limited_dBelieve when=0.9 Make an uproar than about about -14db.Corresponding conclusion also can obtain from Fig. 6 (b).

Claims (4)

1. a kind of space ultrahigh speed maneuvering target detection method based on rmdcft, described rmdcft represent radon- revise from Scattered linear frequency modulation Fourier transformation is it is characterised in that described method includes carrying out in order following step:

1) total echo-signal that space based radar is received is demodulated and process of pulse-compression；

2) target velocity, distance and the tune that the signal after above-mentioned demodulation pulse compression is obtained in priori using rmdcft conversion Target echo signal is scanned for and obtains maneuvering target echo-signal in frequency range, then enterprising in above-mentioned basis of signals The corresponding filtering transformation of row is processed, and estimates to obtain the initial velocity of target according to peak in distance-speed domain, simultaneously Estimated value to target range and frequency modulation rate；

3) in step 2) on the basis of determine the scope of maneuvering target echo-signal mid frequency, target range and frequency modulation rate, and With rmdcft conversion, target echo signal being scanned in the scope of above-mentioned determination, that is, in step 2) maneuvering target that obtains returns It is filtered conversion process again on the basis of wave number evidence, in domain, target is obtained according to peak in frequency modulation rate-mid frequency Frequency modulation rate essence estimated value；

4) utilize step 3) in estimate that the smart estimated value of target frequency modulation rate obtaining compensates to the quadratic phase of echo-signal, And then carry out coherent accumulation along target trajectory on the basis of the just conventional radon- Fourier transformation method of repairing, so that echo letter Number energy is effectively accumulated, and is eventually used for follow-up CA-CFAR detection.

2. the space ultrahigh speed maneuvering target detection method based on rmdcft according to claim 1 it is characterised in that: Step 2) in, described to the signal after demodulation pulse compression using rmdcft convert the target velocity obtaining in priori, away from From with frequency modulation rate in the range of target echo signal scanned for and obtain maneuvering target echo-signal, then in above-mentioned signal base Corresponding filtering transformation process is carried out on plinth, and estimates to obtain the initial velocity of target according to peak in distance-speed domain, The method simultaneously obtaining the estimated value of target range and frequency modulation rate is:

Total echo-signal that space based radar receives is:

$s_r(\hat{r},t_m)=a_r\sin c\[\frac{\mathrm{\π}(\hat{r}-r_0-v_0{t}_m-\frac{1}{2}{a}_0{t}_m^2)}{{\mathrm{\ρ}}_r}\]\exp (-j\frac{4\mathrm{\π}}{\mathrm{\λ}}{r}_0)\exp \{-j\frac{2\mathrm{\π}}{m}\[\left(\frac{l_0}{m}\right)m^2+k_{0}m\]\}---\left(1\right)$

A in formula (1)_rIt is the amplitude of echo-signal, t '=t-mt_rRepresent the fast time,t_rRepresent pulse-recurrence time, t_m =mt_rRepresent the slow time, m=1 ..., m, wherein m represent the umber of pulse in the Coherent processing time, f_cRepresent carrier frequency,For signal wavelength, b_sFor signal bandwidth,r₀Represent the initial distance of target, v₀Represent the initial speed of target Degree, a₀Represent the acceleration of target；Mid frequency for echo-signal；It is echo-signal Frequency modulation rate；

Echo-signal in range gate-slow time domain for the target is taken out one by one, then carries out on this basis revising for the first time Matched Fourier Transform, i.e. correspondingly the coupling speed of target and frequency modulation rate, obtained according to peak in distance-speed domain To the estimated value of target initial velocity, and obtain target range and the estimated value of frequency modulation rate simultaneously.

3. the space ultrahigh speed maneuvering target detection method based on rmdcft according to claim 1 it is characterised in that: Step 3) in, described in step 2) on the basis of determine maneuvering target echo-signal mid frequency, target range and frequency modulation rate Scope, and with rmdcft conversion, target echo signal is scanned in the scope of above-mentioned determination, that is, in step 2) obtain It is filtered conversion process again, according to peak in frequency modulation rate-mid frequency domain on the basis of maneuvering target echo data The method obtaining the frequency modulation rate essence estimated value of target is:

Using step 2) in obtain target initial velocity, target range and adjust frequency estimation, determine target echo center frequency The scope of rate, distance and frequency modulation rate, then carries out mid frequency, frequency modulation rate and distance in this scope to target echo data Smart search process, does the Matched Fourier Transform of second correction on the basis of search obtains target echo data, now Join that Fourier transformation mated is mid frequency and the frequency modulation rate of target, according to peak value position in mid frequency-tune frequency domain Put the frequency modulation rate essence estimated value obtaining target.

4. the space ultrahigh speed maneuvering target detection method based on rmdcft according to claim 1 it is characterised in that: Step 4) in, described utilization step 3) in estimate the smart estimated value of the target frequency modulation rate the obtaining quadratic phase to echo-signal Compensate, and then carry out coherent accumulation along target trajectory on the basis of the just conventional radon- Fourier transformation method of repairing, thus Echo-signal energy is effectively accumulated, the method being eventually used for follow-up CA-CFAR detection is:

With step 3) in obtain target echo signal frequency modulation rate essence estimated result compensate echo-signal quadratic phase, Ran Houjin Row coherent accumulation is processed, and carries out CA-CFAR detection after coherent accumulation target energy, thus judging that target has or not, complete Become detection.