CN102928837A - Space spinning object imaging method based on single range matched filtering (SRMF) and sequence CLEAN - Google Patents

Abstract

The invention relates to a space spinning object imaging method based on single range matched filtering (SRMF) and a sequence CLEAN. According to the method, a coherence CLEAN is replaced by the sequence CLEAN, and the sequence CLEAN is combined with the SRMF, so that the influence of a false scattering point can be effectively eliminated. As the maximum m peak values which are found by using the sequence CLEAN every time can form an m-tree, and before performing the next step, the peak values which are found every time can be used for judging whether the scattering point is a real target scattering point, so that the interference of the false scattering point and noise can be effectively eliminated. According to the simulation result, all scattering points can be accurately found by using the method.

Description

A kind of space spin target imaging method based on SRMF and sequence C LEAN

Technical field

The invention belongs to the radar imagery technical field, relate to the design of space spin target imaging method, be specifically related to a kind of space spin target imaging method based on SRMF and sequence C LEAN.

Background technology

Along with the deep exploitation of the mankind to space circle, the quantity of space junk sharply increases, and while its high-speed motion (being generally 8km/s) is if collide with it and will cause serious threat to the normal operation of spacecraft at the rail spacecraft.Therefore it is carried out detection and Identification becomes an important task.Based on the fact that space debris is carried out simple spin motion around its main shaft, main formation method has at present: single range Doppler is interfered (SRDI), and is single apart from matched filtering (SRMF) and SRMF-CLEAN.

The key step of SRMF-CLEAN method imaging is as follows:

(1) at first utilize the Fourier pair echo data to carry out distance obtains data from single range unit to compression;

(2) adopt SRMF that single range unit data are processed, obtain width of cloth two dimension complex pattern;

(3) extract the maximum corresponding position of peak value and range parameter in the top complex pattern, and utilize this parametric configuration point spread function;

(4) estimate this locational reflection coefficient on the basis of previous step;

(5) deduct this scattering point to the contribution of echo data, with subduing signal afterwards as new input value;

(6) repeating step (2) ~ step (5) is until all scattering points all find or signal energy reaches noise level.

Because the SRMF-CLEAN method is responsive and easily be subjected to the impact of false scattering point on noise ratio, utilize the positional information deviation of the scattering point that the method proposes larger.That (the m value is larger, and the scattering point degree of accuracy that finds is higher for m peak value and sequence C LEAN searches at every turn; But m value too conference causes the side's of fortune speed slack-off), and utilize energy variation before and after subduing to judge whether real impact point of this peak value.

Summary of the invention

The technical matters that solves

For fear of the deficiencies in the prior art part, the present invention proposes a kind of space spin target imaging method based on SRMF and sequence C LEAN.

Technical scheme

A kind of space spin target imaging method based on SRMF and sequence C LEAN is characterized in that step is as follows:

Step 1: according to the distance of simulation objectives model to the orientation to the structure echoed signal;

Step 2: utilize the Fourier pair echoed signal to carry out Range compress and obtain signal S _k(x), wherein k represents iterations, and utilizes formula Calculate the energy of the rear signal of compression, wherein * represents conjugation;

Step 3: utilize the SRMF method that each range unit in the echoed signal is processed, obtain two-dimentional complex pattern I _k

Step 4: extract two-dimentional complex pattern I _kThe m of a middle maximum peak value, and record coordinate corresponding to each peak value and amplitude information;

Step 5: utilize coordinate and the amplitude information structure point spread function of i peak value, i=1,2 ... m; Use S _k(x) deduct the point spread function of this point at correspondence position, obtain the signal S of i peak value choosing in the k time iterative process _{K, i}

Step 6: calculate signal S _{K, i}Corresponding energy $T_i(k+1)={\∫}_{-\∞}^{\∞}{S}_{(k,i)}\left(x\right)S_{(k,i)}^{*}\left(x\right)\mathrm{dx},$ Wherein * represents conjugation;

Step 7: compare T _i(k+1) with T (k), if the former thinks that greater than latter the point of subduing in the step 5 is false target, T relatively _i(k+1) with T (k), if the former thinks that greater than latter the point of subduing in the step 5 is false target, stop the iteration of this point; Otherwise then store position and the amplitude information of this point, then make k=k+1, with S _{K, i}As new echoed signal, repeating step 2-5;

Step 8: to m peak value repeating step 4 ~ step 6;

Step 9: when target has all found or reaches noise level, then stop, obtain an incomplete m fork tree, find that node of signal energy minimum, then from this node searching route that makes progress, and store all nodes corresponding to this path;

The target imaging of " totally " signal that step 10, usefulness " totally " point spread function and these convolution obtain restoring.

Beneficial effect

A kind of space spin target imaging method based on SRMF and sequence C LEAN that the present invention proposes utilizes sequence C LEAN to replace relevant CLEAN and SRMF to combine, and utilizes the method can effectively solve the impact of false scattering point.Because what sequence C LEAN searched is m maximum peak value at every turn, can consist of a m fork tree.And find peak value at every turn can judge this scattering point see whether belong to real target scattering point before carrying out next step.Can effectively solve like this interference of false scattering point and noise.Find by simulation result, the method can find all scattering points more exactly.

Description of drawings

Fig. 1: radar spin object module

Fig. 2: simulated point object module

Fig. 3: SRMF imaging effect figure

Fig. 4: SRMF-SRMF method imaging effect figure

Fig. 5: the imaging effect figure of the inventive method

Embodiment

Now in conjunction with the embodiments, the invention will be further described for accompanying drawing:

Figure a radar spin object module, R is the radar line of sight direction, and Z axis is target rotating axle, and P is any scattering point on the target, and P ' is the subpoint of some P on the radar imagery plane.

1) according to 4 some structure echoed signal shown in the figure b;

2) signal that above-mentioned echoed signal is carried out behind the Range compress is S _k(x) (k represents the number of times of iteration) and calculate the energy of this signal:

$T_i\left(k\right)={\∫}_{-\∞}^{\∞}{S}_k\left(x\right)S_k^{*}\left(x\right)\mathrm{dx}$

3) set up matched signal for radius one by one, the signal behind itself and the Range compress all is transformed on the frequency domain, then both being multiplied each other to obtain two-dimentional complex pattern I _k

4) search image I _k4 peak values of middle maximum, and record position and the amplitude information of each peak value;

5) utilize i (i=1,2,3, the 4) relevant information of individual peak value to construct the point spread function of this point, and deduct this point spread function at the signal of compression, thereby obtain new signal S _{(k, i)}

6) energy of calculating new signal:

$T_i(k+1)={\∫}_{-\∞}^{\∞}{S}_{(k,i)}\left(x\right)S_{(k,i)}^{*}\left(x\right)\mathrm{dx}$

7) compare T _i(k+1) with T (k), if the former thinks the false target that is that this subdues greater than latter, stop the iteration of this point; Otherwise store position and the amplitude of this point, then make k=k+1, with S _{(k, i)}As new input signal, repeating step 1)-5);

8) to m signal repeating step 5)-8);

9) if 4 targets all find, termination of iterations process then.Can obtain incomplete 4 fork trees this moment, find that node of signal energy minimum, then from this node searching route that makes progress, and store all nodes corresponding to this path;

10) " totally " signal that obtains restoring with " totally " point spread function and these convolution.

The image interpretation explanation:

Fig. 3 is SRMF algorithm imaging effect figure, as seen from the figure among the higher and figure of secondary lobe because the secondary lobe stack has formed higher false scattering point, directly affect the identification of emulation scattering point; Fig. 4 is the imaging effect figure with the SRMF-CLEAN algorithm, there is larger deviation in the position of the scattering point that finds as seen from the figure and emulation scattering point, because this algorithm is searched maximum peak value at every turn, do not judge whether this impact point belongs to the real goal scattering point; Fig. 5 is the imaging effect figure of algorithm of the present invention, owing to this algorithm after finding peak point it is judged see whether belong to the real goal scattering point, and this algorithm can find the scattering point of emulation more exactly as seen from the figure.

Claims (1)

1. one kind based on the space of SRMF and sequence C LEAN spin target imaging method, it is characterized in that step is as follows:

Step 1: according to the distance of simulation objectives model to the orientation to the structure echoed signal;

Step 2: utilize the Fourier pair echoed signal to carry out Range compress and obtain signal S _k(x), wherein k represents iterations, and utilizes formula

Calculate the energy of the rear signal of compression, wherein * represents conjugation;

Step 3: utilize the SRMF algorithm that each range unit in the echoed signal is processed, obtain two-dimentional complex pattern I _k

Step 4: extract two-dimentional complex pattern I _kThe m of a middle maximum peak value, and record coordinate corresponding to each peak value and amplitude information;

Step 5: utilize coordinate and the amplitude information structure point spread function of i peak value, i=1,2 ... m; Use S _k(x) deduct the point spread function of this point at correspondence position, obtain the signal S of i peak value choosing in the k time iterative process _{K, i}

Step 6: calculate signal S _{K, i}Corresponding energy $T_i(k+1)={\∫}_{-\∞}^{\∞}{S}_{(k,i)}\left(x\right)S_{(k,i)}^{*}\left(x\right)\mathrm{dx},$ Wherein * represents conjugation;

Step 7: compare T _i(k+1) with T (k), if the former thinks that greater than latter the point of subduing in the step 5 is false target, T relatively _i(k+1) with T (k), if the former thinks that greater than latter the point of subduing in the step 5 is false target, stop the iteration of this point; Otherwise then store position and the amplitude information of this point, then make k=k+1, with S _{K, i}As new echoed signal, repeating step 2-5;

Step 8: to m peak value repeating step 4 ~ step 6;

Step 9: when target has all found or reaches noise level, then stop, obtain an incomplete m fork tree, find that node of signal energy minimum, then from this node searching route that makes progress, and store all nodes corresponding to this path;

The target imaging of " totally " signal that step 10, usefulness " totally " point spread function and these convolution obtain restoring.

Images