CN105676219A - Quadrature-phase modulation-based MIMO radar three-dimensional imaging method - Google Patents

Abstract

The invention discloses a quadrature-phase modulation-based MIMO radar three-dimensional imaging method. According to the method, firstly, an orthogonal phase coding sequence of a corresponding class number is constructed according to the number of transmitting array elements. Secondly, received echo signals are filtered correspondingly, so that the corresponding echo signals of different transmitting array elements can be separated. Thirdly, an entire three-dimensional imaging area is divided, and each of all pixel points in the imaging area is traversed sequentially. In this way, the amplitude value corresponding to the time delay interval point of all transmitting array elements that corresponds to each pixel point can be obtained. Finally, the pixel information of all pixel points in the three-dimensional imaging area and the amplitude values thereof together form a three-dimensional matrix. The three-dimensional matrix is drawn according to the actual size ratio of the imaging area, so that the information of a three-dimensional space target can be obtained. According to the technical scheme of the invention, the three-dimensional target of a ground-penetrating radar can be imaged while the information of the three-dimensional target can be transmitted and received at the same time.

Description

Based on the MIMO radar three-D imaging method of orthogonal phase modulation

Technical field

The present invention relates to GPR Imaging technical field, it is specifically related to a kind of MIMO radar three-D imaging method based on orthogonal phase modulation.

Background technology

GPR utilizes to launch hertzian wave, obtains the radar system of the unknown target information in shallow top layer, underground. Owing to utilizing high resolution three-dimensional imaging algorithm can know the hiding target sizes in different face in three-dimensional space, the information such as position, thus it is widely used in aspects such as hitting anti-terrorism, medical diagnosis, disaster rescue, safety guard.

MIMO radar is the radar of multiple-input and multiple-output, according to the multiple transmitting antenna of layout and receiving antenna, and launching between signal orthogonal, the different array elements of receiving antenna carry out different matched filterings according to target scatter echo, it is achieved to the separation of different signal. Compared with traditional phased array radar, MIMO radar can significantly be improved parameter differentiability, have better resolving power, and higher sensitivity receives domestic and international concern. But, owing to data acquiring mode and traditional synthetic-aperture radar (syntheticapertureradar, SAR) of MIMO radar are different so that conventional radar imaging method is difficult to directly be applied in MIMO radar imaging. Such as the distance doppler (rangeDoppler in synthetic aperture imaging algorithm, RD) algorithm, distance offset (rangemigration, RM) algorithm is all based on transmitting-receiving with the even spatial sampling put, and these are all difficult to the echoed signal of the many transmitting-receivings of directly process.

Summary of the invention

The problem being existing three-dimensional imaging algorithm and being difficult to the echoed signal of the many transmitting-receivings of directly process to be solved by this invention, thering is provided a kind of MIMO radar three-D imaging method based on orthogonal phase modulation, it can realize launching the GPR objective imaging receiving simultaneously and obtaining objective information simultaneously.

For solving the problem, the present invention is achieved by the following technical solutions:

Based on a MIMO radar three-D imaging method for orthogonal phase modulation, comprise the following steps:

Step 1, meets, according to the design of the size in three-dimensional imaging region, the plane array required, and the transmitting array element of given plane array and reception array element; Each output terminal receiving array element is respectively connected to the quantity matched filter identical with the number launching array element, and each matched filter launches array element corresponding to one;

Step 2, constructs the orthogonal PSK sequence of respective sets number according to transmitting element number of array, and each launches the corresponding one group of orthogonal PSK sequence of array element;

Step 3, receives array element and receives echoed signal, and the echoed signal received is sent to matched filter and carries out relevant filtering, thus realizes the different separation launching echoed signal corresponding to array element;

Step 4, divides whole three-dimensional imaging region, travels through the range value of echoed signal corresponding to the time delay spacing point place of the transmitting-receiving array element of each pixel of imaging region successively; The range value of the echoed signal that the time delay spacing point of all transmitting-receiving array elements that cumulative this pixel of summation is corresponding is corresponding successively;

Step 5, coordinate information and the required range value of whole three-dimensional imaging space pixel can form three-dimensional matrice, drawn according to the actual size of imaging region by three-dimensional matrice, so that it may obtain three-dimensional space target information.

It is in above-mentioned steps 2, as follows according to the process launching the orthogonal PSK sequence that element number of array L constructs respective sets number,

Step 2.1, the number of phases M of setting code length N and encoding phase;

Step 2.2, random generation T group coding matrix { s_l(n) }, the line number of every group coding matrix is L, and row number is N, and element is s_l(n), wherein

s_l(n)=exp [j ψ_l(n)]

Wherein, ${\mathrm{\ψ}}_l\left(n\right)\∈\{0,\frac{2\mathrm{\π}}{M},2\·\frac{2\mathrm{\π}}{M},...,(M-1)\·\frac{2\mathrm{\π}}{M}\};$ M is the number of phases of encoding phase; N=1,2 ..., N, N are code length; L=1,2 ..., L, L are for launching element number of array;

Step 2.3, for each group coding matrix, with behavior unit, calculate autocorrelation sidelobe energy A (l) of every a line of this group coding matrix and the cross-correlation energy B (p of every two row, q), and calculate the target function value W of this group coding matrix accordingly:

$W={\mathrm{\λ}}_1\×{\mathrm{\Σ}}_{l=1}^{L}A\left(l\right)+{\mathrm{\λ}}_2\×{\mathrm{\Σ}}_{p=1}^{L-1}{\mathrm{\Σ}}_{q=p+1}^{L}B(p,q)$

Wherein, λ₁For auto-correlation weighting coefficient, λ₂For cross-correlation weighting coefficient, 0 < λ₁< 1,0 < λ₂< 1, λ₁+λ₂=1; L=1,2 ..., L; P=1,2 ..., L; Q ,=1,2 ..., L;

Step 2.4, retains that minimum for target function value W group coding matrix, and according to genetic algorithm, all the other T-1 group coding matrixes is carried out cross and variation, thus forms the encoder matrix of new T group;

Step 2.5, repeating step 2.3-2.4, until when target function value W is less than goal-selling threshold value or iteration number of times reaches default iteration threshold, then the every a line of the encoder matrix that group target function value W this iteration retained is minimum respectively launches the transmitting signal of array element as one.

In above-mentioned steps 2.2, the group number T of the random encoder matrix produced should be greater than the number L launching array element.

In above-mentioned steps 2.3, auto-correlation weighting coefficient λ₁It is greater than cross-correlation weighting coefficient λ₂。

In above-mentioned steps 2.3, auto-correlation weighting coefficient λ₁It is 0.7, cross-correlation weighting coefficient λ₂It is 0.3.

Compared with prior art, the present invention has the following advantages:

(1) transmitting-receiving array element is saved. Have the advantages that M is launched array element and N number of reception array element according to MIMO radar, MN different target echo signal can be obtained in theory. When this just obtains coordinates data amount than conventional array element, save MN-(M+N) individual array element.

(2) Received signal strength noise immunity is reduced. Owing to receiving end adopts correlation reception, non-correlation between noise signal and transmitting signal, so after being correlated with, only through the transmitting signal of target reflection and the noise signal of suppression.

Accompanying drawing explanation

Fig. 1 is that MIMO array is structured the formation mode schematic diagram, and wherein white circle represents transmitting array element, and black circles represents reception array element.

Fig. 2 is formed centrally schematic diagram in virtual phase, and wherein white circle represents transmitting array element, and black circles represents reception array element, and gray circles represents the virtual array element of generation.

Fig. 3 is the autocorrelation function graph of each signal.

Fig. 4 is the cross correlation function figure between each signal.

Fig. 5 is for receiving array element place correlation reception block diagram.

Fig. 6 is multiple goal three-dimensional imaging figure; Wherein (a) is three-dimensional multiple goal BP imaging results, (b) be the degree of depth to 0.2m place imaging slice map, (c) be the degree of depth to 0.3m place imaging slice map, (d) is that the degree of depth is to 0.6m place imaging slice map.

Embodiment

Based on the MIMO radar three-D imaging method of orthogonal phase modulation, comprise step as follows:

The first step, design plane MIMO array, as shown in Figure 1.

According to following principle design plane MIMO array: (1) saves transmitting-receiving array element; (2) array element is launched as far as possible few; (3) launch array element interval big, both can avoid the energy envelope launching signal, coupled interference can be reduced again; (4) receive and dispatch the symmetry of array element, the target of all directions is all had same imaging capability.

The advantage of Fig. 1 design plane MIMO array is: (1) obtains the echoed signal at three-dimensional space difference place as far as possible on a large scale. (2) transmitting-receiving element number of array is saved. (3) form virtual phase center, reduce the graing lobe phenomenon of receiving antenna array.

The formation at virtual phase center, as shown in Figure 2.

Launch array element to target distance R_T, have according to Fei Nieer is approximate:

$R_T=\sqrt{{R_v}^2+d^2-2R_{v}dcos\mathrm{\&delta;}}\&ap;R_v-dcos\mathrm{\&delta;}+\frac{d^2}{2R_v}{\mathrm{sin\&delta;}}^2$

With reason, receive array element to target distance R_R, have according to Fei Nieer is approximate:

$R_R=\sqrt{{R_v}^2+d^2+2R_{v}dcos\mathrm{\&delta;}}\&ap;R_v+dcos\mathrm{\&delta;}+\frac{d^2}{2R_v}{\mathrm{sin\&delta;}}^2$

Wherein,R_vFor the virtual array element formed is to target distance, d is for launching array element to virtual array element distance, and δ is the angle of transmitting-receiving array element place sea line and virtual array element and target place sea line, and λ is the wavelength launching signal; Then

R_T+R_R≈2R_V

Namely bistatic array element is equivalent to and is equipped with a transmitting-receiving with putting array element in transmitting-receiving array element interposition.

2nd step, the transmitting array element that given plane array is concrete and reception array element. One group of orthogonal PSK sequence of identical number is sought to comprise according to transmitting element number of array. Wherein require that the autocorrelative function that each self-emission array element launches signal is similar to impact function, and launch the cross correlation function between signal between two close to 0. The strong and weak directly impact of each cross correlation function launched between signal receives the echoed signal of array element.

It is as follows that concrete PSK sequence produces process:

The l of 2.1L group transmitted wave shape array element launches signal { s_l(n), l=1,2 ... L} represents. Each signal sequence length (i.e. code element number) is N, adopts M phase phase encoding, s_lThe phase place ψ of (n)_lN () represents. s_lN the expression formula of () is:

s_l(n)=exp [j ψ_l(n)]

Wherein n=1,2 ..., N. ψ_lN () is encoding phase,In two-phase encoder matrix, ψ_l(n) only desirable 0 or π. Phase combination { ψ_l(1),ψ_l(2),…,ψ_l(N) } corresponding phase encoding matrix { s is determined_l(1),s_l(2),…,s_l(N) orthogonal property }. s_lThe autocorrelative function of (n) R (ψ_l, k) represent, s_p(n) and s_qThe cross correlation function of (n) C (ψ_p,ψ_q, k) represent, wherein-N < k < Np, q ∈ 1,2 ... L}, and p ≠ q.

2.2, according to the autocorrelative function side-lobe energy of signal and cross correlation function energy structure objective function W, optimize minimumization W to try to achieve encoding sequence. Objective function represents:

$W={\mathrm{\&lambda;}}_1\&times;{\mathrm{\&Sigma;}}_{l=1}^L({\mathrm{\&Sigma;}}_{k=-(N-1)}^{N-1}{\left|R({\mathrm{\&psi;}}_l,k)\right|}^2-{\left|R({\mathrm{\&psi;}}_l,0)\right|}^2)+{\mathrm{\&lambda;}}_2\&times;{\mathrm{\&Sigma;}}_{p=1}^{L-1}{\mathrm{\&Sigma;}}_{q=p+1}^L{\mathrm{\&Sigma;}}_{k=-(N-1)}^{N-1}{\left|C({\mathrm{\&psi;}}_p,{\mathrm{\&psi;}}_q,k)\right|}^2$

Wherein λ₁, λ₂It is the weighting coefficient of objective function, 0 < λ₁< 1,0 < λ₂< 1 and λ₁+λ₂=1, general λ₁>λ₂, it is possible to reach the object suppressing autocorrelation sidelobe energy. As long as given L, M, N and concrete weighting coefficient can adopt genetic Optimization Algorithm optimization object function.

2.3 produce T group phase encoding matrix at random, and wherein each group all comprises L signal group, and each signal group Baud Length is all that N, M phase encodes.According to genetic algorithm, using the random T group coding sequence produced as initial value, calculate the fitness function value (using the objective function built as adaptive value function) of each code character signal in this T group, if this value meets the termination condition pre-set, then terminate. Otherwise, again this T group signal being optimized, retain the signal group that fitness function value is little, other group of functions that cross and variation fitness function value is big, as new initial value, then go to judge whether to meet loop stop conditions, till meeting. Select the group coding sequence making objective function E minimum as final encoding sequence in the T group signal met. The group number T of the random encoder matrix produced can set arbitrarily according to demand, but in order to obtain optimizing preferably effect, the group number T of the random encoder matrix produced should much larger than transmitting element number of array L, i.e. T > > L.

Initialize L=4, M=4, N=120 and weighting coefficient λ₁=0.7, λ₂=0.3, produce PSK sequence according to said process. Namely launching signal and have four groups, four phase encodings, every group coding length 120 code element, each signal auto-correlation function produced, cross correlation function is respectively as shown in Figures 3 and 4.

3rd step, connects L matched filter respectively at each receiving end, does the mode of relevant filtering to all transmitting array element signals successively by the data of receiving end, it is achieved the different object launching echoed signal corresponding to array element of separation. Namely MIMO radar launch simultaneously simultaneously receive mode can obtain far more than reality receive element number of array data. Correlation reception process is such as Fig. 5.

Due to after each receiving end matched filtering, except the autocorrelation signal of useful new number also receives other cross-correlated signal launching signals, jth receives the signal that carry out self-emission array element l of array element after matched filtering separation and represents and be:

r_lj(n)=b_l(n-(t_lq ⁽ⁿ⁾+τ_jq ⁽ⁿ⁾))+∑_{M=1 ... L, m ≠ l}b_lm(n-(τ_mq ⁽ⁿ⁾+τ_jq ⁽ⁿ⁾))

Wherein, receive array element number J and represent, j=1,2 ..., J. b_lN () launches autocorrelative function corresponding to array element l, b_lmN () represents the cross correlation function launched between array element l and other transmittings array element signals m. τ_lq ⁽ⁿ⁾It is launch array element l to count to the sampling interval that the distance of scatter point q is corresponding. Due to noise, signal does not have dependency with launching, so noise is just weakened in the process of correlation reception.

4th step, owing to the obtain manner of MIMO radar data is different with traditional synthetic-aperture radar so that conventional radar imaging method is difficult to directly be applied in MIMO radar imaging. If the distance range and Doppler in synthetic aperture imaging algorithm, distance migration algorithm are all with the even spatial sampling put based on transmitting-receiving, it is difficult to the directly echoed signal of the many transmitting-receivings of process. And general existing three-dimensional distance skew imaging algorithm is based on far-field approximation.

But the rear orientation projection (backprojection according to echo temporal amplitude information, BP) algorithm but can process the echoed signal that different array manifold is arranged, derives according to ordinary two dimensional BP imaging algorithm and is applied to difference and structures the formation the three-dimensional BP imaging algorithm of mode.

Divide whole three-dimensional imaging region E × F × H, travel through each pixel (x successively_e,y_f,z_h), wherein e=1 ..., Ef=1 ..., Fh=1 ..., H. According to echoed signal r_ljN (), calculates the range value I (e, f, h) of each pixel.

$I(e,f,h)={\mathrm{\&Sigma;}}_{l=1}^L{\mathrm{\&Sigma;}}_{j=0}^J{r}_{lj}(n-{\mathrm{\&tau;}}_{l,j}^{(n,e,f,h)})$

Represent that the l is launched array element through scatter point (x_e,y_f,z_h) arrive the time delay that jth receives array element.

5th step, according to each pixel (x_e,y_f,z_h) try to achieve the pixel value I (e, f, h) of corresponding point position, whole three-dimensional imaging space pixel can form three-dimensional matrice, is drawn according to the actual size of imaging region by three-dimensional matrice, so that it may obtain three-dimensional space target information.

For the even fine sand medium of 1m × 1m × 1m, point target position is separately positioned on (0.2,0.8,0.6), (0.6,0.2,0.3), (0.2,0.8,0.2), (0.5,0.5,0.2) place, according to step one to step 5, three-dimensional imaging result is such as Fig. 6, and wherein (a) is three-dimensional multiple goal BP imaging results, and (b) is that the degree of depth is to 0.2m place imaging slice map, (c) be the degree of depth to 0.3m place imaging slice map, (d) is that the degree of depth is to 0.6m place imaging slice map.

Claims (5)

1., based on the MIMO radar three-D imaging method of orthogonal phase modulation, it is characterized in that, comprise the following steps:

Step 1, meets, according to the design of the size in three-dimensional imaging region, the plane array required, and the transmitting array element of given plane array and reception array element; Each output terminal receiving array element is respectively connected to the quantity matched filter identical with the number launching array element, and each matched filter launches array element corresponding to one;

Step 2, constructs the orthogonal PSK sequence of respective sets number according to transmitting element number of array, and each launches the corresponding one group of orthogonal PSK sequence of array element;

Step 3, receives array element and receives echoed signal, and the echoed signal received is sent to matched filter and carries out relevant filtering, thus realizes the different separation launching echoed signal corresponding to array element;

Step 4, divides whole three-dimensional imaging region, travels through the range value of echoed signal corresponding to the time delay spacing point place of the transmitting-receiving array element of each pixel of imaging region successively; The range value of the echoed signal that the time delay spacing point of all transmitting-receiving array elements that cumulative this pixel of summation is corresponding is corresponding successively;

Step 5, coordinate information and the required range value of whole three-dimensional imaging space pixel can form three-dimensional matrice, drawn according to the actual size of imaging region by three-dimensional matrice, so that it may obtain three-dimensional space target information.

2., according to claim 1 based on the MIMO radar three-D imaging method of orthogonal phase modulation, it is characterized in that, in step 2, as follows according to the process of the orthogonal PSK sequence launching element number of array structure respective sets number,

Step 2.1, the number of phases M of setting code length N and encoding phase;

Step 2.2, random generation T group coding matrix { s_l(n) }, the line number of every group coding matrix is L, and row number is N, and element is s_l(n), wherein

s_l(n)=exp [j ψ_l(n)]

Wherein, ${\mathrm{\&psi;}}_l\left(n\right)\&Element;\{0,\frac{2\mathrm{\&pi;}}{M},2\&CenterDot;\frac{2\mathrm{\&pi;}}{M},...,(M-1)\&CenterDot;\frac{2\mathrm{\&pi;}}{M}\};$ M is the number of phases of encoding phase; N=1,2 ..., N, N are code length; L=1,2 ..., L, L are for launching element number of array;

Step 2.3, for each group coding matrix, with behavior unit, calculate autocorrelation sidelobe energy A (l) of every a line of this group coding matrix and the cross-correlation energy B (p of every two row, q), and calculate the target function value W of this group coding matrix accordingly:

$W={\mathrm{\&lambda;}}_1\&times;{\mathrm{\&Sigma;}}_{l=1}^{L}A\left(l\right)+{\mathrm{\&lambda;}}_2\&times;{\mathrm{\&Sigma;}}_{p=1}^{L-1}{\mathrm{\&Sigma;}}_{q=p+1}^{L}B(p,q)$

Wherein, λ₁For auto-correlation weighting coefficient, λ₂For cross-correlation weighting coefficient, 0 < λ₁< 1,0 < λ₂< 1, λ₁+λ₂=1; L=1,2 ..., L; P=1,2 ..., L; Q ,=1,2 ..., L;

Step 2.4, retains that minimum for target function value group coding matrix, and according to genetic algorithm, all the other T-1 group coding matrixes is carried out cross and variation, thus forms the encoder matrix of new T group;

Step 2.5, repeating step 2.3-2.4, until when target function value is less than goal-selling threshold value or iteration number of times reaches default iteration threshold, then the every a line of the encoder matrix that group target function value this iteration retained is minimum respectively launches the transmitting signal of array element as one.

3. according to claim 2 based on the MIMO radar three-D imaging method of orthogonal phase modulation, it is characterized in that, in step 2.2, the group number T of the random encoder matrix produced is greater than transmitting element number of array L, i.e. T > L.

4., according to claim 2 based on the MIMO radar three-D imaging method of orthogonal phase modulation, it is characterized in that, in step 2.3, auto-correlation weighting coefficient l₁It is greater than cross-correlation weighting coefficient λ₂, i.e. λ₁> λ₂。

5., according to claim 4 based on the MIMO radar three-D imaging method of orthogonal phase modulation, it is characterized in that, in step 2.3, auto-correlation weighting coefficient λ₁It is 0.7, cross-correlation weighting coefficient λ₂It is 0.3.