CN101995574A - Near field focusing beam forming positioning method - Google Patents

Abstract

The invention relates to a near field focusing beam forming positioning method. In the method, the response of a signal received by each array element of a virtual array in a frequency domain, to be processed, within the frequency band range [fmin, fmax] is firstly obtained based on an expansion towing array passive synthesized aperture method, and then a target is accurately positioned by utilizing the near field focusing beam forming positioning method, wherein the frequency band range [fmin, fmax] is selected according to the actual requirement, and only the conditions that fmin is greater than 0, and fmax is not greater than one half of data sampling rate are satisfied. The invention effectively combines the expansion towing array passive synthesized aperture method and the near field focusing beam forming positioning method together and meanwhile acquires the benefits brought by the two methods. Moreover, the beam forming arithmetic is realized in the frequency domain; a broadband signal is decomposed into a plurality of frequency point signals, and each frequency point within the signal frequency band range is processed, and finally, results are accumulated; and the processing performances, such as robustness, resolution and the like, are more superior to those of direct processing in a time domain.

Description

A kind of near field focus beam forms localization method

Technical field

The present invention relates to the sonar digital processing field, particularly a kind of near field focus beam forms localization method.

Background technology

At the detection and the measurement of middle far field target, the method for passive positioning under water commonly used has three sub-array localization method, target motion analysis method and a coupling disposal route etc.For the accurate location of near field target, it is comparatively suitable that focus beam forms the location rule.Conventional beam-forming technology is based on the far field plane wave approximation, is equivalent to a spatial filter, thereby carries out the spatial spectrum density that compensation of delay obtains signal and carry out measurement of bearing according to the get it right sound wave that receives of bearing meter.But for the near field target, only consider the orientation and ignore influence that conventional beam-forming technology not only can not be found range, and target is carried out orientation also can go wrong apart from the factor.In this case, just need carry out the compensation addition again of spherical wave form to received signal according to different directions and the last time-delay of distance, obtain target direction and range information from the crest that forms, this just forms technology at the focus beam of near field target.

But the continuous reduction along with the frequency of underwater sound weak signal wants to keep original detection performance, and array aperture just must be increasing, and this makes corresponding engineering use more and more to be difficult to realize.Given this, utilize the passive synthetic aperture technique of expansion towed array can rely on the various benefits of the long battle array of motor-driven acquisition of weak point battle array, provide a kind of feasible thinking for utilizing short gust of accurate location that obtains near field low frequency and even very low frequency (VLF) underwater sound weak signal.The technology that passive synthetic aperture (PASA) is a kind of passive receiving target noise, motion and signal processing method by array manually increase the small-bore array length.Passive synthetic aperture technique commonly used has two kinds: expansion towed array technology (ETAM); The passive synthetic aperture technique of Fast Fourier Transform (FFT) (FFTSA).The former uses overlapping correlator to estimate the phase perturbation that is caused by medium and motion path disturbance between the snap, the latter then be directly with certain hour at interval the phase differential of interior two snap signals as constant, the former excellent performance, the latter is convenient to utilize FFT to calculate fast, but greatly reduces performance.

In general, focus beam forms technology to the near field target effective, but only uses based on actual array in the past, and near field range is limited, and accurately the distance of localizing objects is subjected to the restriction in actual array aperture; Passive synthetic aperture technique, expansion towed array technology has especially wherein increased array aperture effectively, but fails to combine to realize the more accurate location of target with focus beam formation technology in the past.

Summary of the invention

The objective of the invention is to, provide a kind of near field focus beam to form localization method, obtain the advantage that expansion towed array technology and near field focus beam form localization method simultaneously, target is accurately located.

For achieving the above object, the present invention proposes a kind of near field focus beam and forms localization method.This method obtains the signal of each array element reception of virtual array earlier at frequency band range [f to be processed based on the passive synthetic aperture method of expansion towed array _Min, f _Max] interior frequency domain response, then, utilize the near field focus beam to form localization method target is accurately located; Wherein, frequency band range [f _Min, f _Max] choose according to actual needs, only need satisfy f _Min＞0, f _MaxHalf that is not more than data sampling rate gets final product.

The passive synthetic aperture method of described expansion towed array comprises that concrete steps are as follows:

Step 1.1): M array element receives the acoustic pressure time-domain signal in the linear array in the towing process, and wherein, M is the linear array array element number, and M is not less than 2 integer; Get the data of ML+1 snap, each snap data comprises M road sound pressure signal, and wherein, ML is the number of times of aperture expansion, and ML＞0 is chosen according to the actual requirements; By the length of half actual array length as an aperture expansion, the overlapping M/2 array element during expansion of promptly each aperture, calculating twice expansion according to formula (1) is time interval between two snaps;

$\mathrm{\&tau;}=\frac{\mathrm{<figure-callout\; id="2"\; label="M"\; filenames="HSA00000336601500012.png,HSA00000336601500031.png"\; state="\{\{state\}\}">M</figure-callout>}}{2}\&CenterDot;d/v---\left(1\right)$

Wherein, d is an array element distance, determines according to actual needs, usually d ∈ [0.1m, 10m]; V is the towed array movement velocity, v 〉=3m/s, but carry out the aperture expansion of same number of times, speed is big more, needs the time domain data of accumulation few more;

Step 1.2): adjacent two snaps such as k, k+1 snap acoustic pressure time-domain signal are done Fourier transform, obtain corresponding M road frequency-region signal, the k snap of m array element is pressed formula (2) data conversion process:

$x_{k,m}\left(t\right)\stackrel{\mathrm{FFT}}{\&DoubleRightArrow;}{X}_{k,m}\left(f_i\right),m=\mathrm{1,2},...,M---\left(2\right)$

Wherein, 0＜k≤ML;

Step 1.3): at the frequency band range [f of needs processing _Min, f _Max], according to the M/2 of k snap and k+1 snap overlapping array element is calculated M/2 phase differential by formula (3) and get weighted mean again and obtain phase perturbation ψ _K+1(f _i),

${\mathrm{\&psi;}}_{k+1}\left(f_i\right)=\arg \left\{\frac{\underset{m=1}{\overset{M/2}{\mathrm{\&Sigma;}}}{X}_{k,M/2+m}\left(f_i\right)\&times;X_{k+1,m}^{*}\left(f_i\right)\&times;{\mathrm{\&rho;}}_m\left(f_i\right)}{\underset{m=1}{\overset{M/2}{\mathrm{\&Sigma;}}}{\mathrm{\&rho;}}_m\left(f_i\right)}\right\}---\left(3\right)$

Wherein, ρ _m(f _i) expression m array element frequency f _iCompensating factor, its formula is formula (4):

${\mathrm{\&rho;}}_m\left(f_i\right)=\frac{|\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{X}_{k,M/2+m}\left(f_i\right)\&times;X_{k+1,m}^{*}\left(f_i\right)|}{\sqrt{\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{\left|X_{k,M/2+m}\left(f_i\right)\right|}^2\&times;\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{\left|X_{k+1,m}^{*}\left(f_i\right)\right|}^2}}---\left(4\right)$

Wherein, Q is the frequency span of compensating factor;

Step 1.4): the different frequency component to k+1 snap the M/2+1～M array element signals carries out phase compensation respectively, it is invented the M+1～M+M/2 array element of k snap, and the like, this M/2 array element signals can virtually become M+M/2 (k-1)+1～M+M/2 (the k-1)+M/2 array element of the 1st snap, that is to say that whenever carrying out an aperture synthesizes, can fictionalize M/2 array element, suc as formula (5):

$X_{k+1,M/2+m}\left(f_i\right)\stackrel{*\exp (-{\mathrm{\&psi;}}_{k+1}\left(f_i\right))}{\&DoubleRightArrow;}{X}_{k,M+m}\left(f_i\right)\stackrel{*\exp (-\underset{j=2}{\overset{k}{\mathrm{\&Sigma;}}}{\mathrm{\&psi;}}_j\left(f_i\right))}{\&DoubleRightArrow;}{X}_{1,M+M/2*(k-1)+m}\left(f_i\right)---\left(5\right);$

Step 1.5): judge whether to have carried out ML aperture and synthesize; If, then virtual obtaining

The signal that individual array element receives is at [f _Min, f _Max] the interior frequency domain response of scope, wherein, MM is the towed array array number after the expansion, is calculated by actual array number and aperture expansion number of times; If not, then repeating step 1.2), step 1.3) and step 1.4).

Described near field focus beam forms the localization method step and comprises:

Step 2.1): at first get in the near-field scan scope and first array element between distance R _j, wherein, R _j=R ₁, R ₂..., R _NRBe total to NR scanning distance, wherein, R _jGet final product less than near field range, NR＞5 are specifically decided according near field range size and scanning accuracy; Each frequency f in required bandwidth to be processed _iOn, at position angle φ _kScan, wherein, φ _k=φ ₁..., φ _NBBe total to NB scanning angle, NB 〉=181 are specifically determined according to the angle scanning precision; Employing obtains space power spectrum output on each frequency at the near field focus beam formation method of single-frequency point;

Step 2.2): according to formula (6) wave beam on each frequency is formed the result and add up, obtain array in distance R _jLocate the power output on each orientation;

$P(R_j,{\mathrm{\&phi;}}_k)=\underset{f_i=\min }{\overset{\max }{\mathrm{\&Sigma;}}}{P}_{f_i}(R_j,{\mathrm{\&phi;}}_k)---\left(6\right)$

Step 2.3): judging distance R _jWhether scanning is intact; If the towed array two-dimensional power spectrum output P (R on different distance and position angle at last is expanded _j, φ _k), wherein, φ _k=φ ₁..., φ _NB, R _j=R ₁, R ₂..., R _NR, can realize location to target according to this two-dimensional power spectrum; If not, repeating step 2.1) and step 2.2).

Described near field at single-frequency point focuses on conventional wave beam formation method concrete steps and comprises:

Frequency f _iOn array response by shown in the formula (7):

X(f _i)＝[X _1，1(f _i)，X _1，2(f _i)，…，X _1，MM(f _i)] ^T (7)

Frequency f _iOn data covariance matrix by shown in the formula (8):

R(f _i)＝X(f _i)·X(f _i) ^* (8)

When distance is R _jThe time, frequency f _iOn space power spectrum output by shown in the formula (9):

$P_{f_i}(R_j,{\mathrm{\&phi;}}_k)=a^{*}R\left(f_i\right)a,$ φ _k＝φ ₁，…，φ _NB (9)

Wherein,

S=0,1,2 ..., MM-1 is the guiding vector of array; W=2 π f _iTime-delay

d _sThe expression target respectively and the distance between s+1 array element and the 1st array element poor,

P _sThe position of representing s+1 array element, P _s=sd.

When target is narrow band signal, described step 2.1) can adopt the focusing localization method that forms based on the MVDR wave beam to obtain space power spectrum output on each frequency.

Described step 2.1) adopting the near field that forms based on the MVDR wave beam to focus on the localization method concrete steps at single-frequency point comprises:

Frequency f _iOn array response be:

X(f _i)＝[X _1，1(f _i)，X _1，2(f _i)，…，X _1，MM(f _i)] ^T

Frequency f _iOn data covariance matrix be:

R(f _i)＝X(f _i)·X(f _i) ^*

When distance is R _jThe time, frequency f _iOn space power spectrum output type (10):

$P_{f_i}(R_j,{\mathrm{\&phi;}}_k)=\frac{1}{a^{*}R{\left(f_i\right)}^{-1}a},$ φ _k＝φ ₁，…，φ _NB (10)

Wherein,

S=0,1,2 ..., MM-1 is the guiding vector of array; W=2 π f _iTime-delay

d _sThe expression target respectively and the distance between s+1 array element and the 1st array element poor, P _sThe position of representing s+1 array element, P _s=sd.

In step 2.1), step 2.2) and step 2.3) in, can change about azimuthal scanning sequency of distance, promptly can be for a fixed angle, advanced row distance scanning, and then carry out repetitive operation at the different scanning angle.

The invention has the advantages that, to expand passive synthetic aperture method of towed array and near field focus beam formation localization method combines effectively, obtain the benefit that the two brings simultaneously, utilize the motor-driven yardstick of expanding the aperture of actual array, thereby make the near field range of synthetic aperture increase greatly, form the accurate location that can be implemented in the certain limit internal object in conjunction with focus beam.The change in aperture has also improved the resolution and the Detection of weak ability of target azimuth and distance greatly greatly.In addition, wave beam forms computing and realizes at frequency domain, and broadband signal is decomposed into a plurality of frequency signals, and each frequency in the signal band scope is handled, at last the result is added up, handling property such as robustness, resolution geometric ratio are directly handled more superior in time domain.

Description of drawings

Fig. 1 is expansion towed array method principle schematic;

Fig. 2 forms the location synoptic diagram for the near field focus beam;

Fig. 3 forms the localization method process flow diagram for a kind of near field focus beam that the present invention proposes;

Fig. 4 forms locating effect figure for focus on conventional wave beam based on 40 yuan of actual array near fields;

Fig. 5 forms locating effect figure for focusing on conventional wave beam based on the 200 element array near fields that obtained for 8 times by 40 yuan of actual array expansions;

Fig. 6 focuses on the MVDR wave beam based on 40 yuan of actual array near fields to form locating effect figure;

Fig. 7 forms locating effect figure for focusing on the MVDR wave beam based on the 200 element array near fields that obtained for 8 times by 40 yuan of actual array expansions.

Embodiment

The present invention will be described in detail below in conjunction with the drawings and specific embodiments.

Expansion towed array method principle schematic, as shown in Figure 1.The concrete steps of expansion towed array method are as follows: step 1.1): M array element receives the acoustic pressure time-domain signal in the linear array in the towing process, and wherein, M is the linear array array element number, and M is not less than 2 integer; Get the data of ML+1 snap, each snap data comprises M road sound pressure signal, and wherein, ML is the number of times of aperture expansion, and ML＞0 is chosen according to the actual requirements; By the length of half actual array length as an aperture expansion, the overlapping M/2 array element during expansion of promptly each aperture, calculating twice expansion according to formula (1) is time interval between two snaps;

$\mathrm{\&tau;}=\frac{\mathrm{<figure-callout\; id="2"\; label="M"\; filenames="HSA00000336601500012.png,HSA00000336601500031.png"\; state="\{\{state\}\}">M</figure-callout>}}{2}\&CenterDot;d/v---\left(1\right)$

Wherein, d is an array element distance, determines according to actual needs, usually d ∈ [0.1m, 10m]; V is the towed array movement velocity, v 〉=3m/s, but carry out the aperture expansion of same number of times, speed is big more, needs the time domain data of accumulation few more;

Step 1.2): adjacent two snaps such as k, k+1 snap acoustic pressure time-domain signal are done Fourier transform, obtain corresponding M road frequency-region signal, as the k snap data conversion process of m array element suc as formula (2):

$x_{k,m}\left(t\right)\stackrel{\mathrm{FFT}}{\&DoubleRightArrow;}{X}_{k,m}\left(f_i\right),m=\mathrm{1,2},...,M---\left(2\right)$

Wherein, 0＜k≤ML;

Step 1.3): at interested frequency band range [f _Min, f _Max], according to the M/2 of k snap and k+1 snap overlapping array element is calculated M/2 phase differential and get weighted mean again and obtain phase perturbation ψ _K+1(f _i), suc as formula (3):

${\mathrm{\&psi;}}_{k+1}\left(f_i\right)=\arg \left\{\frac{\underset{m=1}{\overset{M/2}{\mathrm{\&Sigma;}}}{X}_{k,M/2+m}\left(f_i\right)\&times;X_{k+1,m}^{*}\left(f_i\right)\&times;{\mathrm{\&rho;}}_m\left(f_i\right)}{\underset{m=1}{\overset{M/2}{\mathrm{\&Sigma;}}}{\mathrm{\&rho;}}_m\left(f_i\right)}\right\}---\left(3\right)$

Wherein, ρ _m(f _i) expression m array element frequency f _iCompensating factor, its formula is suc as formula (4):

${\mathrm{\&rho;}}_m\left(f_i\right)=\frac{|\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{X}_{k,M/2+m}\left(f_i\right)\&times;X_{k+1,m}^{*}\left(f_i\right)|}{\sqrt{\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{\left|X_{k,M/2+m}\left(f_i\right)\right|}^2\&times;\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{\left|X_{k+1,m}^{*}\left(f_i\right)\right|}^2}}---\left(4\right)$

Wherein, Q is the frequency span of compensating factor;

Step 1.4): the different frequency component to k+1 snap, the M/2+1～M array element signals carries out phase compensation respectively, it is invented the M+1～M+M/2 array element of k snap, and the like, this M/2 array element signals can virtually become M+M/2 (k-1)+1～M+M/2 (the k-1)+M/2 array element of the 1st snap.That is to say that whenever to carry out aperture synthetic, can fictionalize M/2 array element, suc as formula (5):

$X_{k+1,M/2+m}\left(f_i\right)\stackrel{*\exp (-{\mathrm{\&psi;}}_{k+1}\left(f_i\right))}{\&DoubleRightArrow;}{X}_{k,M+m}\left(f_i\right)\stackrel{*\exp (-\underset{j=2}{\overset{k}{\mathrm{\&Sigma;}}}{\mathrm{\&psi;}}_j\left(f_i\right))}{\&DoubleRightArrow;}{\mathrm{<figure-callout\; id="1"\; label="X"\; filenames="HSA00000336601500012.png"\; state="\{\{state\}\}">X</figure-callout>}}_{1,M+M/2*(k-1)+m}\left(f_i\right)---\left(5\right);$

Step 1.5): judge whether to have carried out ML aperture and synthesize; If, then virtual obtaining

The signal that individual array element receives is at [f _Min, f _Max] the interior frequency domain response of scope, wherein, MM is the towed array array number after the expansion, is calculated by actual array number and aperture expansion number of times; If not, then repeating step 1.2), step 1.3), step 1.4).

The near field focus beam forms the location synoptic diagram, as shown in Figure 2.The concrete steps that the near field focus beam forms the location are as follows:

Step 2.1): at first get in the near-field scan scope and first array element between distance R _j, wherein, R _j=R ₁, R ₂..., R _NRBe total to NR scanning distance, wherein, R _jGet final product less than near field range, NR＞5 are specifically decided according near field range size and scanning accuracy; Each frequency f in required bandwidth to be processed _iOn, at position angle φ _kScan, wherein, φ _k=φ ₁..., φ _NBBe total to NB scanning angle, NB 〉=181 are specifically determined according to the angle scanning precision; Employing is exported at the space power spectrum that the conventional wave beam formation method of the focusing of single-frequency point obtains on each frequency;

Step 2.2): according to formula (6) wave beam on each frequency is formed the result and add up, obtain array in distance R _jLocate the power output on each orientation;

$P(R_j,{\mathrm{\&phi;}}_k)=\underset{f_i=\min }{\overset{\max }{\mathrm{\&Sigma;}}}{P}_{f_i}(R_j,{\mathrm{\&phi;}}_k)---\left(6\right)$

Step 2.3): judging distance R _jWhether scanning is intact; If the towed array two-dimensional power spectrum output P (R on different distance and position angle at last is expanded _j, φ _k), wherein, φ _k=φ ₁..., φ _NB, R _j=R ₁, R ₂..., R _NR, can realize location to target according to this two-dimensional power spectrum; If not, repeating step 2.1) and step 2.2).

Wherein, focusing on conventional wave beam formation method can be replaced by the focusing localization method that forms based on the MVDR wave beam.

The conventional wave beam formation method of focusing particular content at single-frequency point is as follows:

Frequency f _iOn array data formula (7):

X(f _i)＝[X _1，1(f _i)，X _1，2(f _i)，…，X _1，MM(f _i)] ^T (7)

Frequency f _iOn data covariance matrix formula (8):

R(f _i)＝X(f _i)·X(f _i) ^* (8)

When distance is R _jThe time, frequency f _iOn space power spectrum output type (9):

$P_{f_i}(R_j,{\mathrm{\&phi;}}_k)=a^{*}R\left(f_i\right)a,$ φ _k＝φ ₁，…，φ _NB (9)

Wherein,

S=0,1,2 ..., MM-1 is the guiding vector of array; W=2 π f _iTime-delay d _sThe expression target respectively and the distance between s+1 array element and the 1st array element poor, P _sThe position of representing s+1 array element, P _s=sd.

When target is narrow band signal, described step 2.1) adopt the near field focusing localization method concrete steps that form based on the MVDR wave beam to comprise at single-frequency point:

Frequency f _iOn array response be:

X(f _i)＝[X _1，1(f _i)，X _1，2(f _i)，…，X _1，MM(f _i)] ^T

Frequency f _iOn data covariance matrix be:

R(f _i)＝X(f _i)·X(f _i) ^*

When distance is R _jThe time, frequency f _iOn space power spectrum output type (10):

$P_{f_i}(R_j,{\mathrm{\&phi;}}_k)=\frac{1}{a^{*}R{\left(f_i\right)}^{-1}a},$ φ _k＝φ ₁，…，φ _NB (10)

Wherein,

S=0,1,2 ..., MM-1 is the guiding vector of array; W=2 π f _iTime-delay

d _sThe expression target respectively and the distance between s+1 array element and the 1st array element poor,

P _sThe position of representing s+1 array element, P _s=sd

In step 2.1), step 2.2) and step 2.3) in, can change about azimuthal scanning sequency of distance, promptly can be for a fixed angle, advanced row distance scanning, and then carry out repetitive operation at the different scanning angle.

Specific embodiment

Simulation parameter: actual sound pressure sensor line array array number M=40, array element distance d=1m does ML=8 aperture expansion, the towed array array number MM=200 after the expansion, towed array movement velocity v=4m/s; Signal sampling rate fs=20KHz.Three sinusoidal signals, its frequency, received signal to noise ratio, position angle, distance are respectively: (133Hz, 0dB, 126 °, 300m), (149Hz, 0dB, 150 °, 500m), (157Hz, 0dB, 60 °, 700m).Velocity of sound c=1500m/s, frequency band 130Hz～160Hz is handled in snap length N=2000.The angle scanning scope is [0 a °, 180 °], every 1 ° of run-down, and totally 181 angles.The range sweep scope is [100m, 1000m], every the 10m run-down, and totally 91 distances.

As shown in Figure 3, a kind of near field focus beam formation localization method concrete steps of the present invention's proposition are as follows:

Step 1: 301 in the corresponding diagram 3,40 array elements of array receive spacing wave in the towing process, get the data of ML+1=9 snap, each snap data comprises 40 road sound pressure signals, and 20 array elements of overlapping half linear array are calculated the time interval between the adjacent snap during by the expansion of each aperture.

$\mathrm{\&tau;}=\frac{\mathrm{<figure-callout\; id="2"\; label="M"\; filenames="HSA00000336601500012.png,HSA00000336601500031.png"\; state="\{\{state\}\}">M</figure-callout>}}{2}\&CenterDot;d/v=\frac{40}{2}\&CenterDot;1/4=4s---\left(11\right)$

Step 2: 302 in the corresponding diagram 3, adjacent two snaps such as k, k+1 snap acoustic pressure time-domain signal are done Fourier transform, obtain corresponding M road frequency-region signal, as follows as the k snap data conversion process of m array element:

Step 3: 303 in the corresponding diagram 3 at interested frequency band range [130Hz, 160Hz], calculates 20 phase differential according to the overlapping array element of 20 couple of k snap and k+1 snap and gets weighted mean again and obtain phase perturbation ψ _K+1(f _i).

${\mathrm{\&psi;}}_{k+1}\left(f_i\right)=\arg \left\{\frac{\underset{m=1}{\mathrm{\&Sigma;}}{X}_{k,20+m}\left(f_i\right)\&times;X_{k+1,m}^{*}\left(f_i\right)\&times;{\mathrm{\&rho;}}_m\left(f_i\right)}{\underset{m=1}{\overset{20}{\mathrm{\&Sigma;}}}{\mathrm{\&rho;}}_m\left(f_i\right)}\right\}---\left(13\right)$

Wherein, ρ _m(f _i) expression m array element frequency f _iCompensating factor:

${\mathrm{\&rho;}}_m\left(f_i\right)=\frac{|\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{X}_{k,20+m}\left(f_i\right)\&times;X_{k+1,m}^{*}\left(f_i\right)|}{\sqrt{\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{\left|X_{k,20+m}\left(f_i\right)\right|}^2\&times;\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{\left|X_{k+1,m}^{*}\left(f_i\right)\right|}^2}}---\left(14\right)$

Wherein, Q is the frequency span of compensating factor, is taken as 10.

Here, interested frequency band range is [130Hz, 160Hz], signal sampling rate is 20KHz, snap length is that FFT counts is 2000, and then the discrete frequency of actual treatment is [130,160]/20K*2000=[13,16], therefore get final product through only handling after the FFT to the discrete frequency of this section.So above-mentioned two formula medium frequency f _iSpan be [13,16].

Step 4: 304 in the corresponding diagram 3, to the different frequency component X of k+1 snap, the 21st～40 array element signals _{K+1, m}(f _i) be ψ respectively _K+1(f _i) phase compensation, it is invented the 41st～60 array element of k snap.Be ψ more respectively _k(f _i) phase compensation, it is invented the 61st～80 array element of k-1 snap.And the like, these 20 array element signals can virtually become 40+20 (k-1)+1～40+20k array element of the 1st snap.That is to say that whenever carrying out an aperture synthesizes, and can fictionalize 20 array elements.

$X_{k+1,20+m}\left(f_i\right)\stackrel{*\exp (-{\mathrm{\&psi;}}_{k+1}\left(f_i\right))}{\&DoubleRightArrow;}{X}_{k,40+m}\left(f_i\right)\stackrel{*\exp (-\underset{j=2}{\overset{k}{\mathrm{\&Sigma;}}}{\mathrm{\&psi;}}_j\left(f_i\right))}{\&DoubleRightArrow;}{\mathrm{<figure-callout\; id="1"\; label="X"\; filenames="HSA00000336601500012.png"\; state="\{\{state\}\}">X</figure-callout>}}_{1,40+20(k-1)+m}\left(f_i\right)---\left(15\right)$

Step 5: repeating step 2,3 and 4, the aperture of carrying out 8 times is synthetic, can the same frequency domain response X of signal in [130Hz, 160Hz] scope that takes reception soon of the virtual 40+20*8=200 of an obtaining array element _ETA

X _ETA＝[X(13)，X(14)，…，X(16)] (16)

X(i)＝[X ₁(i)，X ₂(i)，…，X ₄₀(i)，…，X ₂₀₀(i)] ^T，i＝13，14，…，16 (17)

Step 6: 305,306 in the corresponding diagram 3, get a certain fixed range R in the near-field scan scope _j, R _jDistance between expression and first array element, in the present embodiment, R _j=100m, 110m ..., 1000m, the discrete frequency f of each of bandwidth of interest correspondence _iOn, in the present embodiment, f _i=13,14 ..., 16 at position angle φ _kScan, in the present embodiment, φ _k=0 °, 1 ° ..., 180 °, totally 181 scanning angles utilize the space power spectrum that obtains on each frequency at the conventional wave beam formation method of the focusing of single-frequency point to export.

Frequency f _iOn array data be X (f _i), frequency f _iOn data covariance matrix be

R(f _i)＝X(f _i)·X(f _i) ^* (18)

When distance is R _jThe time, frequency f _iOn space power spectrum be output as

$P_{f_i}(R_j,{\mathrm{\&phi;}}_k)=a^{*}R\left(f_i\right)a,$ φ _k＝0°，1°，…，180° (19)

Wherein,

S=0,1,2 ..., 199, be the guiding vector of array,

Time-delay

d _sThe expression target is respectively and the distance between s+1 array element and the 1st array element poor:

P _sThe position of representing s+1 array element: P _s=sd.

Step 7: 307 in the corresponding diagram 3, the wave beam of synthetic different frequency forms the result, obtains array in distance R _jLocate the power output on each orientation.

$P(R_j,{\mathrm{\&phi;}}_k)=\underset{f_i=13}{\overset{16}{\mathrm{\&Sigma;}}}{P}_{f_i}(R_j,{\mathrm{\&phi;}}_k)---\left(20\right)$

Step 8: 308 in the corresponding diagram 3, to different distance R _j, in the present embodiment, R _j=100m, 110m ..., 1000m, totally 91 scanning distances, repeating step 6 and step 7, the towed array two-dimensional power spectrum output P (R on different distance and position angle at last is expanded _j, φ _k), φ wherein _k=0 °, 1 ° ..., 180 °, R _j=100m, 110m ..., 1000m.Can realize location according to this two-dimensional power spectrum to target.

Can keep other step constant and step 6 is replaced with MVDR wave beam as described below forms and focus on localization method, to obtain better bearing accuracy and resolution at the narrow band signal target.

Step 6: 305,306 in the corresponding diagram 3, get a certain fixed range R in the near-field scan scope _j, R _jDistance between expression and first array element, in the present embodiment, R _j=100m, 110m ..., 1000m, the discrete frequency f of each of bandwidth of interest correspondence _iOn, in the present embodiment, f _i=13,14 ..., 16 at position angle φ _kScan, in the present embodiment, φ _k=0 °, 1 ° ..., 180 °, totally 181 scanning angles utilize the MVDR wave beam to form to focus on localization method to obtain space power spectrum output on each frequency.

Frequency f _iOn array data be X (f _i), frequency f _iOn data covariance matrix be

R(f _i)＝X(f _i)·X(f _i) ^* (21)

When distance is R _jThe time, frequency f _iOn space power spectrum be output as

$P_{f_i}(R_j,{\mathrm{\&phi;}}_k)=\frac{1}{a^{*}R{\left(f_i\right)}^{-1}a},$ φ _k＝0°，1°，…，180° (22)

Wherein,

S=0,1,2 ..., 199, be the guiding vector of array,

Time-delay

d _sThe expression target is respectively and the distance between s+1 array element and the 1st array element poor:

P _sThe position of representing s+1 array element: P _s=sd.

Need to prove, in the step 6,7 and 8, can change about azimuthal scanning sequency of distance, promptly can be for a fixed angle, advanced row distance scans, and then carries out repetitive operation at the different scanning angle.

Fig. 4 is that not adopt focusing on conventional wave beam based on 40 yuan of actual array near fields and forming locating effect figure of expansion towed array technology, Fig. 5 be to adopt the 200 element array near fields based on being obtained for 8 times by 40 yuan of actual array expansions of expansion towed array technology to focus on conventional wave beam to form locating effect figure.As can be seen from Figure 4, target for 300m, 500m and 700m unit place, focus on conventional wave beam formation based on 40 yuan of actual array near fields and not only extract the fall short range information, and azimuth discrimination is also very fuzzy, even can not differentiate two little targets of position angle difference; Fig. 5 then shows, focusing on the resolution characteristic that conventional wave beam forms for the target azimuth based on the near field of 200 element array that obtained for 8 times by the expansion of 40 yuan of actual array improves greatly, and the range information that can roughly find out three targets comes, in fact, the peak value of getting wave beam formation can accurately position target, seems that just range information is not obvious.

Fig. 6 is based on 40 yuan of actual array near fields and focuses on MVDR wave beam formation locating effect figure, and Fig. 7 is based on the 200 element array near fields focusing MVDR wave beam that is obtained by 40 yuan of actual array expansions 8 times and forms locating effect figure.As can be seen from Figure 6, for the target at 300m, 500m and 700m unit place, the azimuthal resolution that focuses on the formation of MVDR wave beam based on 40 yuan of actual array near fields is very high, but the scanning of adjusting the distance is insensitive, can't see the range information of target; Fig. 7 then shows, focusing on the MVDR wave beam based on the 200 element array near fields that obtained by 40 yuan of actual array expansion 8 times forms for higher than 40 yuan of actual array of the resolution of target azimuth, and the range information that can clearly find out three targets comes, simultaneously, getting the peak value that wave beam forms can more accurately position target.

Fig. 4, Fig. 5 and Fig. 6, Fig. 7 contrast as can be seen, and behind employing the present invention, array is to target direction resolution more nearby and accurately all raisings greatly of station-keeping ability.In addition, as can be seen, the performance of near field focusing MVDR wave beam formation location is better than the near field and focuses on conventional wave beam formation location from their contrast.

In fact, for linear array, near field range is:

$r\&le;\frac{{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HSA00000336601500012.png,HSA00000336601500031.png"\; state="\{\{state\}\}">L</figure-callout>}}^2}{\mathrm{\&lambda;}}---\left(23\right)$

Wherein, r represents the radius size of near field range, and L is a basic matrix length, and λ is a signal wavelength.

Formula (23) illustrates, near field range and array aperture out to out square are directly proportional, and are inversely proportional to signal wavelength.The expansion towed array is because its aperture is expanded greatly, and its near field range also increases greatly, so the present invention has enlarged the accurately scope of localizing objects greatly, is the long signal of wavelength for low frequency or very low frequency signal especially.Fig. 4, Fig. 5, Fig. 6 and Fig. 7 have embodied this point just.For 40 yuan of actual array, its length is 39m, target in the simulated conditions is considerably beyond its near field range, and 200 yuan of expansion towed arrays that form for expansion, its length is 199m, near field range expands original 25 times to, and the target in the simulated conditions is all in its near field range, so locating effect is fine.In addition, the increase in aperture also helps the location of weak target and the detection of far field weak target more.

In a word, the present invention can be effectively carries out hi-Fix to target more nearby.

It should be noted last that above embodiment is only unrestricted in order to technical scheme of the present invention to be described.Although the present invention is had been described in detail with reference to embodiment, those of ordinary skill in the art is to be understood that, technical scheme of the present invention is made amendment or is equal to replacement, do not break away from the spirit and scope of technical solution of the present invention, it all should be encompassed in the middle of the claim scope of the present invention.

Claims (6)

1. a near field focus beam forms localization method, and this method obtains the signal of each array element reception of virtual array earlier at frequency band range [f to be processed based on the passive synthetic aperture method of expansion towed array _Min, f _Max] interior frequency domain response, then, utilize the near field focus beam to form localization method target is accurately located; Wherein, frequency band range [f _Min, f _Max] choose according to actual needs, only need satisfy f _Min＞0, f _MaxHalf that is not more than data sampling rate gets final product.

2. near field according to claim 1 focus beam forms localization method, it is characterized in that the passive synthetic aperture method of described expansion towed array comprises that concrete steps are as follows:

Step 1.1): M array element receives the acoustic pressure time-domain signal in the linear array in the towing process, and wherein, M is the linear array array element number, and M is not less than 2 integer; Get the data of ML+1 snap, each snap data comprises M road sound pressure signal, and wherein, ML is the number of times of aperture expansion, and ML＞0 is chosen according to the actual requirements; By the length of half actual array length as an aperture expansion, the overlapping M/2 array element during expansion of promptly each aperture, calculating twice expansion according to formula (1) is time interval between two snaps;

$\mathrm{\&tau;}=\frac{M}{2}\&CenterDot;d/v---\left(1\right)$

Wherein, d is an array element distance, determines according to actual needs, usually d ∈ [0.1m, 10m]; V is the towed array movement velocity, v 〉=3m/s, but carry out the aperture expansion of same number of times, speed is big more, needs the time domain data of accumulation few more;

Step 1.2): adjacent two snaps such as k, k+1 snap acoustic pressure time-domain signal are done Fourier transform, obtain corresponding M road frequency-region signal, the k snap of m array element is pressed formula (2) data conversion process:

$x_{k,m}\left(t\right)\stackrel{\mathrm{FFT}}{\&DoubleRightArrow;}{X}_{k,m}\left(f_i\right),m=\mathrm{1,2},...,M---\left(2\right)$

Wherein, 0＜k≤ML;

Step 1.3): at the frequency band range [f of needs processing _Min, f _Max], according to the M/2 of k snap and k+1 snap overlapping array element is calculated M/2 phase differential by formula (3) and get weighted mean again and obtain phase perturbation ψ _K+1(f _i),

${\mathrm{\&psi;}}_{k+1}\left(f_i\right)=\arg \left\{\frac{\underset{m=1}{\overset{M/2}{\mathrm{\&Sigma;}}}{X}_{k,M/2+m}\left(f_i\right)\&times;X_{k+1,m}^{*}\left(f_i\right)\&times;{\mathrm{\&rho;}}_m\left(f_i\right)}{\underset{m=1}{\overset{M/2}{\mathrm{\&Sigma;}}}{\mathrm{\&rho;}}_m\left(f_i\right)}\right\}---\left(3\right)$

Wherein, ρ _m(f _i) expression m array element frequency f _iCompensating factor, its formula is formula (4):

${\mathrm{\&rho;}}_m\left(f_i\right)=\frac{|\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{X}_{k,M/2+m}\left(f_i\right)\&times;X_{k+1,m}^{*}\left(f_i\right)|}{\sqrt{\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{\left|X_{k,M/2+m}\left(f_i\right)\right|}^2\&times;\underset{i=-Q/2}{\overset{Q/2}{\mathrm{\&Sigma;}}}{\left|X_{k+1,m}^{*}\left(f_i\right)\right|}^2}}---\left(4\right)$

Wherein, Q is the frequency span of compensating factor;

Step 1.4): the different frequency component to k+1 snap the M/2+1～M array element signals carries out phase compensation respectively, it is invented the M+1～M+M/2 array element of k snap, and the like, this M/2 array element signals can virtually become M+M/2 (k-1)+1～M+M/2 (the k-1)+M/2 array element of the 1st snap, that is to say that whenever carrying out an aperture synthesizes, can fictionalize M/2 array element, suc as formula (5):

$X_{k+1,M/2+m}\left(f_i\right)\stackrel{*\exp (-{\mathrm{\&psi;}}_{k+1}\left(f_i\right))}{\&DoubleRightArrow;}{X}_{k,M+m}\left(f_i\right)\stackrel{*\exp (-\underset{j=2}{\overset{k}{\mathrm{\&Sigma;}}}{\mathrm{\&psi;}}_j\left(f_i\right))}{\&DoubleRightArrow;}{X}_{1,M+M/2*(k-1)+m}\left(f_i\right)---\left(5\right);$

Step 1.5): judge whether to have carried out ML aperture and synthesize; If, then virtual obtaining

The signal that individual array element receives is at [f _Min, f _Max] the interior frequency domain response of scope, wherein, MM is the towed array array number after the expansion, is calculated by actual array number and aperture expansion number of times; If not, then repeating step 1.2), step 1.3) and step 1.4).

3. near field according to claim 1 focus beam forms localization method, it is characterized in that, described near field focus beam forms the localization method step and comprises:

Step 2.1): at first get in the near-field scan scope and first array element between distance R _j, wherein, R _j=R ₁, R ₂..., R _NRBe total to NR scanning distance, wherein, R _jGet final product less than near field range, NR＞5 are specifically decided according near field range size and scanning accuracy; Each frequency f in required bandwidth to be processed _iOn, at position angle φ _kScan, wherein, φ _k=φ ₁..., φ _NBBe total to NB scanning angle, NB 〉=181 are specifically determined according to the angle scanning precision; Employing obtains space power spectrum output on each frequency at the near field focus beam formation method of single-frequency point;

Step 2.2): according to formula (6) wave beam on each frequency is formed the result and add up, obtain array in distance R _jLocate the power output on each orientation;

$P(R_j,{\mathrm{\&phi;}}_k)=\underset{f_i=\min }{\overset{\max }{\mathrm{\&Sigma;}}}{P}_{f_i}(R_j,{\mathrm{\&phi;}}_k)---\left(6\right)$

Step 2.3): judging distance R _jWhether scanning is intact; If the towed array two-dimensional power spectrum output P (R on different distance and position angle at last is expanded _j, φ _k), wherein, φ _k=φ ₁..., φ _NB, R _j=R ₁, R ₂..., R _NR, can realize location to target according to this two-dimensional power spectrum; If not, repeating step 2.1) and step 2.2).

4. near field according to claim 3 focus beam forms localization method, it is characterized in that described near field at single-frequency point focuses on conventional wave beam formation method concrete steps and comprises:

Frequency f _iOn array response by shown in the formula (7):

X(f _i)＝[X _1，1(f _i)，X _1，2(f _i)，…，X _1，MM(f _i)] ^T (7)

Frequency f _iOn data covariance matrix by shown in the formula (8):

R(f _i)＝X(f _i)·X(f _i) ^* (8)

When distance is R _jThe time, frequency f _iOn space power spectrum output by shown in the formula (9):

$P_{f_i}(R_j,{\mathrm{\&phi;}}_k)=a^{*}R\left(f_i\right)a,$ φ _k＝φ ₁，…，φ _NB (9)

Wherein,

S=0,1,2 ..., MM-1 is the guiding vector of array; W=2 π f _iTime-delay

d _sThe expression target respectively and the distance between s+1 array element and the 1st array element poor,

P _sThe position of representing s+1 array element, P _s=sd.

5. near field according to claim 3 focus beam forms localization method, it is characterized in that described step 2.1) adopt the focusing localization method that forms based on the MVDR wave beam to obtain space power spectrum output on each frequency.

6. near field according to claim 5 focus beam forms localization method, it is characterized in that described step 2.1) adopt the near field focusing localization method concrete steps that form based on the MVDR wave beam to comprise at single-frequency point:

Frequency f _iOn array response be:

X(f _i)＝[X _1，1(f _i)，X _1，2(f _i)，…，X _1，MM(f _i)] ^T

Frequency f _iOn data covariance matrix be:

R(f _i)＝X(f _i)·X(f _i) ^*

When distance is R _jThe time, frequency f _iOn space power spectrum output type (10):

$P_{f_i}(R_j,{\mathrm{\&phi;}}_k)=\frac{1}{a^{*}R{\left(f_i\right)}^{-1}a},$ φ _k＝φ ₁，…，φ _NB (10)

Wherein,

S=0,1,2 ..., MM-1 is the guiding vector of array; W=2 π f _iTime-delay

d _sThe expression target respectively and the distance between s+1 array element and the 1st array element poor,

P _sThe position of representing s+1 array element, P _s=sd.

Images