CN101504458A - Phase filtering based beam forming method - Google Patents

Abstract

The invention relates to the field of sonar signal processing, in particular to a phase filtration-based beam forming method. In the method, the result of beam forming is subjected to phase filtration according to a phase difference sequence of split beams so as to form a phase filtration-based beam forming output. The novel phase filtration-based beam forming method provided by the invention has the advantages of making full use of phase information, narrowing the major lobes of beams, suppressing reverberation return signals in directions excluding the directions of the main axes of the beams as much as possible, improving the precision of the estimation of the time of arrival (TOA) of following waves, along with simple calculation method, small calculated amount and easy implementation.

Description

A kind of wave beam formation method based on phase filtering

Technical field

The present invention relates to the sonar signal process field, be specifically related to the Array Signal Processing field of multi-beam sonar, mainly is a kind of wave beam formation method based on phase filtering.

Background technology

It is the hot research problem that sonar signal is handled that wave beam forms.Wave beam formation system is the core component of modern sonar system, is that sonar has good tactics, the basis of technical feature.Conventional wave beam formation method (Coventional beamforming is called for short CBF) is the classical way in the Array Signal Processing, and it is simple in structure, is easy to realize, and various rugged environments are had adaptive faculty preferably.

Multi-beam sonar receives the seabed back-scattered signal with same frequency from different directions, its intensity reduces and decays with glancing angle, and change with seabed substrate, angle of inclination, the echoed signal of its seabed sampled point is the scattering wave on the certain area in seabed.Therefore the received beam of conventional beamforming algorithm has certain width, and does not have steep front, has increased the difficulty that follow-up ripple reaches time (TOA) estimation.The wave beam main lobe width is narrowed down, and suppress the wave beam main shaft echoed signal of direction in addition as far as possible, become the multi-beam sonar wave beam and form the problem of being concerned about.

Summary of the invention

Technical matters to be solved by this invention provides a kind of wave beam formation method based on phase filtering that is suitable for multi-beam sonar, makes to suppress the wave beam main shaft reverberating echo signal of direction in addition as far as possible, finally improves the precision that follow-up ripple reaches time (TOA) estimation.

For solving above technical matters, the present invention proposes following technical scheme to realize:

It may further comprise the steps:

Step 1: adopt the M unit received signal of formation reception battle array arbitrarily, the basic matrix received signal is carried out preposition pre-service such as time filtering, A/D conversion, amplification by preposition preprocessor.Utilize Hilbert (Hilbert) conversion of signal, the analytic signal that obtains delayed output signals is:

$\tilde{x}\left(n\right)={[{\tilde{x}}_1\left(n\right),{\tilde{x}}_2\left(n\right),\·\·\·,{\tilde{x}}_M\left(n\right)]}^T---\left(1\right)$

Wherein, Be the 1st, 2 ..., the analytic signal of M delayed output signals, subscript " T " expression transposition.

Step 2: basic matrix is divided into two submatrixs, and these two submatrixs are normally overlapping.The 1st, 2 ..., N array element is formed a battle array A, L, L+1 ..., M array element is formed a battle array B, so two submatrixs separately the analytic signal of output signal be:

${\tilde{x}}_A\left(n\right)={[{\tilde{x}}_1\left(n\right),{\tilde{x}}_2\left(n\right),\·\·\·,{\tilde{x}}_N\left(n\right)]}^T---\left(2\right)$

${\tilde{x}}_B\left(n\right)={[{\tilde{x}}_L\left(n\right),{\tilde{x}}_{L+1}\left(n\right),\·\·\·,{\tilde{x}}_M\left(n\right)]}^T---\left(3\right)$

Step 3: suppose to form the wave beam in θ orientation, the driving vector that two sub-array beams of then whole battle array and A, B form is respectively:

W(θ)＝[w ₁(θ)，w ₂(θ)，…，w _M(θ)] (4)

W _A(θ)＝[w ₁(θ)，w ₂(θ)，…，w _N(θ)] (5)

W _B(θ)＝[w _L(θ)，w _L+1(θ)，…，w _M(θ)] (6)

Wherein, w _j(f, θ)=exp (j2 π f τ _j(θ)) be the driving vector of j array element, τ _j(θ) be the delay compensation of j array element to reference point (generally being made as true origin), j=1,2 ..., N.Then two sub-array beam outputs of the whole battle array in θ orientation and A, B are respectively:

$y(\mathrm{\θ},n)=W\left(\mathrm{\θ}\right)\′*\tilde{x}\left(n\right)---\left(7\right)$

$y_A(\mathrm{\θ},n)=W_A\left(\mathrm{\θ}\right)\′*{\tilde{x}}_A\left(n\right)---\left(8\right)$

$y_B(\mathrm{\θ},n)=W_B\left(\mathrm{\θ}\right)\′*{\tilde{x}}_B\left(n\right)---\left(9\right)$

Step 4: the phase difference sequence that calculates submatrix A and B is:

Wherein, angle[ ] represent sequence is asked phase place, subscript " ^*" expression asks conjugation.When receiving the seabed involuting wave signal, the value of the phase difference sequence of two submatrixs was near zero point.Set a less phase differential threshold delta according to practical situations, make and work as phase difference sequence

(θ, during n) greater than δ, (θ is zero n) to y, and then the wave beam that will try to achieve in θ orientation output result is

The present invention can bring following beneficial effect:

The present invention is according to the phase difference sequence of split beam, and the result that wave beam is formed carries out phase filtering, obtains forming output based on the wave beam of phase filtering.The invention provides a kind of novel, computing method are simple, calculated amount is little, the wave beam formation method that is easy to realize based on phase filtering, made full use of phase information, the main lobe of wave beam is narrowed down, suppressed the wave beam main shaft reverberating echo signal of direction in addition as much as possible, improve the precision that follow-up ripple reaches time (TOA) estimation.

Description of drawings:

Fig. 1: wave beam of the present invention forms FB(flow block);

Fig. 2: 8 yuan of circular arc hydrophone array structural drawing;

Fig. 3: the figure as a result that adopts conventional wave beam to form;

Fig. 4: adopt the figure as a result that forms based on the phase filtering wave beam;

Embodiment:

The present invention will be further described below in conjunction with specific embodiments and the drawings:

Wave beam of the present invention forms FB(flow block) as shown in Figure 1, and this wave beam formation method based on phase filtering specifically may further comprise the steps:

Step 1: adopt the M unit received signal of formation reception battle array arbitrarily, the basic matrix received signal is carried out preposition pre-service such as time filtering, A/D conversion, amplification by preposition preprocessor.Make the array element numbering be respectively 1,2 ..., M, then delayed output signals is:

x(n)＝[x ₁(n)，x ₂(n)，…，x _M(n)] ^T (1)

Wherein, x ₁(n), x ₂(n) ..., x _M(n) be the 1st, 2 ..., the output signal of M array element, subscript " T " expression transposition.

Step 2: utilize Hilbert (Hilbert) conversion of signal, the analytic signal that obtains delayed output signals is:

$\tilde{x}\left(n\right)={[{\tilde{x}}_1\left(n\right),{\tilde{x}}_2\left(n\right),\·\·\·,{\tilde{x}}_M\left(n\right)]}^T---\left(2\right)$

Wherein ${\tilde{x}}_j\left(n\right)=x_j\left(n\right)+j{\hat{x}}_j\left(n\right),$

Be j delayed output signals x _j(n) Hilbert transform.

Step 3: basic matrix is divided into two submatrixs, and these two submatrixs are normally overlapping.The 1st, 2 ..., N array element is formed a battle array A, L, L+1 ..., M array element is formed a battle array B, so two submatrixs separately the analytic signal of output signal be:

${\tilde{x}}_A\left(n\right)={[{\tilde{x}}_1\left(n\right),{\tilde{x}}_2\left(n\right),\·\·\·,{\tilde{x}}_N\left(n\right)]}^T---\left(3\right)$

${\tilde{x}}_B\left(n\right)={[{\tilde{x}}_L\left(n\right),{\tilde{x}}_{L+1}\left(n\right),\·\·\·,{\tilde{x}}_M\left(n\right)]}^T---\left(4\right)$

Step 4: suppose to form the wave beam in θ orientation, the driving vector that two sub-array beams of then whole battle array and A, B form is respectively:

W(θ)＝[w ₁(θ)，w ₂(θ)，…，w _M(θ)] (5)

W _A(θ)＝[w ₁(θ)，w ₂(θ)，…，w _N(θ)] (6)

W _B(θ)＝[w _L(θ)，w _L+1(θ)，…，w _M(θ)] (7)

Wherein, w _j(f, θ)=exp (j2 π f τ _j(θ)) be the driving vector of j array element, τ _j(θ) be the delay compensation of j array element to reference point (generally being made as true origin), j=1,2 ..., N.Then two sub-array beam outputs of the whole battle array in θ orientation and A, B are respectively:

$y(\mathrm{\θ},n)=W\left(\mathrm{\θ}\right)\′*\tilde{x}\left(n\right)---\left(8\right)$

$y_A(\mathrm{\θ},n)=W_A\left(\mathrm{\θ}\right)\′*{\tilde{x}}_A\left(n\right)---\left(9\right)$

$y_B(\mathrm{\θ},n)=W_B\left(\mathrm{\θ}\right)\′*{\tilde{x}}_B\left(n\right)---\left(10\right)$

Step 5: the phase difference sequence that calculates submatrix A and B is:

Wherein, angle Expression is asked phase place to sequence, subscript " ^*" expression asks conjugation.When then receiving the seabed involuting wave signal, the value of the phase difference sequence of two submatrixs was near zero point.Set a less phase differential threshold delta according to practical situations, make and work as phase difference sequence

(θ, during n) greater than δ, (θ is zero n) to y, and then the wave beam in θ orientation output result is

Instantiation adopts 8 yuan of circular arc hydrophone arraies, and structure as shown in Figure 2.The array element received signal is the lakebed echoed signal in certain lake examination, and signal adopts the 10kHz simple signal.The azimuth-range figure of Fig. 3 for adopting conventional wave beam to form.Fig. 4 is for adopting the azimuth-range figure that forms based on the phase filtering wave beam.As seen, method that the present invention gives is better than conventional wave beam formation method, can suppress the wave beam main shaft reverberating echo signal of direction in addition, improves the precision that follow-up ripple reaches time (TOA) estimation.

In addition to the implementation, the present invention can also have other embodiments.All employings are equal to the technical scheme of replacement or equivalent transformation formation, all drop on the protection domain that this achievement requires.

Claims (4)

1, a kind of wave beam formation method based on phase filtering, it is characterized in that: the method includes the steps of:

Step 1: adopt the M unit received signal of formation reception battle array arbitrarily, utilize the Hilbert transform of signal after the preposition pre-service, the analytic signal that obtains delayed output signals is:

$\tilde{x}\left(n\right)={[{\tilde{x}}_1\left(n\right),{\tilde{x}}_2\left(n\right),\·\·\·,{\tilde{x}}_M\left(n\right)]}^T---\left(1\right)$

Wherein,

Be the 1st, 2 ..., the analytic signal of M delayed output signals, subscript " T " expression transposition;

Step 2: basic matrix is divided into two submatrixs, the 1st, 2 ..., N array element is formed a battle array A, L, L+1 ..., M array element is formed a battle array B, so two submatrixs separately the analytic signal of output signal be:

${\tilde{x}}_A\left(n\right)={[{\tilde{x}}_1\left(n\right),{\tilde{x}}_2\left(n\right),\·\·\·,{\tilde{x}}_N\left(n\right)]}^T---\left(2\right)$

${\tilde{x}}_B\left(n\right)={[{\tilde{x}}_L\left(n\right),{\tilde{x}}_{L+1}\left(n\right),\·\·\·,{\tilde{x}}_M\left(n\right)]}^T---\left(3\right)$

Step 3: suppose to form the wave beam in θ orientation, the driving vector that two sub-array beams of then whole battle array and A, B form is respectively:

W(θ)＝[w ₁(θ)，w ₂(θ)，…，w _M(θ)] (4)

W _A(θ)＝[w ₁(θ)，w ₂(θ)，…，w _N(θ)] (5)

W _B(θ)＝[w _L(θ)，w _L+1(θ)，…，w _M(θ)]?(6)

Wherein, w _j(f, θ)=exp (j2 π f τ _j(θ)) be the driving vector of j array element, τ _j(θ) be the delay compensation of j array element to reference point, j=1,2 ..., N, then two sub-array beam outputs of the whole battle array in θ orientation and A, B are respectively:

$y(\mathrm{\θ},n)=W\left(\mathrm{\θ}\right)\′*\tilde{x}\left(n\right)---\left(7\right)$

$y_A(\mathrm{\θ},n)=W_A\left(\mathrm{\θ}\right)\′*{\tilde{x}}_A\left(n\right)---\left(8\right)$

$y_B(\mathrm{\θ},n)=W_B\left(\mathrm{\θ}\right)\′*{\tilde{x}}_B\left(n\right)---\left(9\right)$

Step 4: the phase difference sequence that calculates submatrix A and B is:

Wherein,

Expression is asked phase place to sequence, and conjugation is asked in subscript " * " expression, and when receiving the seabed involuting wave signal, the value of the phase difference sequence of two submatrixs was near zero point, set a phase differential threshold delta, made and worked as phase difference sequence During greater than δ, y (θ is zero n), and then the wave beam that will try to achieve in θ orientation output result is:

2, the wave beam formation method based on phase filtering according to claim 1 is characterized in that: preposition pre-service is meant the processing of the basic matrix received signal being carried out time filtering, A/D conversion, amplification by preposition preprocessor.

3, the wave beam formation method based on phase filtering according to claim 1 is characterized in that: described two submatrixs are overlapping.

4, the wave beam formation method based on phase filtering according to claim 1, it is characterized in that: described true origin is made as reference point.

Images