CN115469265A - Acoustic vector array joint processing azimuth estimation method - Google Patents

Acoustic vector array joint processing azimuth estimation method Download PDF

Info

Publication number
CN115469265A
CN115469265A CN202211073071.5A CN202211073071A CN115469265A CN 115469265 A CN115469265 A CN 115469265A CN 202211073071 A CN202211073071 A CN 202211073071A CN 115469265 A CN115469265 A CN 115469265A
Authority
CN
China
Prior art keywords
covariance matrix
matrix
theta
vector
observation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211073071.5A
Other languages
Chinese (zh)
Inventor
时胜国
张旭
杨德森
朱晓春
朱中锐
徐付佳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Harbin Engineering University
Original Assignee
Harbin Engineering University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Harbin Engineering University filed Critical Harbin Engineering University
Priority to CN202211073071.5A priority Critical patent/CN115469265A/en
Publication of CN115469265A publication Critical patent/CN115469265A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S3/00—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
    • G01S3/80—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using ultrasonic, sonic or infrasonic waves
    • G01S3/802—Systems for determining direction or deviation from predetermined direction
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S3/00—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
    • G01S3/80—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using ultrasonic, sonic or infrasonic waves
    • G01S3/86—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using ultrasonic, sonic or infrasonic waves with means for eliminating undesired waves, e.g. disturbing noises
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Mathematical Optimization (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Analysis (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computational Mathematics (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)

Abstract

The invention discloses an acoustic vector array joint processing azimuth estimation method, which comprises the steps of establishing an acoustic vector array output signal model, and constructing a covariance matrix R of acoustic pressure and vibration velocity joint processing according to the model V The covariance matrix R V Decomposed into an observation coefficient matrix and a residual covariance matrix R uv And stripping the observation direction from the sound pressure vibration velocity cross covariance matrix, reconstructing a Hermite covariance matrix by decomposing the singular value of the residual covariance matrix, and finally performing space spectrum estimation by using the reconstructed covariance matrix to obtain the estimated direction. The invention avoids the problem that certain azimuth signals are filtered or weakened due to the fixed observation azimuth, does not need to scan the observation azimuth, solves the problem of overhigh algorithm complexity, and increases the noise by decomposing and reconstructing singular values of the sound pressure vibration velocity cross covariance matrixThe anti-noise capability of the acoustic vector array processing is enhanced, and the multi-target resolution and the azimuth estimation performance under the condition of low signal-to-noise ratio are improved.

Description

Acoustic vector array joint processing azimuth estimation method
Technical Field
The invention belongs to the field of underwater acoustic array signal processing, relates to an acoustic vector array combined processing azimuth estimation method, and particularly relates to an acoustic vector array sound pressure and vibration velocity combined processing azimuth estimation method based on covariance matrix decomposition.
Background
The acoustic vector sensor is generally formed by compounding a sound pressure sensor and a particle vibration velocity sensor, and can be used for synchronously picking up sound pressure and particle vibration velocity information in a sound field at a spatial concurrent point. Because the perception sound field information is more comprehensive, the signal processing mode of the acoustic vector sensor array is more diversified, and the acoustic vector sensor array can be basically divided into two types based on a Nehorai processing framework and acoustic pressure vibration velocity combined processing.
An acoustic vector array nehiri Processing framework was proposed by the article nehiri in 1994 (article nehiri, ethylene pad. Acoustics vector-sensor array Processing, J. IEEE trans. Signal Processing,1994,42 (9)), which considers the vibration velocity component output by the acoustic vector sensor as information independent from the sound pressure component, and although this Processing method enables the spatial spectrum estimation method based on the acoustic pressure array to be extended to the acoustic vector array, the performance improvement still is not sufficient to meet the target orientation estimation requirement of the underwater low signal-to-noise ratio environment.
The core idea of the sound pressure and vibration velocity combined processing method is that the space correlation characteristic of the sound pressure and vibration velocity is fully utilized to suppress noise, and Bai Xingyu and the like (Bai Xingyu, jiang Yu and Zhao Chunhui. Acoustic vector array source number detection and orientation estimation [ J ] acoustics report based on sound pressure and vibration velocity combined processing, 2008 (01): 56-61.) realize high-resolution detection and orientation of a remote target by constructing a sound pressure and vibration velocity cross covariance matrix; the straight line image of the Yao expands the concept of sound pressure vibration speed cross covariance matrix (the straight line image of the Yao, hu Jinhua, yao Dongming. An acoustic vector array orientation estimation algorithm [ J ]. Acoustics academic newspaper (Chinese edition), 2008 (04): 305-309.) based on a multiple signal classification method, utilizes the combined directional gain of a vector sensor, further reduces the signal-to-noise ratio threshold of orientation estimation, and obtains better performance in the aspects of multi-target resolving power, resolving probability and the like. However, in the process of constructing the sound pressure vibration velocity cross covariance matrix, the vibration velocity component needs to be projected to a certain observation direction to obtain the combined vibration velocity in the direction, and the sound pressure vibration velocity joint processing method has a certain spatial filtering capability, and designating the observation direction as a certain fixed value may cause some direction signals to be filtered or weakened, thereby causing output signal-to-noise ratio to be reduced or information source to be missed, and if the observation direction is scanned along with the spatial spectrum search direction, the cross covariance matrix is changed, thereby causing the algorithm complexity to be increased dramatically.
Disclosure of Invention
In view of the prior art, the technical problem to be solved by the present invention is to provide an acoustic vector array sound pressure vibration velocity joint processing azimuth estimation method based on covariance matrix decomposition, which strips the observation azimuth from the sound pressure vibration velocity cross covariance matrix, avoids the problem that some azimuth signals are filtered or weakened due to fixed observation azimuth, and solves the problem of excessive algorithm complexity without scanning the observation azimuth.
In order to solve the technical problem, the invention provides an acoustic vector array joint processing azimuth estimation method, which comprises the steps of establishing an acoustic vector array output signal model, and constructing a covariance matrix R of sound pressure and vibration velocity joint processing according to the model V The covariance matrix R V Decomposed into an observation coefficient matrix and a residual covariance matrix R uv And stripping the observation direction from the sound pressure vibration velocity cross covariance matrix, reconstructing a Hermite covariance matrix by decomposing the singular value of the residual covariance matrix, and finally performing space spectrum estimation by using the reconstructed covariance matrix to obtain the estimated direction.
Further, the method for estimating the orientation by jointly processing the acoustic vector arrays comprises the following steps:
step 1, establishing an acoustic vector array output signal model of M array elements in any geometric shape, establishing an x and y reference rectangular coordinate system by taking the array element at the leftmost end of the array as an origin, and taking the mth array element coordinate as (x) m ，y m ) Obtaining the output vector p (n) of sound pressure channel and the output vector v of vibration speed x, y channel of acoustic vector array x (n)，v y (n)：
Figure BDA0003830023520000021
Wherein s (n) = [ s ] 1 (n)，…，s K (n)] T Representing the source vector, K the number of sources, symbol T the transposition operation, a (θ) is the sound pressure array manifold matrix, a (θ) = [ a (θ) = 1 )，a(θ 2 )，…，a(θ K )]，θ k For the azimuth of the kth source, the acoustic pressure array steering vector a (θ) k )＝[1，exp(-j2πf 0 τ 2 )，…，exp(-j2πf 0 τ M )] T ，τ m ＝(x m cosθ k +y m sinθ k ) C, C is the speed of sound, f 0 Is the center frequency of the signal source, [ phi ] vx ＝diag[cos(θ 1 )，…，cos(θ K )]，Φ vy ＝diag[sin(θ 1 )，…，sin(θ K )]Respectively, x, y channel coefficient matrix of vibration speed, n p (n)，n vx (n)，n vy (n) are background noise vectors of x and y channels of sound pressure and vibration velocity respectively;
step 2, combining the output vector v of the vibration speed x and y channels x (n)，v y (n) obtaining an observation azimuth θ r Combined vibration velocity of time:
v c (n)＝cos(θ r )v x (n)+sin(θ r )v y (n)，
constructing a covariance matrix R of sound pressure and vibration velocity combined processing V ，R V Is prepared from (p + v) c )v c The combined mode of the combined treatment is obtained,
Figure BDA0003830023520000022
e {. Denotes the desired operation;
step 3, the covariance matrix R V Decomposition into v c (n) observation coefficient matrix T v (θ)、p(n)+v c (n) observation coefficient matrix T u (theta) with the residual covariance matrix R uv ；
Step 4, for the residual covariance matrix R uv Singular value decomposition to obtain diagonal matrix composed of non-zero singular valuesLambda, and left and right singular vectors U, R composed of Lambda corresponding columns uv ＝UΛV H And selecting U or V to construct a new covariance matrix R C ＝UΛU H Or R C ＝VΛV H ；
Step 5, new guide vector a C (theta) is generated from an observation coefficient matrix T (theta) and a steering vector a (theta), a C (θ)＝T H (theta) a (theta), wherein, when R is C ＝UΛU H When, T (θ) = T u (θ) when R C ＝VΛV H When, T (θ) = T v (θ)；
Step 6, utilizing covariance matrix R C And a guide vector a C And (theta) implementing a spatial spectrum estimation method to obtain an estimated azimuth.
Further, step 3 combines the covariance matrix R V Decomposed into a matrix of observation coefficients T v (θ)、T u (theta) with the residual covariance matrix R uv The method comprises the following specific steps:
step 3-1, setting an observation direction theta r Searching orientation theta for spatial spectrum, extracting v c (n) medium observation coefficient matrix T v (θ)：
Figure BDA0003830023520000031
In the formula (I), the compound is shown in the specification,
Figure BDA0003830023520000032
υ(θ)＝[cos(θ)，sin(θ)] T upsilon (theta) is an array manifold of single vibration velocity sensors;
step 3-2, extracting p (n) + v in the covariance matrix c (n) term observation coefficient matrix T u (θ)：
Figure BDA0003830023520000033
In the formula (I), the compound is shown in the specification,
Figure BDA0003830023520000034
u(θ)＝[1，cos(θ)，sin(θ)] T u (θ) is the array manifold of the single vector sensor;
step 3-3, the sound pressure and the vibration speed are jointly processed to form a covariance matrix R V Decomposed into an observation coefficient matrix and a residual covariance matrix R uv The form of multiplication:
Figure BDA0003830023520000035
in the formula, R uv Is a residual covariance matrix, specifically:
Figure BDA0003830023520000036
further, the spatial spectrum estimation method is a CBF method, an MVDR method, or a MUSIC method.
Further, the spatial spectrum estimation method is an MVDR method, and the spatial spectrum P C The expression (θ) is:
Figure BDA0003830023520000037
further, the covariance matrix R V From pv c The combined mode of the joint processing is obtained, then the covariance matrix R is obtained V Decomposition into v c (n) observation coefficient matrix T v (theta) with the residual covariance matrix R uv 。
Further, the covariance matrix R V From p (p + v) c ) The combined mode of the joint processing is obtained, then the covariance matrix R is obtained V Decomposition into p (n) + v c (n) observation coefficient matrix T u (theta) with the residual covariance matrix R uv 。
Further, the covariance matrix R V Is prepared from (p + v) c ) 2 The combined mode of the joint processing is obtained, then the covariance matrix R is obtained V Decomposition into p (n) + v c (n) observation coefficient matrix T u (θ)、p(n)+v c (n) observation coefficient matrix T u (theta) and residual covariance matrixR uv 。
The invention has the beneficial effects that: the invention provides a new sound vector array sound pressure and vibration velocity combined processing method, which comprises the following steps: aiming at the contradiction between the azimuth estimation performance and the algorithm calculated amount caused by the observation azimuth selection in the traditional sound pressure and vibration velocity combined processing, the observation azimuth is stripped from the sound pressure and vibration velocity cross covariance matrix, the Hermite covariance matrix is reconstructed by decomposing the singular value of the residual covariance matrix, and finally the spatial spectrum estimation algorithm is implemented by using the covariance matrix, so that the problem that some azimuth signals are filtered or weakened due to the fixed observation azimuth is avoided, the observation azimuth does not need to be scanned, and the problem of overhigh algorithm complexity is solved. According to the invention, observation direction does not need to be selected, and the anti-noise capability of acoustic vector array processing is enhanced by decomposing and reconstructing the sound pressure vibration velocity cross covariance matrix, so that the multi-target resolution and direction estimation performance under the condition of low signal-to-noise ratio are improved.
Drawings
FIG. 1 is a flow chart of the algorithm;
FIG. 2 is a model of an acoustic vector uniform linear array;
fig. 3 (a) is a spatial spectrum comparison plot of SNR =0 dB;
FIG. 3 (b) is a spatial spectrum comparison plot of SNR = -5 dB;
FIG. 4 is a graph of the variation of the target resolution probability with the signal-to-noise ratio;
figure 5 root mean square error versus signal to noise ratio.
Detailed Description
The invention is further described with reference to the drawings and examples.
The invention aims at the contradiction that in the traditional sound pressure and vibration velocity combined processing method, the observation azimuth is fixed, which may cause that some azimuth signals are weakened or filtered due to spatial filtering, and the observation azimuth scanning causes the algorithm complexity to be increased violently.
The first embodiment is as follows:
the technical scheme adopted by the invention for solving the technical problem comprises the following steps:
step 1, establishing an acoustic vector array output signal model of M array elements in any geometric shape, establishing an x and y reference rectangular coordinate system by taking the array element at the leftmost end of the array as an origin, and taking the mth array element coordinate as (x) m ,y m ) Obtaining the output vector p (n) of the sound pressure channel and the output vector v of the vibration speed x, y channel of the acoustic vector array x (n),v y (n)：
Figure BDA0003830023520000051
Wherein s (n) = [ s ] 1 (n)，…，s K (n)] T Representing source vector, K is the number of sources, symbol T represents transposition operation, A (theta) is a sound pressure array manifold matrix, and A (theta) = [ a (theta) = 1 )，a(θ 2 )，…，a(θ K )]，θ k For the azimuth of the kth source, the acoustic pressure array steering vector a (θ) k )＝[1，exp(-j2πf 0 τ 2 )，…，exp(-j2πf 0 τ M )] T ，τ m ＝(x m cosθ k +y m sinθ k ) C, C is the speed of sound, f 0 Is the center frequency of the signal source, [ phi ] vx ＝diag[cos(θ 1 )，…，cos(θ K )]，Φ vy ＝diag[sin(θ 1 )，…，sin(θ K )]Respectively, x, y channel coefficient matrix of vibration speed, n p (n)，n vx (n)，n vy (n) background noise vectors of sound pressure and vibration speed x and y channels respectively;
step 2, combining the output vector v of the vibration speed x and y channels x (n)，v y (n) obtaining an observation azimuth θ r Combined vibration velocity v of time c (n)，
v c (n)＝cos(θ r )v x (n)+sin(θ r )v y (n)，
And constructing a covariance matrix R of sound pressure and vibration velocity combined processing V ，R V Can be prepared from but not limited to pv c 、p(p+v c )、(p+v c )v c And (p + v) c ) 2 The combined processing method is to obtain the product with (p + v) c )v c The combination is taken as an example and the combination,
Figure BDA0003830023520000052
e {. Cndot } represents the desired operation;
step 3, the covariance matrix R V Decomposed into an observation coefficient matrix and a residual covariance matrix R uv ；
Step 4, for the residual covariance matrix R uv Singular value decomposition to obtain diagonal matrix Lambda composed of non-zero singular values, and left singular vector U and right singular vector V composed of Lambda corresponding columns, with R uv ＝UΛV H And selecting U or V to construct a new covariance matrix R C ＝UΛU H Or R C ＝VΛV H ；
Step 5, new guide vector a C (theta) is generated from an observation coefficient matrix T (theta) and a steering vector a (theta), where a C (θ)＝T H (theta) a (theta), wherein, when R is C ＝UΛU H When, T (θ) = T u (θ) when R C ＝VΛV H Time, T (θ) = T v (θ)；
Step 6, utilizing covariance matrix R C And a guide vector a C (θ) implement a spatial spectrum estimation method including, but not limited to, the CBF, MVDR, and MUSIC methods, etc., where the spatial spectrum P of the MVDR method C (theta) is expressed as
Figure BDA0003830023520000053
Step 3 covariance matrix R of the invention V The decomposition specifically comprises the following steps:
step 3-1, setting an observation direction theta r Searching orientation theta for spatial spectrum, extracting v c (n) medium observation coefficient matrix T v (θ)
Figure BDA0003830023520000054
In the formula (I), the compound is shown in the specification,
Figure BDA0003830023520000055
υ(θ)＝[cos(θ)，sin(θ)] T upsilon (theta) is an array manifold of single vibration velocity sensors;
step 3-2, extracting p (n) + v in the covariance matrix c (n) term observation coefficient matrix T u (θ)
Figure BDA0003830023520000056
In the formula (I), the compound is shown in the specification,
Figure BDA0003830023520000061
u(θ)＝[1，cos(θ)，sin(θ)] T u (θ) is the array manifold of the single vector sensor;
step 3-3, the sound pressure and the vibration speed are jointly processed to form a covariance matrix R V Decomposed into an observation coefficient matrix and a residual covariance matrix R uv Form of multiplication, wherein (p + v) c )v c Covariance matrix R V Can be decomposed into
Figure BDA0003830023520000062
In the formula, R uv For the remaining covariance matrix it is possible to,
Figure BDA0003830023520000063
example two:
with reference to fig. 1, the present invention comprises the following steps:
step 1, taking an acoustic vector uniform linear array as an example, establishing an array output signal model, as shown in fig. 2, considering a two-dimensional isotropic noise field, wherein the array takes the position of a first array element as a reference point, M array elements are distributed at equal intervals along the positive direction of a y axis, the distance between adjacent array elements is d, and K independent far-field equal-power narrow-band signals existSource, incident angle theta k K =1,2,.. K, defined as the angle between the source and the positive direction of the x-axis, the output vector p (n) of the sound pressure channel and the output vector v of the vibration speed x, y channel x (n)，v y (n) is:
Figure BDA0003830023520000064
wherein s (n) = [ s ] 1 (n)，…，s K (n)] T Representing a source vector; a (θ) is a sound pressure array manifold matrix having
A(θ)＝[a(θ 1 )，a(θ 2 )，…，a(θ K )] (2)
Wherein a (θ) k ) Steering vector on acoustic pressure sensor array for kth source
a(θ k )＝[1，exp(-j2πd sinθ k /λ)，…，exp(-j(M-1)2πd sinθ k /λ)| T (3)
In the formula, λ represents the source incident wavelength, symbol T represents the transposition operation, Φ vx ＝diag[cos(θ 1 )，…，cos(θ K )]，Φ vy ＝diag[sin(θ 1 )，…，sin(θ K )]Respectively, x, y channel coefficient matrix of vibration speed, n p (n)，n vx (n)，n vy (n) are background noise vectors of x and y channels of sound pressure and vibration velocity respectively;
step 2, combining the output vector v of the vibration speed x and y channels x (n)，v y (n) obtaining an observation azimuth θ r Combined vibration velocity v of time c (n)，
v c (n)＝cos(θ r )v x (n)+sin(θ r )v y (n) (4)
Constructing a covariance matrix R of sound pressure and vibration velocity combined processing V ，R V Can be prepared from but not limited to pv c 、p(p+v c )、(p+v c )v c And (p + v) c ) 2 Combined with a joint process to obtain (p + v) c )v c The combination is taken as an example and the combination,
Figure BDA0003830023520000065
wherein E {. Cndot } represents an expected operation;
step 3, the covariance matrix R V Decomposed into an observation coefficient matrix and a residual covariance matrix R uv To (p + v) c )v c Covariance matrix R of combined form V The decomposition steps are as follows:
step 3-1, setting an observation direction theta r Searching orientation theta for spatial spectrum, extracting v c (n) medium observation coefficient matrix T v (θ)
Figure BDA0003830023520000071
In the formula (I), the compound is shown in the specification,
Figure BDA0003830023520000072
υ(θ)＝[cos(θ)，sin(θ)] T upsilon (theta) is an array manifold of single vibration velocity sensors;
step 3-2, extracting p (n) + v in the covariance matrix c (n) term observation coefficient matrix T u (θ)
Figure BDA0003830023520000073
In the formula (I), the compound is shown in the specification,
Figure BDA0003830023520000074
u(θ)＝[1，cos(θ)，sin(θ)] T u (θ) is the array manifold of the single vector sensor;
step 3-3, the sound pressure and the vibration speed are jointly processed to form a covariance matrix R V Decomposed into an observation coefficient matrix and a residual covariance matrix R uv The form of multiplication is as follows
Figure BDA0003830023520000075
In the formula, R uv For the remaining covariance matrix it is possible to,
Figure BDA0003830023520000076
step 4, for the residual covariance matrix R uv The singular value decomposition is carried out, and the singular value decomposition,
R uv ＝U u Λ uv V H (10)
in the formula, Λ uv Is R uv Singular value matrix of (a), which can be written as uv ＝[Λ，0 2M×M ] T Let λ be m Λ = diag (λ) for the mth non-zero singular value 1 ，...，λ 2M ) And λ 1 ≥λ 2 ≥…≥λ 2M (ii) a V is R uv Right singular vector, U u Is R uv The left singular vector U of the column corresponding to the Lambda is taken as,
U＝U u [I 2M×2M ，0 2M×M ] T (11)
in the formula, 0 2M×M Is a 2M × M dimensional zero matrix, I 2M×2M Is a 2 Mx 2M dimensional unit matrix, selects Lambda and singular vectors U or V to construct a new covariance matrix
R C ＝UΛU H Or R C ＝VΛV H (12)
Step 5, new guide vector a C (theta) is generated from the corresponding observation coefficient matrix T (theta) and the steering vector a (theta)
a C (θ)＝T H (θ)a(θ) (13)
Wherein when R is C ＝UΛU H When, T (θ) = T u (θ) when R C ＝VΛV H When, T (θ) = T v (θ)；
Step 6, utilizing covariance matrix R C And a guide vector a C (θ) implementing a spatial spectrum estimation method including, but not limited to, the CBF, MVDR and MUSIC methods, exemplified by the MVDR method, spatial spectrum P C (theta) is expressed as
Figure BDA0003830023520000081
Taking a spatial spectrum P C And (theta) obtaining the far-field target azimuth estimated value by the first K maximum values.
The above description is directed to the embodiments of the present invention, and the following description is directed to the simulation examples.
Assuming that a noise field is isotropic, considering an acoustic vector equally-spaced linear array with the array element number of 8, the array element distance is half wavelength, two uncorrelated far-field equal-power narrow-band signals exist, the incident angles are-4 degrees and 5 degrees respectively, assuming that noise is stable white Gaussian noise, setting the snap-shot number to be 1000, and setting the Monte Carlo test frequency to be 100. Selecting (p + v) c )v c The method comprises the steps of constructing cross covariance matrixes in a combined mode, comparing three conditions that observation directions are set to be 0 degrees and 50 degrees respectively in a conventional method based on a Nehorai processing framework and a traditional sound pressure vibration velocity combined processing method and scanning along with a spectrum searching direction, and comparing the performance of the method, wherein the method selects left singular value vectors to construct a new cross covariance matrix.
Fig. 3 (a) and 3 (b) are graphs comparing the spatial spectrum of each method under the conditions of SNR =0dB and SNR = -5dB, respectively. The method not only has lower spatial spectrum background level and sharper target spectral peaks than other methods, but also can effectively estimate the directions of two targets when the two target spectral peaks are mixed and are difficult to distinguish by other methods under the condition that the signal-to-noise ratio is-5 dB.
Fig. 4 is a graph of the target resolution probability versus the signal-to-noise ratio for each method. Compared with other methods, the method has the advantages that the target successful resolution probability is higher under the condition of low signal to noise ratio, and the target successful resolution probability can reach 1 by using a lower signal to noise ratio threshold.
FIG. 5 is a plot of Root Mean Square Error (RMSE) versus signal to noise ratio for each method. It can be found that the RMSE of each method is gradually reduced with the improvement of the target resolution probability above-6 dB, wherein the performance of the method of the present invention is optimal, when the signal-to-noise ratio is greater than-4 dB, the RMSE of the method of the present invention is lower than 0.5 °, and if the RMSE is close to that when SNR = -4dB, the signal-to-noise ratio required by other methods is greater than 0dB.

Claims (8)

1. A method for estimating the orientation of acoustic vector array joint processing is characterized in that: establishing an acoustic vector array output signal model, and constructing a covariance matrix R of sound pressure and vibration velocity combined processing according to the model V The covariance matrix R V Decomposed into an observation coefficient matrix and a residual covariance matrix R uv And stripping the observation direction from the sound pressure vibration velocity cross covariance matrix, reconstructing a Hermite covariance matrix by decomposing the singular value of the residual covariance matrix, and finally performing space spectrum estimation by using the reconstructed covariance matrix to obtain the estimated direction.
2. The method of claim 1, comprising the steps of:
step 1, establishing an acoustic vector array output signal model of M array elements in any geometric shape, establishing an x and y reference rectangular coordinate system by taking the array element at the leftmost end of the array as an origin, and taking the mth array element coordinate as (x) m ,y m ) Obtaining the output vector p (n) of the sound pressure channel and the output vector v of the vibration speed x, y channel of the acoustic vector array x (n),v y (n)：
Figure FDA0003830023510000011
Wherein s (n) = [ s ] 1 (n),…,s K (n)] T Denotes a source vector, K is the number of sources, symbol T denotes a transposition operation, a (θ) is a sound pressure array manifold matrix, a (θ = [ a (θ) ] 1 ),a(θ 2 ),…,a(θ K )]，θ k For the azimuth of the kth source, the acoustic pressure array steering vector a (θ) k )＝[1,exp(-j2πf 0 τ 2 ),…,exp(-j2πf 0 τ M )] T ，τ m ＝(x m cosθ k +y m sinθ k ) C, C is the speed of sound, f 0 Is the center frequency of the signal source, [ phi ] vx ＝diag[cos(θ 1 ),…,cos(θ K )]，Φ vy ＝diag[sin(θ 1 ),…,sin(θ K )]Respectively, the x and y channel coefficient matrixes of the vibration speed, n p (n),n vx (n),n vy (n) background noise vectors of sound pressure and vibration speed x and y channels respectively;
step 2, combining the output vector v of the vibration speed x and y channels x (n),v y (n) obtaining an observation azimuth θ r Combined vibration velocity of time:
v c (n)＝cos(θ r )v x (n)+sin(θ r )v y (n)，
constructing a covariance matrix R of sound pressure and vibration velocity combined processing V ，R V Is prepared from (p + v) c )v c The combined mode of the combined treatment is obtained,
Figure FDA0003830023510000012
e {. Denotes the desired operation;
step 3, the covariance matrix R V Decomposition into v c (n) observation coefficient matrix T v (θ)、p(n)+v c (n) observation coefficient matrix T u (theta) and residual covariance matrix R uv ；
Step 4, for the residual covariance matrix R uv Singular value decomposition is carried out to obtain a diagonal matrix Lambda consisting of non-zero singular values and a left singular vector U and a right singular vector V, R consisting of Lambda corresponding columns uv ＝UΛV H And selecting U or V to construct a new covariance matrix R C ＝UΛU H Or R C ＝VΛV H ；
Step 5, new guide vector a C (theta) is generated from an observation coefficient matrix T (theta) and a steering vector a (theta), a C (θ)＝T H (theta) a (theta), wherein, when R is C ＝UΛU H When, T (θ) = T u (θ) when R C ＝VΛV H When, T (θ) = T v (θ)；
Step 6, utilizing covariance matrix R C And a guide vector a C (theta) trueAnd applying a space spectrum estimation method to obtain an estimated direction.
3. The method of claim 1, wherein the method comprises: step 3. The covariance matrix R V Decomposed into a matrix of observation coefficients T v (θ)、T u (theta) with the residual covariance matrix R uv The method specifically comprises the following steps:
step 3-1, setting an observation direction theta r Searching orientation theta for spatial spectrum, extracting v c (n) medium observation coefficient matrix T v (θ)：
Figure FDA0003830023510000021
In the formula (I), the compound is shown in the specification,
Figure FDA0003830023510000022
υ(θ)＝[cos(θ)，sin(θ)] T upsilon (theta) is an array manifold of single vibration velocity sensors;
step 3-2, extracting p (n) + v in the covariance matrix c (n) term observation coefficient matrix T u (θ)：
Figure FDA0003830023510000023
In the formula (I), the compound is shown in the specification,
Figure FDA0003830023510000024
u(θ)＝[1,cos(θ)，sin(θ)] T u (θ) is the array manifold of the single vector sensor;
step 3-3, the sound pressure and the vibration speed are jointly processed to form a covariance matrix R V Decomposed into an observation coefficient matrix and a residual covariance matrix R uv The form of multiplication:
Figure FDA0003830023510000025
in the formula, R uv Is a residual covariance matrix, specifically:
Figure FDA0003830023510000026
4. the acoustic vector array joint processing azimuth estimation method according to claim 2, wherein: the spatial spectrum estimation method is a CBF method, an MVDR method or a MUSIC method.
5. The acoustic vector array joint processing azimuth estimation method according to claim 2, wherein: the spatial spectrum estimation method is an MVDR method, and the spatial spectrum P C The expression (θ) is:
Figure FDA0003830023510000027
6. the method of claim 2, wherein the method comprises: the covariance matrix R V From pv c The combined mode of the joint processing is obtained, then the covariance matrix R is obtained V Decomposition into v c (n) observation coefficient matrix T v (theta) with the residual covariance matrix R uv 。
7. The acoustic vector array joint processing azimuth estimation method according to claim 2, wherein: the covariance matrix R V From p (p + v) c ) The combined mode of the joint processing is obtained, then the covariance matrix R is obtained V Decomposition into p (n) + v c (n) observation coefficient matrix T u (theta) with the residual covariance matrix R uv 。
8. The acoustic vector array joint processing azimuth estimation method according to claim 2, wherein: the protocolVariance matrix R V Is prepared from (p + v) c ) 2 The combined mode of the joint processing is obtained, then the covariance matrix R is obtained V Decomposition into p (n) + v c (n) observation coefficient matrix T u (θ)、p(n)+v c (n) observation coefficient matrix T u (theta) and residual covariance matrix R uv 。
CN202211073071.5A 2022-09-02 2022-09-02 Acoustic vector array joint processing azimuth estimation method Pending CN115469265A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211073071.5A CN115469265A (en) 2022-09-02 2022-09-02 Acoustic vector array joint processing azimuth estimation method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211073071.5A CN115469265A (en) 2022-09-02 2022-09-02 Acoustic vector array joint processing azimuth estimation method

Publications (1)

Publication Number Publication Date
CN115469265A true CN115469265A (en) 2022-12-13

Family

ID=84369018

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211073071.5A Pending CN115469265A (en) 2022-09-02 2022-09-02 Acoustic vector array joint processing azimuth estimation method

Country Status (1)

Country Link
CN (1) CN115469265A (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080198695A1 (en) * 2006-08-02 2008-08-21 Ali Abdi System and method for using acoustic field parameters for communication
CN104391268A (en) * 2014-11-21 2015-03-04 哈尔滨工业大学深圳研究生院 Direction-of-arrival angle detection method
CN106199505A (en) * 2016-06-28 2016-12-07 哈尔滨工程大学 The sane direction estimation method in a kind of acoustic vector circle battle array mode territory
CN107132503A (en) * 2017-03-23 2017-09-05 哈尔滨工程大学 Acoustic vector circle battle array broadband coherent source direction estimation method based on vector singular value decomposition
CN110082712A (en) * 2019-03-14 2019-08-02 哈尔滨工程大学 A kind of acoustic vector circle battle array Coherent Targets direction estimation method
CN112130112A (en) * 2020-09-20 2020-12-25 哈尔滨工程大学 Information source number estimation method based on acoustic vector array joint information processing

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080198695A1 (en) * 2006-08-02 2008-08-21 Ali Abdi System and method for using acoustic field parameters for communication
CN104391268A (en) * 2014-11-21 2015-03-04 哈尔滨工业大学深圳研究生院 Direction-of-arrival angle detection method
CN106199505A (en) * 2016-06-28 2016-12-07 哈尔滨工程大学 The sane direction estimation method in a kind of acoustic vector circle battle array mode territory
CN107132503A (en) * 2017-03-23 2017-09-05 哈尔滨工程大学 Acoustic vector circle battle array broadband coherent source direction estimation method based on vector singular value decomposition
CN110082712A (en) * 2019-03-14 2019-08-02 哈尔滨工程大学 A kind of acoustic vector circle battle array Coherent Targets direction estimation method
CN112130112A (en) * 2020-09-20 2020-12-25 哈尔滨工程大学 Information source number estimation method based on acoustic vector array joint information processing

Non-Patent Citations (8)

* Cited by examiner, † Cited by third party
Title
AIFER LIU ET AL.: "Efficient DOA Estimation Method with Ambient Noise Elimination for Array of Underwater Acoustic Vector Sensors", 《2021 IEEE/CIC INTERNATIONAL CONFERENCE ON COMMUNICATIONS IN CHINA (ICCC WORKSHOPS)》, 26 November 2021 (2021-11-26), pages 250 - 255 *
LONGKAI LIANG ET AL.: "Two-demensional DOA estimation method of acoustic vector sensor array based on sparse recovery", 《DIGITAL SIGNAL PROCESSING》, 31 January 2022 (2022-01-31), pages 1 - 6 *
向悠扬 等: "声矢量阵联合信息处理的信源数估计算法", 《哈尔滨工程大学学报》, vol. 43, no. 5, 31 May 2022 (2022-05-31), pages 706 - 711 *
安妍妍;李赢;时胜国;时洁;: "声矢量圆阵宽带相干信号的方位估计", 南京大学学报(自然科学), no. 04, 30 July 2017 (2017-07-30), pages 7 - 14 *
杨德森 等: "声矢量圆阵相位模态域目标方位估计", 《声学学报》, vol. 39, no. 1, 31 January 2014 (2014-01-31), pages 19 - 25 *
白兴宇;姜煜;赵春晖;: "基于声压振速联合处理的声矢量阵信源数检测与方位估计", 声学学报(中文版), no. 01, 15 January 2008 (2008-01-15), pages 58 - 63 *
白兴宇;杨德森;赵春晖;: "基于声压振速联合信息处理的声矢量阵相干信号子空间方法", 声学学报(中文版), no. 05, 10 September 2006 (2006-09-10), pages 28 - 35 *
白兴宇;杨德森;赵春晖;: "声矢量阵远程定向技术(一)――新的协方差矩阵生成方法", 信号处理, no. 03, 30 June 2006 (2006-06-30), pages 88 - 92 *

Similar Documents

Publication Publication Date Title
CN111239678B (en) Two-dimensional DOA estimation method based on L-shaped array
CN112180329B (en) Automobile noise source acoustic imaging method based on array element random uniform distribution spherical array deconvolution beam forming
CN107561484B (en) Direction-of-arrival estimation method based on interpolation co-prime array covariance matrix reconstruction
CN110927661A (en) Single-basis expansion co-prime array MIMO radar DOA estimation method based on MUSIC algorithm
CN107544051A (en) Wave arrival direction estimating method of the nested array based on K R subspaces
CN107703478B (en) Extended aperture two-dimensional DOA estimation method based on cross-correlation matrix
CN109507636B (en) Direction-of-arrival estimation method based on virtual domain signal reconstruction
CN109375227B (en) Deconvolution beam forming three-dimensional acoustic imaging method
CN110837076A (en) Tensor decomposition-based vector hydrophone array orientation estimation method
CN111736118A (en) Linear array expansion method
CN111693947A (en) Improved MUSIC method based on co-prime array DOA estimation
CN109541526B (en) Circular ring array azimuth estimation method by using matrix transformation
CN112130112B (en) Information source number estimation method based on acoustic vector array joint information processing
CN110632605B (en) Wide-tolerance large-aperture towed linear array time domain single-beam processing method
CN115469265A (en) Acoustic vector array joint processing azimuth estimation method
CN112087235A (en) Sparsity self-adaptive DOA estimation method and system based on pseudo-inverse perception dictionary
CN114563760B (en) Second-order super-beam forming method, equipment and medium based on SCA array
CN108830843B (en) Ultrasonic phased array sparse full-focusing imaging method based on array element sparse optimization and new edge guiding interpolation
CN110927663A (en) Three-dimensional compressed sensing dimension reduction method for near-field sound source parameter estimation
CN114371441A (en) Virtual array direction of arrival estimation method, device, product and storage medium
CN110632579B (en) Iterative beam forming method using subarray beam domain characteristics
CN109061564B (en) Simplified near-field positioning method based on high-order cumulant
CN114184999B (en) Method for processing generated model of cross-coupling small-aperture array
CN113420411B (en) High-resolution narrowband DOA estimation algorithm for wireless signals and implementation method
CN114755627A (en) Compressed sensing and minimization processing combined co-prime area array two-dimensional DOA estimation method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination