阅读说明：本技术 一种中心对称声矢量圆阵快速空间谱压缩超分辨方位估计方法 (A kind of quick spatial spectrum of central symmetry acoustic vector circle battle array compresses super-resolution direction estimation method ) 是由 时洁 李赫颖 杨德森 时胜国 张宇涵 李志超 朱中锐 柳艾飞 于 2019-09-25 设计创作，主要内容包括：本发明提供一种中心对称声矢量圆阵快速空间谱压缩超分辨方位估计方法,首先将声矢量圆阵上、下半个圆环的传感器分别按特定顺序排列,采集声信号,构造出满足<Image he="64" wi="356" file="DDA0002214284940000011.GIF" imgContent="drawing" imgFormat="GIF" orientation="portrait" inline="no"></Image>的导向矢量；然后根据声矢量传感器的平均声强抗噪原理,采用声压振速联合处理的方法构建出协方差矩阵,降低了矢量阵导向矢量矩阵维度同时改善了抗噪能力；随后根据真实源和与真实源相差180度的虚拟源的空间谱构造出一个新的空间谱函数进行半谱搜索,实现了对声矢量圆阵空间谱的压缩；最后通过本发明构造的判别式快速地判别出真实的声源方位。本发明首次实现了对矢量圆阵的空间谱进行压缩,在保证高分辨率的同时极大的提高了运算效率,实现了声矢量圆阵的快速高效的超分辨率方位估计。(The present invention provides a kind of quick spatial spectrum compression super-resolution direction estimation method of central symmetry acoustic vector circle battle array, first the sensor of acoustic vector circle upper and lower half of the annulus of battle array is pressed particular order arrangement respectively, acquires acoustical signal, construct satisfaction Steering vector；Then according to the average sound intensity anti-noise principle of acoustic vector sensors, covariance matrix is constructed using the method for acoustic pressure vibration velocity Combined Treatment, vector array steering vector matrix dimensionality is reduced while improving anti-noise ability；The spatial spectrum that the virtual source of 180 degree is then differed according to real source and with real source constructs a new space spectral function and carries out half spectrum search, realizes the compression to acoustic vector circle battle array spatial spectrum；True sound bearing is rapidly determined finally by the discriminate that the present invention constructs.The present invention realizes for the first time compresses the spatial spectrum of vector circle battle array, greatly improves operation efficiency while guaranteeing high-resolution, realizes the super-resolution orientation estimation rapidly and efficiently of acoustic vector circle battle array.)

1. a kind of quick spatial spectrum of central symmetry acoustic vector circle battle array compresses super-resolution direction estimation method, it is characterised in that: step It is as follows:

Step 1: acoustic vector circle battle array is divided into two semicircular rings up and down, collects vector circle battle array upper half annulus counterclockwise Sensor received signal data are collected the sensor received signal data of vector circle battle array lower half annulus by clock-wise order, Then entire acoustic vector circle battle array is constructed in the receipt signal matrix of t moment sound pressure channel and the direction x, y vibration velocity channelWith

Step 2: carrying out azimuth sweep, constructs respectively to the sensor of upper half annulus and lower half annulus and receives signal phase Then corresponding steering vector synthesizes the steering vector matrix of entire acoustic vector circle battle arrayAt this time To any one azimuth angle theta, steering vector matrix meets:

Wherein,WithRespectively indicate the sound pressure channel of acoustic vector circle battle array and the guiding in the direction x, y channel Vector, " * " represent conjugate operation；

Step 3: two vibration velocity components are weighted linear combination, using azimuth scanning angle θ as the guidance side of rotary electronic Position obtains joint vibration velocity of the acoustic vector circle battle array on channeling direction θAre as follows:

Step 4: dimension-reduction treatment is carried out to vector array covariance matrix data according to average sound intensity anti-noise principle, construction orientation is swept Retouch covariance matrix R when angle is θ_pv(θ),

Wherein, expectation is asked in E [] expression, and subscript " H " indicates conjugate transposition operation；

Step 5: the space spectral function P suitable for the quick spatial spectrum compression super-resolution orientation estimation of acoustic vector circle battle array is constructed (θ)；

Step 6: using the space spectral function P (θ) of construction, azimuth spectrum functional value corresponding to scanning angle θ is obtained；

Step 7: step 2 is repeated to step 6, in θ ∈ [0, π] or θ_kHalf spectrum scanning is carried out in the range of ∈ [π, 2 π], is drawn The image of half space spectrum out, obtains spectrum peak position θ by observation space spectrogram picture_k；

Step 8: θ is utilized_kThe position θ of another spectral peak entire spatial spectrum [0,2 π] range Nei is calculated_k ^*, calculation formula are as follows:

Step 9: spectral peak angle information brings conventional MVDR algorithm spatial spectrum letter at two obtained by being composed in searching for vector circle battle array half In number, two corresponding functional value P (θ are obtained_k) and P (θ_k ^*):

Step 10: two functional values compared are differentiated by following formula:

The position for determining real sources in spectral peak from two, determines the arrival bearing of sound source.

2. a kind of quick spatial spectrum of central symmetry acoustic vector circle battle array according to claim 1 compresses super-resolution orientation estimation side Method, it is characterised in that: step 5 specifically:

(1) by the covariance matrix R after dimensionality reduction_pv(θ) introduces MVDR Beam-former, obtains the space spectral function of real sources Are as follows:

Wherein, subscript " -1 " indicates inversion operation；

(2) met using steering vector matrix after reconstruction signalThis relationship is obtained in [0,2 π] model The space spectral function of 180 ° of virtual source is differed in enclosing with real sources incident angle are as follows:

(3) spatial spectrum of real source and virtual source is fitted, constructs the space spectral function for being suitable for half spectrum search:

Technical field

The present invention relates to a kind of quick spatial spectrums of central symmetry acoustic vector circle battle array to compress super-resolution direction estimation method, belongs to Underwater Vector signal processing field.

Background technique

Vector hydrophone can measure the acoustic pressure and particle velocity information of sound field simultaneously, and bring for simplest linear array The fuzzy ability in left and right is differentiated, is can be realized without fuzzy positioning in the total space the orientation in, is improved the detection of acoustic pressure battle array and estimate energy Power has many advantages, such as.Circle battle array has a wide range of applications in Sonar system, and uniform circular array can get 360 ° omni-directional without fuzzy Azimuth information and approximately uniform angular resolution, have compared with even linear array and have an enormous advantage.But and line of vector Battle array great successes are compared, and the research of related vector circle battle array is less.

Yang Desen et al. utilizes phase mode shift theory, and vector sound field is decomposed into a series of mutually orthogonal phase modes State proposes acoustic vector circle battle array phase mode domain method for estimating target azimuth；Ye, Zhongfu et al. utilize central symmetry circle battle array Symmetry characteristic, propose space average theory, have studied irrelevant target orientation estimation performance；Open the artificial solution base such as common vetch It cannot be directly used to this problem of uniform circular array in even linear array vector reconstruction method, propose that the reconstruct of one mode space vector is calculated Method；Wang Yichuan et al. proposes that vector circle battle array time domain parses MVDR algorithm, has studied the original of vector circle battle array time domain parsing MVDR algorithm Reason and implementation process；Shi Jie et al. obtains width by the way that the received acoustic pressure of vector circle battle array and vibration velocity data are carried out sub-bands decomposition Band focuses covariance matrix, proposes a kind of acoustic vector circle battle array robust wideband MVDR direction estimation method；Shi Shengguo et al. basis The architectural characteristic of central symmetry acoustic vector circle battle array, in time domain by acoustic pressure vibration velocity Combined Treatment theory and front and back to space average algorithm Combine, propose it is a kind of based on front and back to the acoustic vector of space average circle battle array direction estimation method.Above method is in improvement sound Noiseproof feature, robustness and the resolution ratio aspect of vector circle battle array orientation estimation achieve preferable effect, but due to above method Calculation amount in practical applications is huge, so being not directly applicable sonar set.Do not have yet at present suitable for acoustic vector circle The Fast Azimuth estimation method of battle array.

Based on this background, the present invention provides a kind of quick spatial spectrums of central symmetry acoustic vector circle battle array to compress super-resolution Direction estimation method.Using the centre symmetry of acoustic vector circle battle array, the sensor of acoustic vector circle upper and lower half of the annulus of battle array is distinguished It is arranged by particular order, acquires acoustical signal, construct satisfactionSteering vector；Then it is passed according to acoustic vector The average sound intensity anti-noise principle of sensor constructs covariance matrix using the method for acoustic pressure vibration velocity Combined Treatment, reduces vector Battle array steering vector matrix dimensionality further improves computational efficiency, while improving anti-noise ability；Then according to real source and with The spatial spectrum of the virtual source of real source difference 180 degree constructs a new space spectral function and carries out half spectrum search, in half of sky Between scan for can be obtained the azimuth information of entire spatial spectrum in spectrum, realize the compression to acoustic vector circle battle array spatial spectrum, protecting Spectrum peak search efficiency is improved while demonstrate,proving high-resolution；It is rapidly determined really finally by the discriminate that the present invention constructs Sounnd source direction.The method achieve the super-resolution orientation estimation of acoustic vector circle battle array rapidly and efficiently, can be satisfied with acoustic vector circle The Fast Azimuth of battle array estimates application demand, can be applied in the signal processings of types sonar such as underwater buoy sonar, carrier-borne sonar.

Summary of the invention

The purpose of the present invention is the symmetry using acoustic vector circle battle array, provide a kind of center using spatial spectrum compress technique The quick spatial spectrum of symmetrical acoustic vector circle battle array compresses super-resolution direction estimation method.

The object of the present invention is achieved like this: steps are as follows:

Step 1: acoustic vector circle battle array is divided into two semicircular rings up and down, collects vector circle battle array upper semi-circle counterclockwise The sensor received signal data of ring are collected the sensor received signal number of vector circle battle array lower half annulus by clock-wise order According to then constructing entire acoustic vector circle battle array in the receipt signal matrix of t moment sound pressure channel and the direction x, y vibration velocity channel With

Step 2: carrying out azimuth sweep, constructs respectively to the sensor of upper half annulus and lower half annulus and receives letter Then number corresponding steering vector synthesizes the steering vector matrix of entire acoustic vector circle battle array At this time to any one azimuth angle theta, steering vector matrix meets:

Wherein,WithRespectively indicate sound pressure channel and the direction x, y channel of acoustic vector circle battle array Steering vector, " * " represent conjugate operation；

Step 3: two vibration velocity components are weighted linear combination, drawing using azimuth scanning angle θ as rotary electronic Orientation is led, joint vibration velocity of the acoustic vector circle battle array on channeling direction θ is obtainedAre as follows:

Step 4: dimension-reduction treatment, construction side are carried out to vector array covariance matrix data according to average sound intensity anti-noise principle Covariance matrix R when bit scan angle is θ_pv(θ),

Wherein, expectation is asked in E [] expression, and subscript " H " indicates conjugate transposition operation；

Step 5: the spatial spectrum letter suitable for the quick spatial spectrum compression super-resolution orientation estimation of acoustic vector circle battle array is constructed Number P (θ)；

Step 6: using the space spectral function P (θ) of construction, azimuth spectrum functional value corresponding to scanning angle θ is obtained；

Step 7: step 2 is repeated to step 6, in θ ∈ [0, π] or θ_kHalf spectrum scanning is carried out in the range of ∈ [π, 2 π], The image for drawing out half space spectrum, obtains spectrum peak position θ by observation space spectrogram picture_k；

Step 8: θ is utilized_kThe position θ of another spectral peak entire spatial spectrum [0,2 π] range Nei is calculated_k ^*, calculate public Formula are as follows:

Step 9: spectral peak angle information brings conventional MVDR algorithm space at two obtained by being composed in searching for vector circle battle array half In spectral function, two corresponding functional value P (θ are obtained_k) and P (θ_k ^*):

Step 10: two functional values compared are differentiated by following formula:

The position for determining real sources in spectral peak from two, determines the arrival bearing of sound source.

The invention also includes structure features some in this way:

1. step 5 specifically:

(1) by the covariance matrix R after dimensionality reduction_pv(θ) introduces MVDR Beam-former, obtains the spatial spectrum letter of real sources Number are as follows:

Wherein, subscript " -1 " indicates inversion operation；

(2) met using steering vector matrix after reconstruction signalThis relationship, obtains [0,2 π] differed with real sources incident angle in range 180 ° virtual source space spectral function are as follows:

(3) spatial spectrum of real source and virtual source is fitted, constructs the space spectral function for being suitable for half spectrum search:

Compared with prior art, the beneficial effects of the present invention are: the present invention, which is realized, receives signal to acoustic vector circle battle array Quick super-resolution signal processing solves current vector circle array signals processing method calculation amount is larger, computational efficiency obtains not To the problem of improving aspect, can be applied in the signal processings of types sonar such as underwater buoy sonar, carrier-borne sonar.1) this method The compression to circle battle array spatial spectrum is realized for the first time, and the orientation letter that entire spectral space is obtained in search result is composed in vector circle battle array half Breath, significantly reduces spectrum peak search calculation amount, improves the efficiency of the orientation estimation of vector circle battle array, has in engineer application huge Big advantage；2) this method takes full advantage of the advantage of vector array element and MVDR algorithm in resolution ratio, has when orientation is estimated Higher resolution ratio；3) vector array is compared has significant advantage compared with acoustic pressure battle array, can significantly improve resolution ratio suppressed sidelobes, Constructing covariance matrix not only using acoustic pressure vibration velocity Combined Treatment mode can be improved vector array computational efficiency but also has preferable anti- It makes an uproar performance；4) this method is targetedly solved suitable for the signal processing of the types sonars such as underwater buoy sonar, carrier-borne sonar The problem of calculating inefficiency in the orientation estimation certainly with the underwater acoustic system of a burst of type of circle.

Detailed description of the invention

Fig. 1 is acoustic vector circle battle array measurement model.

Fig. 2 is the flow chart of the quick spatial spectrum compression super-resolution direction estimation method of central symmetry acoustic vector circle battle array.

It composes using tradition MVDR method and the method for the present invention half in acoustic pressure battle array when Fig. 3 is simple sund source and searches for resulting azimuth spectrum Figure.

The method of the present invention composes for acoustic pressure battle array and vector array half and searches for resulting orientation spectrogram when Fig. 4 is simple sund source.

It composes using tradition MVDR method and the method for the present invention half in acoustic pressure battle array when Fig. 5 is double sound source and searches for resulting azimuth spectrum Figure.

The method of the present invention composes for acoustic pressure battle array and vector array half and searches for resulting orientation spectrogram when Fig. 6 is double sound source.

Specific embodiment

Present invention is further described in detail with specific embodiment with reference to the accompanying drawing.

The Array Model of the method for the present invention as shown in Figure 1, signal processing flow figure as shown in Fig. 2, specific embodiment such as Under:

Central symmetry vector circle battle array is divided into two semicircular rings up and down, pressed by the first step using the symmetry of acoustic vector circle battle array Sequence counter-clockwise collects the sensor received signal data of vector circle battle array upper half annulus, collects vector circle battle array by clock-wise order Then the sensor received signal data of lower half annulus construct entire acoustic vector circle battle array in t moment sound pressure channel and x, the side y To the receipt signal matrix of vibration velocity channelWithIt is implemented as follows:

1) vector circle battle array measurement model is as shown in Figure 1, radius is r, by the two-dimentional acoustic vector sensors group of M first (M is even number) At acoustic vector circle battle array be located in xoy plane, acoustic vector circle battle array on M sensor counterclockwise successively be labeled as No. 1, No. 2 to No. M.The center of circle is overlapped with coordinate origin, and two vibration velocity channel directions of acoustic vector sensors are respectively along x and y-axis Positive direction.

Assuming that N number of narrow band signal s₁(t),…,s_n(t),…,s_N(t) from the direction of xoy plane is parallel to respectively with incidence Angle, θ₁(t),…,θ_n(t),…,θ_N(t) it is incident to acoustic vector circle battle array, obtains two-dimentional acoustic vector circle battle array in m-th of array element of t moment Received acoustic pressure and the direction x channel, the direction y channel vibration velocity signal are respectively P_m(t) and V_xm(t)、V_ym(t)。

2) 1,2 is pressed ..., the sequence counter-clockwise of M/2 collects the sensor received signal data of vector circle battle array upper half annulus:

In formula, P_up(t) and V_xup(t)、V_yup(t) sensor for respectively indicating two-dimentional acoustic vector circle battle array upper half annulus is received Acoustic pressure and the direction x, the direction y channel vibration velocity component.

3) M ... is then pressed, the clock-wise order of M/2+1 collects the sensor received signal of vector circle battle array lower half annulus Data:

In formula, P_down(t) and V_xdown(t)、V_ydown(t) sensor for respectively indicating two-dimentional acoustic vector circle battle array lower half annulus connects The acoustic pressure of receipts and the direction x, the direction y channel vibration velocity component.

4) entire acoustic vector circle battle array is constructed in the receipt signal matrix in each channel of t moment are as follows:

Wherein,WithThe sound pressure signal of entire acoustic vector circle battle array after respectively indicating reconstruct and the side x To the receipt signal matrix of the, direction y vibration velocity channel, subscript " T " indicates transposition operation.

Second step carries out azimuth sweep, and θ is any one azimuth in scanning range, to upper half annulus and lower half circle The sensor of ring constructs steering vector corresponding with signal is received respectively, then synthesizes the guiding arrow of entire acoustic vector circle battle array Moment matrixIt is implemented as follows:

1) for upper half ring configuration steering vector are as follows:

For lower half ring configuration steering vector are as follows:

In formula, A_up(θ) and A_xup(θ)、A_yup(θ) respectively indicate acoustic vector circle battle array upper semi-circle ring sensor sound pressure channel and X, the corresponding steering vector of the direction y channel receiving signal, A_down(θ) and A_xdown(θ)、A_ydown(θ) respectively indicates acoustic vector circle battle array The sound pressure channel and the corresponding steering vector of the direction x, y channel receiving signal of lower half circle ring sensor, a_m(θ)、a_ym(θ)、a_ym(θ) Indicate the sound pressure channel and the corresponding steering vector of the direction x, y channel receiving signal of m-th of array element, in which:

In formula,For the angle of m bugle call vector sensor and positive direction of the x-axis, k=2 π f/c is Wave number, c are the velocity of sound, and f is signal frequency.

2) the steering vector matrix of entire acoustic vector circle battle array is then constructed are as follows:

According to:

This critical nature derives it is found that any one azimuth angle theta, and the steering vector of construction is all satisfied:

Wherein, " * " indicates conjugate operation.

Two vibration velocity components are weighted linear combination, drawing using azimuth scanning angle θ as rotary electronic by third step Orientation is led, joint vibration velocity of the acoustic vector circle battle array on channeling direction θ is obtainedAre as follows:

4th step carries out dimension-reduction treatment, construction side to vector array covariance matrix data according to average sound intensity anti-noise principle Covariance matrix R when bit scan angle is θ_pvAre as follows:

Wherein, expectation is asked in E [] expression, and subscript " H " indicates conjugate transposition operation.

Theoretically,

Wherein: ψ_c=diag (cos (θ₁-θ),cos(θ₂-θ),…,cos(θ_n- θ)), diag () expression is made of () Diagonal matrix, N_c(t)=N_x(t)cos(θ_r)+N_y(t)sin(θ_r) it is the noise signal for combining vibration velocity signal, N (t), N_x(t)、 N_y(t) sound pressure channel is respectively represented, the noise signal that the direction x, the direction y vibration velocity channel receive.Array is in isotropic In noise field, theoretically for uniform vector circle battle array, the noise signal that acoustic pressure and vibration velocity unit receive be it is incoherent, i.e.,Therefore the covariance matrix R of this method construction_pvHave in isotropic noise field preferable anti- It makes an uproar performance.

5th step constructs the spatial spectrum letter suitable for the quick spatial spectrum compression super-resolution orientation estimation of acoustic vector circle battle array Number P (θ), is implemented as follows:

1) by the covariance matrix R after dimensionality reduction_pv(θ) introduces MVDR Beam-former, obtains the spatial spectrum letter of real sources Number are as follows:

Wherein, subscript " -1 " indicates inversion operation.

2) met using steering vector matrix after reconstruction signalThis relationship is obtained at [0,2 π] The space spectral function of 180 ° of virtual source is differed in range with real sources incident angle are as follows:

3) spatial spectrum of real source and virtual source is fitted, constructs the space spectral function for being suitable for half spectrum search:

Meet A (θ)=A (θ+π) according to circle battle array steering vector^*Known to this property: P (θ)==P (θ+π), therefore utilize This spatial spectrum scans for, and the azimuth information of entire spatial spectrum can be compressed in half space spectrum.

6th step finds out azimuth spectrum functional value corresponding to scanning angle θ using the space spectral function P (θ) of construction.

7th step repeats second step to the 6th step, in θ ∈ [0, π] or θ_kHalf spectrum scanning is carried out in the range of ∈ [π, 2 π], The image for drawing out half space spectrum, obtains spectrum peak position θ by observation space spectrogram picture_k。

8th step, utilizes θ_kEntire spatial spectrum θ is calculated_kThe position θ of another spectral peak in ∈ [0,2 π] range_k ^*, meter Calculate formula are as follows:

9th step brings conventional MVDR algorithm space into spectral peak angle information at gained two in the half spectrum search of vector circle battle array In spectral function, two corresponding functional value P (θ are obtained_k) and P (θ_k ^*):

Tenth step, two functional values compared, is differentiated by following formula:

Spectral peak can be generated in real angle on the spatial spectrum of conventional MVDR algorithm, and will not be generated at virtual source angle Spectral peak differentiate that according to above formula, can distinguish vector circle battle array half composes real source that search obtains and virtual according to this property Source, so that it is determined that the true arrival bearing of sound source.

The specific embodiment of summary of the invention is described in detail above.By above step, this method is realized The spatial spectrum of vector circle battle array is compressed, the side of entire space spectral space has been obtained by vector circle battle array half space spectrum search Position information, improves computational efficiency while guaranteeing high-resolution, while having stronger anti-isotropic noise ability, real The super-resolution orientation estimation of acoustic vector circle battle array rapidly and efficiently is showed.The present invention is done below by emulation experiment and is further retouched It states.

Example one: simple sund source treatment effect analysis

One parameter setting of example is as follows: 8 yuan of uniform circular arrays, and circle battle array radius R=0.5m, frequency of source is the narrowband sound of 1kHz Signal is incident on round battle array from far field, and incident angle is 220 °, and number of snapshots 500, step-size in search is 0.5 °, signal-to-noise ratio 10dB, Noise is white Gaussian noise.Fig. 3 show distinct methods under simple sund source and carries out the orientation that half spectrum search obtains in [0,180 °] Spectrogram, wherein the curve of band " * " represents the conventional MVDR method for using acoustic pressure array element, and the curve with "+", which represents, uses acoustic pressure battle array The method of the present invention when first, i.e., quick MVDR method.Fig. 4 show distinct methods under simple sund source and carries out half spectrum in [0,180 °] Search for obtained orientation spectrogram, wherein the method for the present invention when curve of band " * " is represented using acoustic pressure array element, the curve with "+" Be represented as present invention ultimately provides the quick MVDR method based on acoustic vector sensor array.

Example two: double sound source treatment effect analysis

Two parameter setting of example is as follows: 8 yuan of uniform circular arrays, and circle battle array radius R=0.5m, frequency is the narrowband acoustical signal of 1kHz It is incident on round battle array from far field, incident direction is 220 ° and 260 °, and number of snapshots 500, step-size in search is 0.5 °, and signal-to-noise ratio is 10dB, noise are white Gaussian noise.Fig. 5 show distinct methods under double sound source and carries out what half spectrum search obtained in [0,180 °] Orientation spectrogram, wherein the curve of band " * " represents the conventional MVDR method for using acoustic pressure array element, and the curve with "+", which represents, uses sound The method of the present invention when member of keeping a situation well under one's control, i.e., quick MVDR method.Fig. 6 show distinct methods under double sound source and carries out in [0,180 °] The orientation spectrogram that half spectrum search obtains, wherein the method for the present invention when curve of band " * " is represented using acoustic pressure array element, with "+" Curve be represented as present invention ultimately provides the quick MVDR method based on acoustic vector sensor array.

Example three: orientation estimated efficiency analysis

Three parameter setting of example is as follows: 8 yuan of uniform acoustic vector circle battle arrays, circle battle array radius R=0.5m, frequency is two of 1kHz The irrelevant acoustical signal in narrowband is incident on round battle array from far field, and number of snapshots 500, signal-to-noise ratio 10dB, noise is white Gaussian noise, is The reliability for guaranteeing simulation result, has carried out 200 Monte Carlo Experiments.Using tic, toc function in MATLAB, record is calculated Two methods of the time that method is run in computer, comparison different lower DOA for searching rope Walk long in acoustic vector circle battle array estimate institute's used time Between, as a result as shown in the table:

Two methods of vector circle battle array DOA estimates the time (unit/s) used

Example one and two shows all may be used in acoustic vector circle battle array and acoustic pressure circle battle array using quick MVDR algorithm provided by the invention Obtain complete, accurate azimuth information in half spectrum search to realize, resolution ratio is almost the same with routine MVDR method, with compared with High resolution ratio；Fig. 3 and Fig. 5 in example show that when incident angle is not in the spectral limit of search, conventional MVDR algorithm is aobvious It cannot so be composed half and estimate aspect in search, and quick MVDR algorithm provided by the invention can obtain in half spectrum search The true incident direction of sound source can be sought by differentiate spectral peak angle the 7th~9 step of progress to azimuth information；Fig. 4 and Fig. 6 shows to compare compared with acoustic pressure battle array using its noiseproof feature of acoustic pressure vibration velocity combination treatment method with significant excellent in vector circle battle array Gesture, can obvious suppressed sidelobes fluctuating raising spatial resolution；Example three shows that this method can significantly improve acoustic vector circle battle array side The efficiency of position estimation method realizes the Fast Azimuth estimation of acoustic vector circle battle array.

Comprehensive three examples can be seen that the method for the present invention and improve acoustic vector circle battle array while guaranteeing high-resolution Computational efficiency significantly improves the performance of acoustic vector circle battle array orientation estimation.The present invention successfully provides a kind of high-resolution, calculating The Fast Azimuth estimation method of high-efficient central symmetry acoustic vector circle battle array, solves in current vector circle array signals processing method The problem of resolution ratio constantly improves but calculation amount is increasing, computational efficiency cannot improve.

To sum up, the present invention is to provide a kind of quick spatial spectrum compression super-resolution orientation of central symmetry acoustic vector circle battle array to estimate Meter method.First the sensor of acoustic vector circle upper and lower half of the annulus of battle array is pressed into particular order arrangement respectively, acquires acoustical signal, structure Produce satisfactionSteering vector；Then it according to the average sound intensity anti-noise principle of acoustic vector sensors, uses The method of acoustic pressure vibration velocity Combined Treatment constructs covariance matrix, reduces vector array steering vector matrix dimensionality while improving Anti-noise ability；The spatial spectrum that the virtual source of 180 degree is then differed according to real source and with real source constructs a new space Spectral function carries out half spectrum search, realizes the compression to acoustic vector circle battle array spatial spectrum；The discriminate constructed finally by the present invention Rapidly determine true sound bearing.The present invention realizes for the first time compresses the spatial spectrum of vector circle battle array, is guaranteeing Operation efficiency is greatly improved while high-resolution, the super-resolution orientation rapidly and efficiently for realizing acoustic vector circle battle array is estimated Meter.