阅读说明：本技术 自适应波束形成器生成方法及系统 (Method and system for generating adaptive beam former ) 是由 杨鑫 于 2019-10-30 设计创作，主要内容包括：本申请揭示了一种自适应波束形成器生成方法及系统,该方法包括根据采样信号矩阵获取采样协方差矩阵,将所述采样协方差矩阵作为初始协方差矩阵；利用近似最小方差算法计算采样信号的方位谱矩阵和噪声功率矩阵；利用所述方位谱矩阵和所述噪声功率矩阵重构所述初始协方差矩阵；利用重构的协方差矩阵计算自适应波束形成器的加权向量。本申请利用超分辨方位估计得到的方位谱信息,利用阵列模型重构出阵列协方差矩阵,同时与现有较成熟的稳健自适应波束形成方法结合,实现具有强干扰抑制能力的稳健高增益波束形成方法。(The application discloses a method and a system for generating a self-adaptive beam former, wherein the method comprises the steps of acquiring a sampling covariance matrix according to a sampling signal matrix, and taking the sampling covariance matrix as an initial covariance matrix; calculating an azimuth spectrum matrix and a noise power matrix of the sampling signal by using an approximate minimum variance algorithm; reconstructing the initial covariance matrix using the azimuth spectrum matrix and the noise power matrix; the weight vector of the adaptive beamformer is calculated using the reconstructed covariance matrix. The method utilizes the azimuth spectrum information obtained by super-resolution azimuth estimation, utilizes the array model to reconstruct the array covariance matrix, and is combined with the existing mature robust adaptive beam forming method to realize the robust high-gain beam forming method with strong interference suppression capability.)

1. An adaptive beamformer generation method, wherein the adaptive beamformer generation comprises:

acquiring a sampling covariance matrix according to the sampling signal matrix, and taking the sampling covariance matrix as an initial covariance matrix;

calculating an azimuth spectrum matrix and a noise power matrix of the sampling signal by using an approximate minimum variance algorithm;

reconstructing the initial covariance matrix using the azimuth spectrum matrix and the noise power matrix;

the weight vector of the adaptive beamformer is calculated using the reconstructed covariance matrix.

2. The adaptive beamformer generation method of claim 1, wherein the reconstructing the initial covariance matrix using the azimuth spectrum matrix and the noise power matrix comprises:

reconstructing the initial covariance matrix using the azimuth spectrum matrix and the noise power matrix;

iterating the reconstructed covariance matrix, and stopping iteration when the iteration stop condition is met;

the calculating the weighting vector of the adaptive beam former by using the reconstructed covariance matrix comprises:

and calculating the weighting vector of the self-adaptive beam former by using the reconstructed covariance matrix obtained by the last iteration.

3. The adaptive beamformer generation method of claim 1, wherein the iteration stop condition is: and the error between the covariance matrix obtained by the (i + 1) th iteration and the covariance matrix obtained by the (i) th iteration is smaller than a preset error value.

4.The adaptive beamformer generation method of any one of claims 1-3, wherein the covariance matrix is R APA^H+σ²I, the orientation spectrum matrix

the weighting vector of the standard MVDR beam former isThe obtained weighting vector of the adaptive beam former

5. An adaptive beamformer generation system, comprising a processor and a sonar receiver electrically connected to the processor, the sonar receiver configured to receive acoustic signals, the processor configured to perform the adaptive beamformer generation method of any one of claims 1-4.

Technical Field

The invention belongs to the field of design and manufacture of underwater acoustic equipment, and relates to a method and a system for generating a high-gain anti-interference beam former under a complex background.

Background

The range, accuracy and reliability of sonar remote detection and the ability to extract and identify target signals are largely dependent on the capabilities of the underwater acoustic array signal processing techniques, with the key techniques being beamforming and direction of arrival (DOA) estimation. The research on the beam forming technology can effectively inhibit interference, which is beneficial to improving the output signal-to-noise ratio of the array, thereby increasing the sonar working distance and improving the remote sensing capability of the sonar system.

Based on the background, the design of how to pass through the anti-interference wave beam former of high gain under the background of strong interference is researched, and the influence of multi-target interference is furthest suppressed at the signal processing end, and the detection capability of the shore-based sonar to the underwater weak target is improved, so that the method has very important practical significance for improving and upgrading the existing-service shore-based sonar or developing a novel sonar model.

The sonar array signal processing mainly faces the problems of few snapshots, multi-target dense distribution, strong target interference, weak target detection and the like, and under the background, how to inhibit interference to the maximum extent is to design an anti-interference beam forming algorithm to obtain the optimal spatial filtering effect is the core problem of the sonar array beam forming algorithm.

The conventional beam former realizes beam directivity by compensating time delay information among array elements, but the conventional beam forming has a wide beam main lobe, a first side lobe of about-13 dB and no strong interference suppression capability. The self-adaptive beam forming method is a data-driven beam forming method, and the main lobe of a beam pattern is narrower, the side lobe is lower, and the resolution is higher. The most typical adaptive beamforming method is the MVDR beamforming method, however, the algorithm is sensitive to array manifold mismatch, and when array manifold errors occur, the performance of the algorithm is drastically reduced.

Disclosure of Invention

In order to solve the problems in the related art, the application provides a method and a system for forming a high-gain anti-interference beam in a complex background. The technical scheme is as follows:

in a first aspect, the present application provides an adaptive beamformer generation comprising:

acquiring a sampling covariance matrix according to the sampling signal matrix, and taking the sampling covariance matrix as an initial covariance matrix;

calculating an azimuth spectrum matrix and a noise power matrix of the sampling signal by using an approximate minimum variance algorithm;

reconstructing the initial covariance matrix using the azimuth spectrum matrix and the noise power matrix;

the weight vector of the adaptive beamformer is calculated using the reconstructed covariance matrix.

Optionally, the reconstructing the initial covariance matrix by using the azimuth spectrum matrix and the noise power matrix includes:

reconstructing the initial covariance matrix using the azimuth spectrum matrix and the noise power matrix;

iterating the reconstructed covariance matrix, and stopping iteration when the iteration stop condition is met;

the calculating the weighting vector of the adaptive beam former by using the reconstructed covariance matrix comprises:

and calculating the weighting vector of the self-adaptive beam former by using the reconstructed covariance matrix obtained by the last iteration.

Optionally, the iteration stop condition is: and the error between the covariance matrix obtained by the (i + 1) th iteration and the covariance matrix obtained by the (i) th iteration is smaller than a preset error value.

Optionally, the covariance matrix is R ═ APA^H+σ²I, the orientation spectrum matrix

And the noise power matrix

The covariance matrix after reconstruction is

Wherein A is the flow vector a (theta) of the array_s) Forming an array flow matrix, wherein P is the output power of the wave beam, and k is the number of signals;

the weighting vector of the standard MVDR beam former is

The obtained weighting vector of the adaptive beam formerWherein the flow vector a (theta) is arrayed_s) Is composed of

M is the number of array elements.

In a second aspect, the present application also provides an adaptive beamformer generation system comprising a processor and a sonar receiver, the processor being electrically connected to the sonar receiver, the sonar receiver being configured to receive acoustic signals, the processor being configured to perform the adaptive beamformer generation method provided in the first aspect and the various alternatives of the first aspect.

The method utilizes the azimuth spectrum information obtained by super-resolution azimuth estimation, utilizes the array model to reconstruct the array covariance matrix, and is combined with the existing mature robust adaptive beam forming method to realize the robust high-gain beam forming method with strong interference suppression capability.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.

Fig. 1 is a simplified flow diagram of an adaptive beamformer generation method provided in one embodiment of the present invention;

FIG. 2 is an orientation map obtained from CBF, MVDR and SAMV, respectively;

FIG. 3 is a graphical representation of experimental results of a beamformer formed after reconstruction of a sampled covariance matrix;

FIG. 4a is a time azimuth plot of a target plotted using a conventional beamforming algorithm;

FIG. 4b is a time azimuth history plot of the target plotted using the approximate minimum variance algorithm provided herein;

fig. 5 is a diagram of a beam response curve of a conventional MVDR beamformer and a covariance-reconstructed adaptive beamformer.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the invention, as detailed in the appended claims.

In recent years, some steady beam forming algorithms which are strict theoretically are proposed successively, such as a worst performance optimal algorithm, a steady Capon algorithm, an iteration steady beam forming algorithm and the like, the algorithms are proved to be diagonal loading algorithms by nature, a specific mathematical expression of a diagonal loading factor is obtained by using an uncertain set, and the problem that the loading factor is difficult to determine is solved. In order to solve the influence of mismatching of a desired signal guide vector in snapshot data on beam quality, several different feature subspace algorithms are proposed in sequence, and under the condition that the number of signals is known, the algorithm projects the desired signal guide vector to a signal subspace, so that the mismatching degree of the guide vector is reduced by correcting the signal guide vector, but when the signal-to-noise ratio is low, entanglement occurs between the signal subspace and a noise subspace, and the error of the corrected signal guide vector is still large.

In order to thoroughly eliminate the influence of a desired signal in snapshot data, a beam forming algorithm based on interference noise covariance matrix reconstruction is provided. The algorithm reconstructs the interference and noise covariance matrix outside the desired signal direction through the orientation spectrum estimation. The reconstructed matrix does not contain the expected signal, so that the robustness of the algorithm to the mismatching of the expected signal steering vector is improved. The adaptive beam former reconstructed by covariance used in the method can generate grooves in strong interference directions; the conventional MVDR beamformer is affected by errors and environments in actual use, has poor robustness and hardly has the capability of interference suppression.

(1) Robust adaptive beamformer architecture based on covariance matrix reconstruction

The design principle of the standard MVDR beamformer is to keep the signals of the bearing of interest undistorted and minimize the beam output power. The narrowband array signal model can be expressed as

x(t)＝A(Θ_s)s(t)+n(t),t＝1,2,...,T (1)

Array manifold matrix A in formula (theta)_s)＝[a(θ₁),a(θ₂),...,a(θ_K)]K is the number of signals, and for the K signal, the array manifold vector

Assume that the complex weight vector is w ═ w₁,w₂,…,w_M]^TThe beam output sequence y (t) can be expressed as

y(t)＝w^Hx(t),t＝1,2,...,T (2)

The beam output power is

P＝E{y(t)y^*(t)}＝w^HRw (3)

Wherein R ═ E { x (t) x^H(t) }. Thus, according to the design principles of the standard MVDR beamformer, one can obtain:

subject to w^Ha(θ_s)＝1 (4)

in the formula [ theta ]_sIs the direction of incidence of the signal of interest.

The constraint optimization problem in the equations is usually solved using standard Lagrange multiplier techniques. Constructing a cost function of

J＝w^HRw+λ[w^Ha(θ_s)-1](5)

Where λ is the Lagrange multiplier. The formula (5) to w^HDifferentiating and making it zero to obtain

w＝-λR^-1a(θ_s) (6)

Substituting equation (6) for the equation constraint in equation (4) yields the weight vector of the standard MVDR beamformer as

Considering the theoretical formula of the covariance matrix:

R＝APA^H+σ²I (8)

orientation spectrum matrix obtained by using approximate minimum variance algorithm, i.e. in formula

And a noise power matrix, i.e.Substituting the data into a formula (8) to obtain a reconstructed covariance matrix, namely

Substituting the reconstructed covariance matrix into formula (7) to obtain the weighting vector of the high-gain robust adaptive beam former:

the beam is formed using a beamformer having the weighting vector of equation (10).

Based on the above theory and formula, please refer to fig. 1, which is a flowchart of a method for generating an adaptive beamformer provided in an embodiment of the present invention, and the method for generating an adaptive beamformer provided in the present application includes the following steps:

step 101, acquiring a sampling covariance matrix according to a sampling signal matrix, and taking the sampling covariance matrix as an initial covariance matrix;

102, calculating an orientation spectrum matrix and a noise power matrix of a sampling signal by using an approximate minimum variance algorithm;

the orientation spectrum matrix here may be:

the noise power matrix may be:

103, reconstructing an initial covariance matrix by using the azimuth spectrum matrix and the noise power matrix;

when the initial covariance matrix is reconstructed by using the azimuth spectrum matrix and the noise power matrix, the initial covariance matrix is reconstructed by using the azimuth spectrum matrix and the noise power matrix; and iterating the reconstructed covariance matrix, and stopping iteration when the iteration stop condition is met.

In one possible implementation, the iteration stop condition may be: and the error between the covariance matrix obtained by the (i + 1) th iteration and the covariance matrix obtained by the (i) th iteration is smaller than a preset error value.

And step 104, calculating a weighting vector of the adaptive beam former by using the reconstructed covariance matrix.

And calculating the weighting vector of the self-adaptive beam former by using the reconstructed covariance matrix obtained by the last iteration.

(2) Improved measures

1. When the number of snapshots is small, an error exists between an estimated value and a true value of a covariance matrix, so that the direction and the power of an interference signal obtained by traditional Capon spectrum estimation are inaccurate, the discrete degree of the estimated value of noise power is large, the noise characteristic value of the estimated matrix is relatively divergent, the noise characteristic value with the relatively large divergence degree can be brought into a newly reconstructed covariance matrix to influence the beam quality, and in order to solve the problem that the number of snapshots is small and the influence of the inaccurate estimation of the traditional azimuth space spectrum on beam forming, the super-resolution azimuth estimation performance under the conditions of low signal-to-noise ratio and limited number of snapshots is realized by using an approximate minimum variance algorithm of parameter estimation.

2. The reconstruction algorithm needs to perform integral reconstruction in all regions in the range of the incoming wave direction of the undesired signal, the calculated amount is large, and the actual application of the algorithm is influenced. The present invention provides a robust adaptive beamforming algorithm based on sparse interference covariance matrix reconstruction. The new algorithm utilizes the maximum eigenvalue and the noise subspace eigenvalue mean value of the received data matrix as interference and noise power respectively, utilizes the space sparsity of interference signals, performs matrix reconstruction only in the range of the interference incoming wave direction, and finally utilizes the new reconstruction matrix to calculate the weighting vector.

3. In addition, the null formed by the algorithm is narrow, in practical application, due to the fact that an interference source moves rapidly, interference signals are easily moved out of the null, the signal-to-interference-and-noise ratio of output signals is reduced seriously, the estimated interference signal direction is corrected by setting a direction fluctuation parameter, the null is widened through the fluctuation parameter, the sensitivity to the snapshot times and estimation errors is reduced on the basis that the mismatching of the direction of the expected signals is guaranteed to be stable, the formed beam side lobe level is lower, the null is deeper, and the null is widened.

(3) Verification of measured test data

Lake test-effect of inhibiting strong interference of dam power station

Experimental results show that, as shown in fig. 2, CBF (conventional beamforming algorithm) is limited by rayleigh criterion, and spatial azimuth resolution is poor; the applicant tries to adopt a super-resolution azimuth estimation method of Sparse signals, and proposes a Sparse approximation Minimum Variance algorithm(s) with variable factors (shortly referred to as SAMV). The algorithm utilizes a compromise parameter to carry out compromise processing of a maximum likelihood estimation value and sparseness performance, changes an exponential factor of a sparse approximation minimum variance algorithm (SAMV) in an iterative process to obtain an azimuth spectrogram with strong sparseness performance and ultralow sidelobes, realizes super-resolution azimuth estimation and coherent processing performance of adjacent targets, and is suitable for signal processing of an underwater large-scale horizontal array.

Based on a high-resolution orientation estimation algorithm, orientation spectrum information is obtained, a sampling covariance matrix is reconstructed, and the test result is shown in fig. 3 below.

The strong interference target located at the 60-degree azimuth is a dam power station, and compared with the adaptive beam forming, the high-gain anti-interference adaptive beam forming algorithm based on covariance matrix reconstruction obviously has stronger interference suppression capability.

Offshore test-effect of inhibiting multi-target interference

The target is detected using a sensor array. The time azimuth history of the target is plotted using the conventional beamforming algorithm and the approximate minimum variance algorithm used in the present application, respectively, as shown in fig. 4(a) and (b), respectively. The frequency band of the treatment ranges from 80Hz to 300Hz, and the full array is used. Comparing fig. 4(a) and (b), it can be seen that the approximate minimum variance algorithm used in the present item can achieve higher spatial resolution capability, and can also resolve clearly for targets that are very close in space.

According to the test record and the reported data, the weak signal strength target exists in the-94-degree direction, and the rest is strong interference, and a high-gain anti-interference beam former is designed to realize the extraction of the weak target signal. FIG. 5 is a beam response curve of a conventional MVDR beamformer and a covariance-reconstructed adaptive beamformer, which in contrast can be found to produce notches in strong interference orientations, such as-156 deg., -133 deg., -104 deg., -83 deg., and-51 deg., for the covariance-reconstructed adaptive beamformer used in this project; the conventional MVDR beamformer is affected by errors and environments in actual use, has poor robustness and hardly has the capability of interference suppression.

In addition, the present application also provides an adaptive beamformer generating system, which includes a processor and a transceiver, where the processor is electrically connected to the transceiver, the transceiver is used for receiving an acoustic wave signal, and the processor is used for executing the above adaptive beamformer generating method.

Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.

It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is limited only by the appended claims.