Target orientation estimation method, system and equipment based on single vector hydrophone

Abstract

The invention discloses a target orientation estimation method, system and equipment based on a single-vector hydrophone, and belongs to the technical field of underwater acoustic signal processing. The method comprises the following steps: acquiring a sound pressure signal in a sound pressure channel of the single-vector hydrophone and a particle vibration velocity signal in a three-dimensional vibration velocity channel; constructing a matrix based on the sound pressure signal and the mass point vibration velocity signal; calculating weights, wherein the weights comprise sound energy flow weights and spectral entropy weights; introducing an adaptive coefficient to fuse the sound energy flow weight and the spectral entropy weight to obtain a joint weight; weighting the covariance matrix through the joint weight to obtain a feature R, and performing decomposition and dimension reduction on the feature R to obtain a dimension-reduced covariance matrix; and constructing an MVDR beam former based on the dimensionality-reduced covariance matrix, and carrying out azimuth spectrum scanning so as to carry out target azimuth estimation. According to the method, the energy information and the spectrum structure information of the vector hydrophone are fused, and the noise is suppressed in combination with the dimension reduction technology, so that high-precision target orientation estimation in a complex environment is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of underwater acoustic signal processing, and in particular to a target direction estimation method, system and device based on a single vector hydrophone, which is particularly suitable for target detection and direction estimation in a complex marine environment. BACKGROUND

[0002] In underwater acoustic engineering, target direction estimation is one of the core technologies of passive sonar systems. A single vector hydrophone integrates a sound pressure sensor and an orthogonal particle velocity sensor, can simultaneously acquire sound pressure signals and three-dimensional velocity signals, and has vector field detection capability, which has a natural advantage over traditional scalar hydrophones in a low signal-to-noise ratio environment.

[0003] The traditional minimum variance distortion response (MVDR) algorithm suppresses noise by optimizing the weight vector, but its performance depends on accurate covariance matrix estimation. In the actual marine environment, factors such as spatial correlation of noise, multipath effect, Doppler shift caused by target motion, etc. will cause covariance matrix estimation error, and thus reduce the direction estimation accuracy. In the existing improved method, the acoustic energy flow weighted MVDR algorithm uses the cross-spectral information of sound pressure and velocity to enhance the signal direction energy, but ignores the signal spectral structure characteristics; while the dimension reduction method based on feature decomposition can suppress noise, but does not fully combine the multi-dimensional information of the vector hydrophone.

[0004] Therefore, there is a need for a direction estimation method, system and device that can fuse the acoustic energy flow energy characteristics and signal spectral structure characteristics, and effectively reduce the noise subspace interference, to improve the target direction estimation accuracy in complex environments. SUMMARY

[0005] The present application provides a target direction estimation method, system and device based on a single vector hydrophone, which fuses the energy information and spectral structure information of the single vector hydrophone, combines the dimension reduction technology to suppress noise, and realizes high-precision target direction estimation in a complex environment.

[0006] The technical solution of the present application to solve the above technical problems is as follows: In a first aspect, the present application provides a target direction estimation method based on a single vector hydrophone.

[0007] A target direction estimation method based on a single vector hydrophone, comprising: Collecting sound pressure signals in the sound pressure channel of the single vector hydrophone and particle velocity signals in the three-dimensional velocity channel, and constructing an initial signal matrix based on the sound pressure signals and the particle velocity signals; Calculating weights, the weights including acoustic energy flow weights and spectral entropy weights; Introducing an adaptive coefficient to fuse the acoustic energy flow weights and the spectral entropy weights to obtain joint weights; The covariance matrix is ​​weighted by joint weights to obtain the feature R. The feature R is decomposed and its dimension reduced to obtain the dimension-reduced covariance matrix. An MVDR beamformer is constructed based on a dimension-reduced covariance matrix, and an azimuth spectrum scan is performed to estimate the target azimuth.

[0008] Furthermore, the calculation of the acoustic energy flow weights includes: The sound pressure signal and vibration velocity signal are converted to the frequency domain, and the components of the sound energy flow in the x and y directions are calculated by cross spectrum. Calculate the acoustic energy flow intensity and normalize it to obtain the acoustic energy flow weight.

[0009] Furthermore, the calculation of spectral entropy weights includes: The spectral structure characteristics of a sound pressure signal are characterized using its spectral entropy. The power spectrum of the sound pressure signal is... ; The power spectral probability density is ; In the formula, The sound pressure spectrum; The spectral entropy H is calculated using the information entropy formula, which is: ; The spectral entropy H is mapped to weights using an exponential function; these are the spectral entropy weights. The calculation formula is: ; In the formula, As a regulating factor, The maximum value of the spectral entropy. M represents the number of effective frequency points.

[0010] Furthermore, the introduction of adaptive coefficients to fuse the acoustic energy flow weight and spectral entropy weight to obtain the joint weight includes: Introducing adaptive coefficients , Where SNR represents the signal-to-noise ratio; This coefficient Based on the dynamic adjustment of the signal-to-noise ratio (SNR), at low SNR, the weight approaches 1, emphasizing acoustic energy flow; at high SNR, the weight approaches 0, emphasizing spectral entropy. The final joint weight is: ; In the formula, This represents the acoustic energy flow weight.

[0011] Furthermore, the step of weighting the covariance matrix using joint weights to obtain the feature R, and then decomposing and reducing the dimensionality of the feature R to obtain the dimensionality-reduced covariance matrix includes: The covariance matrix characterizes the statistical properties of multi-channel signals. The feature R is obtained by weighting the covariance matrix using joint weights, and the calculation formula is as follows: ; In the formula, It is a frequency signal vector; For regularization terms; Perform eigenvalue decomposition on R. And sort the eigenvalues ​​in descending order; Principal component K is selected based on the cumulative energy criterion; In the formula, Represents feature value; threshold Take a value of 0.85~0.95; The top K principal eigenvectors by energy percentage are retained. and corresponding eigenvalues Reconstruct the dimension-reduced covariance matrix to obtain the dimension-reduced covariance matrix. ,Right now: ; In the formula, This represents the eigenvector.

[0012] Furthermore, the construction of the MVDR beamformer based on the dimensionality-reduced covariance matrix and the azimuth spectrum scanning include: Calculate the MVDR weight vector In the formula, Plane waveguide vector, The target location; The formula for calculating azimuth spectrum scanning is: .

[0013] Secondly, the present invention provides a target orientation estimation system based on a single-vector hydrophone.

[0014] A target orientation estimation system based on a single-vector hydrophone includes: The signal acquisition module is used to acquire the sound pressure signal in the single-vector hydrophone sound pressure channel and the particle velocity signal in the three-dimensional vibration velocity channel, and to construct a matrix based on the sound pressure signal and the particle velocity signal. The weight calculation module is used to calculate the acoustic energy flow weight and the spectral entropy weight; The weight fusion module is used to fuse acoustic energy flow weights and spectral entropy weights to obtain joint weights; The feature dimensionality reduction module is used to reduce the dimensionality of the features to obtain the reduced covariance matrix. The target azimuth estimation module is used to construct an MVDR beamformer based on a dimension-reduced covariance matrix and perform azimuth spectrum scanning to estimate the target azimuth.

[0015] Thirdly, the present invention provides an electronic device.

[0016] An electronic device includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the target orientation estimation method based on a single-vector hydrophone as described above.

[0017] In summary, compared with the prior art, the beneficial effects of the above technical solution are: (1) The joint weight fusion proposed in this invention innovatively combines acoustic energy flow weight and spectral entropy weight. Acoustic energy flow weight can effectively capture the energy propagation direction of the signal in space by calculating the cross spectrum of sound pressure and particle velocity, and strengthen the energy contribution of the target signal in a specific location. Spectral entropy weight, on the other hand, measures the orderliness of the signal from the perspective of the signal spectrum structure by quantifying the concentration of frequency domain energy distribution. In complex marine environments, background noise often exhibits different energy distributions and spectral characteristics than the target signal. Traditional single-weighting methods struggle to simultaneously enhance the directional energy and optimize the spectral structure of the signal. The joint weight fusion method of this invention dynamically adjusts the ratio of the two weights through adaptive coefficients: in low signal-to-noise ratio (SNR) environments, the acoustic energy flow weight dominates, prioritizing the enhancement of the signal's directional energy and improving the algorithm's ability to capture weak signals; in high SNR environments, the spectral entropy weight plays a greater role, further refining the signal by suppressing the disordered spectral structure of noise. This complementary mechanism enables the algorithm to comprehensively respond to the directional characteristics of signals in complex backgrounds, significantly improving the accuracy and robustness of azimuth estimation.

[0018] (2) By utilizing spectral entropy theory, effective noise suppression is achieved through precise measurement of the concentration of signal energy distribution in the frequency domain. Specifically, spectral entropy quantifies the degree of order in the signal spectrum by calculating the probability distribution entropy value of the signal power spectrum. Target signals typically have a relatively concentrated energy distribution, exhibiting a low spectral entropy value; while noise, due to its randomness and broadband characteristics, has a high spectral entropy value. Traditional methods are prone to misjudging strong noise, mistaking noise energy for signal energy. However, the spectral entropy weighting of this invention transforms the spectral entropy value into weighting coefficients through exponential mapping, which can keenly identify the disordered spectral structure of noise and dynamically reduce the weight of noise frequency components.

[0019] (3) Traditional MVDR algorithms rely solely on the covariance matrix of the received signal for weight vector calculation, failing to fully exploit the multidimensional information potential of a single-vector hydrophone. This invention, through joint weight fusion, utilizes both the spatial energy information of acoustic energy flow and the frequency domain structure information of spectral entropy to extract and enhance signals from multiple dimensions. This comprehensive utilization of multidimensional information enables the algorithm to more fully describe the characteristics of the target signal, effectively compensating for the shortcomings of traditional methods in terms of insufficient information utilization.

[0020] Traditional methods often employ fixed weighting strategies or parameter settings when facing complex and ever-changing marine environments, making it difficult to adapt to different signal-to-noise ratios, noise characteristics, and target motion states. This invention introduces adaptive coefficients to dynamically adjust the joint weights and adaptively selects principal components based on energy thresholds during eigenvalue decomposition and dimensionality reduction, giving the algorithm strong environmental adaptability.

[0021] Traditional noise suppression methods are typically based on a single energy criterion or simple statistical properties, which can easily damage the signal while suppressing noise. The spectral entropy weighted noise suppression and eigenvalue decomposition dimensionality reduction techniques of this invention deeply suppress noise from the perspectives of signal spectral structure and covariance matrix feature space, respectively. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating a target orientation estimation method based on a single-vector hydrophone, according to an embodiment of the present invention. Detailed Implementation

[0023] The principles and features of the present invention are described below with reference to all the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0024] This invention discloses a target orientation estimation method, system, and device based on a single-vector hydrophone.

[0025] In a first aspect, embodiments of the present invention provide a target orientation estimation method based on a single-vector hydrophone.

[0026] Reference Figure 1 A target orientation estimation method based on a single-vector hydrophone includes the following steps: Acquire sound pressure signals from the single-vector hydrophone sound pressure channel and vibration velocity signals from the three-dimensional vibration velocity channel, and construct an initial signal matrix based on the sound pressure signals and vibration velocity signals; Calculate the weights, which include acoustic energy flow weights and spectral entropy weights; An adaptive coefficient is introduced to fuse the acoustic energy flow weight and the spectral entropy weight to obtain a joint weight; The covariance matrix is ​​weighted by joint weights to obtain the feature R. The feature R is decomposed and its dimension reduced to obtain the dimension-reduced covariance matrix. An MVDR beamformer is constructed based on a dimension-reduced covariance matrix, and an azimuth spectrum scan is performed to estimate the target azimuth.

[0027] The method provided by this invention achieves high-precision target orientation estimation in complex environments by fusing the energy information and spectral structure information of a vector hydrophone and combining it with dimensionality reduction techniques to suppress noise.

[0028] The following is a detailed explanation of each step.

[0029] S1: Acquire the sound pressure signal in the single-vector hydrophone sound pressure channel and the particle velocity signal in the three-dimensional vibration velocity channel, and construct the initial signal matrix based on the sound pressure signal and the particle velocity signal.

[0030] Specifically, the received signals of the single-vector hydrophone include the sound pressure signal in the sound pressure channel and the particle velocity signal in the three-dimensional vibration velocity channel. An initial signal matrix x(t) is constructed based on the sound pressure signal and the particle velocity signal, denoted as... In the formula, It represents the sound pressure signal, reflecting the pressure change of the sound wave; This represents the particle vibration velocity in the x-direction; This represents the particle vibration velocity in the y-direction; This represents the particle vibration velocity in the z-direction. According to plane wave theory, the target orientation is... At that time, the plane waveguide vector is: Therefore, we only need to consider the two-dimensional case.

[0031] S2: Calculate the weights.

[0032] Specifically, the weights include acoustic energy flow weights and spectral entropy weights.

[0033] S1.1: Calculate the acoustic energy flow weight.

[0034] Sound energy flow represents the sound energy passing through a unit area perpendicular to the direction of sound wave propagation per unit time. It is a vector and can be represented as a sound pressure signal. With particle velocity The product of, i.e.: .

[0035] The average acoustic energy flow refers to the time-averaged vector magnitude of the acoustic energy flow at a point in the sound field. The steady-state energy distribution is obtained through short-time integration, i.e.: .

[0036] The specific calculation includes the following sub-steps: S1.1.1: Convert the sound pressure signal and vibration velocity signal to the frequency domain, and calculate the components of the acoustic energy flow in the x and y directions using cross-spectrum analysis, expressed as follows: and In the formula, Represents the sound pressure spectrum. This represents the vibration velocity spectrum in the x-direction. This represents the vibration velocity spectrum in the y-direction; Indicates taking the real part, Indicates conjugate.

[0037] S1.1.2: Calculate the acoustic energy flow intensity and normalize the acoustic energy flow intensity to obtain the acoustic energy flow weight.

[0038] Acoustic intensity is expressed as ; Normalize the acoustic energy flow intensity to obtain the acoustic energy flow weight. The calculation formula is: This weight reflects the energy distribution of the signal at different frequency points; the higher the energy of the frequency component, the greater its weighting percentage.

[0039] S2.1: Calculate the spectral entropy weights.

[0040] Spectral entropy can measure the uniformity of signal power distribution, and specifically includes the following steps: S2.1.1: The spectral structure characteristics of the sound pressure signal are characterized using the spectral entropy of the sound pressure signal. The power spectrum of the sound pressure signal is... The power spectral probability density is In the formula, The sound pressure spectrum; S2.1.2: Calculate the spectral entropy H according to the information entropy formula. The formula is as follows: ; Spectral entropy characterizes the degree of disorder in a signal's spectrum; ordered signals have lower spectral entropy, while noise has spectral entropy close to its maximum value. M is the number of effective frequency points; S2.1.3: Map the spectral entropy H to weights using an exponential function; Specifically, spectral entropy weights The calculation formula is: In the formula, As an adjustment factor, the weight decay rate is controlled; the lower the spectral entropy, the higher the weight, thereby suppressing the influence of noise. This represents the maximum value of the spectral entropy.

[0041] S3: Introduce adaptive coefficients to fuse acoustic energy flow weights and spectral entropy weights to obtain joint weights.

[0042] Specifically, the acoustic energy flow weight and spectral entropy weight are fused to fully leverage the advantages of both weights, and an adaptive coefficient is introduced. The acoustic energy flow weight and spectral entropy weight are integrated, where the adaptive coefficients are... SNR stands for Signal-to-Noise Ratio, i.e. or Adaptive coefficient The adjustment is made dynamically based on the signal-to-noise ratio (SNR). At low SNR... Approaching 1, the emphasis is on acoustic energy flow weighting; at high signal-to-noise ratios, Approaching 0, the weighting emphasizes spectral entropy, and the final joint weight is: .

[0043] S4: Weight the covariance matrix using joint weights to obtain the feature R, decompose the feature R to reduce its dimensionality, and obtain the dimensionality-reduced covariance matrix.

[0044] The covariance matrix characterizes the statistical properties of multi-channel signals, specifically: S4.1: The feature R is obtained by weighting the covariance matrix using joint weights, and is expressed as: In the formula, It is a frequency signal vector. This is a regularization term to avoid matrix singularities; S4.2: Decompose the feature R, i.e. And sort the eigenvalues ​​in descending order; S4.3: Select the principal component K according to the cumulative energy criterion, i.e. In the formula, Represents feature value; threshold Take a value of 0.85~0.95; S4.4: Retain the top K principal eigenvectors by energy percentage and corresponding eigenvalues Reconstruct the dimension-reduced covariance matrix to obtain the dimension-reduced covariance matrix. ,Right now: ; In the formula, This represents the eigenvector.

[0045] S5: Construct an MVDR beamformer based on the dimensionality-reduced covariance matrix and perform azimuth spectrum scanning to estimate the target azimuth.

[0046] Specifically, the covariance matrix based on dimensionality reduction reconstruction Construct a minimum variance distortionless response (MVDR) beamformer; Calculate the MVDR weight vector This weighting ensures maximum gain in the direction of the target signal while suppressing noise. Scanning azimuth spectrum: .

[0047] The following results were obtained through simulation, as shown in Table 1: Table 1 Simulation Results

[0048] This invention innovatively combines acoustic energy flow weighting and spectral entropy weighting. Acoustic energy flow weighting, by calculating the cross-spectrum of sound pressure and particle velocity, effectively captures the energy propagation direction of the signal in space, enhancing the energy contribution of the target signal in a specific location. Meanwhile, spectral entropy weighting, from the perspective of the signal's spectral structure, measures the orderliness of the signal by quantifying the concentration of energy distribution in the frequency domain.

[0049] In complex marine environments, background noise often exhibits different energy distributions and spectral characteristics than the target signal. Traditional single-weighting methods struggle to simultaneously enhance the directional energy and optimize the spectral structure of the signal. The joint weighting fusion module of this invention dynamically adjusts the ratio of the two weights through adaptive coefficients: in low signal-to-noise ratio (SNR) environments, the acoustic energy flow weight dominates, prioritizing the enhancement of the signal's directional energy and improving the algorithm's ability to capture weak signals; in high SNR environments, the spectral entropy weight plays a greater role, further refining the signal by suppressing the disordered spectral structure of noise. This complementary mechanism enables the algorithm to comprehensively respond to the directional characteristics of signals in complex backgrounds, significantly improving the accuracy and robustness of azimuth estimation.

[0050] By leveraging spectral entropy theory, effective noise suppression is achieved through precise measurement of the concentration of energy distribution in the signal's frequency domain. Specifically, spectral entropy quantifies the degree of order in the signal spectrum by calculating the probability distribution entropy value of the signal power spectrum. Target signals typically have a relatively concentrated energy distribution, exhibiting a low spectral entropy value; while noise, due to its randomness and broadband characteristics, has a high spectral entropy value. Traditional methods are prone to misjudging high-energy noise, mistaking noise energy for signal energy. However, the spectral entropy weighting of this invention transforms the spectral entropy value into weighting coefficients through exponential mapping, enabling it to accurately identify the disordered spectral structure of noise and dynamically reduce the weight of noise frequency components.

[0051] Traditional MVDR algorithms rely solely on the covariance matrix of the received signal for weight vector calculation, failing to fully exploit the multidimensional information potential of a single-vector hydrophone. This invention, through a joint weight fusion module, utilizes both the spatial energy information of the acoustic energy flow and the frequency domain structure information of the spectral entropy to extract and enhance signal features from multiple dimensions. This comprehensive utilization of multidimensional information allows the algorithm to more fully describe the characteristics of the target signal, effectively overcoming the shortcomings of traditional methods in terms of insufficient information utilization.

[0052] Traditional methods often employ fixed weighting strategies or parameter settings when facing complex and ever-changing marine environments, making it difficult to adapt to different signal-to-noise ratios, noise characteristics, and target motion states. This invention introduces adaptive coefficients to dynamically adjust the joint weights and adaptively selects principal components based on energy thresholds during eigenvalue decomposition and dimensionality reduction, giving the algorithm strong environmental adaptability.

[0053] Traditional noise suppression methods are typically based on a single energy criterion or simple statistical properties, which can easily damage the signal while suppressing noise. The spectral entropy weighted noise suppression module and eigenvalue decomposition dimensionality reduction technique of this invention deeply suppress noise from the perspectives of the signal spectral structure and the covariance matrix feature space, respectively.

[0054] Secondly, embodiments of the present invention also provide a target orientation estimation system based on a single-vector hydrophone.

[0055] A target orientation estimation system based on a single-vector hydrophone includes: The signal acquisition module is used to acquire the sound pressure signal in the single-vector hydrophone sound pressure channel and the particle velocity signal in the three-dimensional vibration velocity channel, and to construct a matrix based on the sound pressure signal and the particle velocity signal. The weight calculation module is used to calculate the acoustic energy flow weight and the spectral entropy weight; The weight fusion module is used to fuse acoustic energy flow weights and spectral entropy weights to obtain joint weights; The feature dimensionality reduction module is used to reduce the dimensionality of the features to obtain the reduced covariance matrix. The target azimuth estimation module is used to construct an MVDR beamformer based on a dimension-reduced covariance matrix and perform azimuth spectrum scanning to estimate the target azimuth.

[0056] Thirdly, embodiments of the present invention also provide an electronic device.

[0057] An electronic device includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the target orientation estimation method based on a single-vector hydrophone as described above.

[0058] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A single vector hydrophone based target bearing estimation method, characterized in that, The method comprises the following steps: Collecting sound pressure signals in a sound pressure channel of a single vector hydrophone and particle velocity signals in a three-dimensional velocity channel, and constructing an initial signal matrix based on the sound pressure signals and the particle velocity signals; Calculating weights, including sound energy flow weights and spectral entropy weights; Introducing adaptive coefficients to fuse the sound energy flow weights and the spectral entropy weights to obtain joint weights; Weighting the covariance matrix by the joint weights to obtain a feature R, decomposing the feature R to reduce dimensions, and obtaining a reduced covariance matrix; Constructing an MVDR beamformer based on the reduced covariance matrix and performing bearing spectrum scanning to estimate the target bearing.

2. The target direction estimation method based on single vector hydrophone according to claim 1, characterized in that: The calculation of the sound energy flow weights comprises the following steps: Converting the sound pressure signals and the velocity signals to the frequency domain, and calculating the components of the sound energy flow in the x and y directions by using the cross spectrum; Calculating the sound energy flow intensity, normalizing the sound energy flow intensity, and obtaining the sound energy flow weights.

3. The target direction estimation method based on single vector hydrophone according to claim 2, characterized in that: The calculation of the spectral entropy weights comprises the following steps: The signal spectrum structure feature is characterized by using sound pressure signal spectrum entropy, wherein the power spectrum of the sound pressure signal is ; The power spectral probability density is ; In the formula, is the sound pressure spectrum; Calculating the spectral entropy H according to the information entropy formula, and the calculation formula is: ； The spectral entropy H is mapped to a weight by an exponential function, the spectral entropy weight The calculation formula is: ； In the formula, is a modulation factor, is a maximum value of the spectral entropy, M is the number of effective frequency points.

4. The target direction estimation method based on single vector hydrophone according to claim 3, characterized in that: The introduction of the adaptive coefficients to fuse the sound energy flow weights and the spectral entropy weights to obtain the joint weights comprises the following steps: Introducing an adaptive coefficient , where SNR denotes the signal-to-noise ratio. The coefficient According to the dynamic adjustment of the signal-to-noise ratio, when the signal-to-noise ratio is low, it tends to 1, and the sound energy flow weight is emphasized; when the signal-to-noise ratio is high, it tends to 0, and the spectrum entropy weight is emphasized, and the final joint weight is: ； In the formula, represents the acoustic energy flow weight.

5. The single vector hydrophone based target bearing estimation method of claim 4, wherein: The weighting of the covariance matrix by the joint weights to obtain the feature R, the decomposition of the feature R to reduce dimensions, and the obtaining of the reduced covariance matrix comprise the following steps: The covariance matrix represents the statistical characteristics of the multi-channel signals, the covariance matrix is weighted by the joint weights to obtain the feature R, and the calculation formula is: ； In the formula, is a frequency signal vector; is a regularization term; characteristic decomposition of R, and arranging the eigenvalues in descending order; Selecting the principal components K according to the cumulative energy criterion; ; wherein, denotes the eigenvalue; threshold value takes 0.85-0.95; Reserve energy proportion of top K principal eigenvectors And the corresponding eigenvalue , Reconstruct the dimensionality reduction covariance matrix to obtain the dimensionality reduction covariance matrix That is: ； In the formula, represents a feature vector.

6. The single vector hydrophone based target bearing estimation method of claim 5, wherein: The construction of the MVDR beamformer based on the reduced covariance matrix and the bearing spectrum scanning comprise the following steps: Computing MVDR weight vectors ; where, Plane waveguide steering vectors, is the target direction; The calculation formula of the bearing spectrum scanning is: 。 7. A single vector hydrophone based target bearing estimation system, characterized by, The method comprises the following steps: A signal acquisition module is configured to collect sound pressure signals in a sound pressure channel of a single vector hydrophone and particle velocity signals in a three-dimensional velocity channel, and construct a matrix based on the sound pressure signals and the particle velocity signals; A weight calculation module is configured to calculate sound energy flow weights and spectral entropy weights; A weight fusion module is configured to fuse the sound energy flow weights and the spectral entropy weights, and obtain joint weights; A feature dimension reduction module is configured to decompose a feature to reduce dimensions, and obtain a reduced covariance matrix; A target bearing estimation module is configured to construct an MVDR beamformer based on the reduced covariance matrix, and perform bearing spectrum scanning to estimate the target bearing.

8. An electronic device, comprising: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the

Images