A combined acoustic energy flow single vector hydrophone positioning method

Abstract

The present invention belongs to the field of measurement technology, and specifically relates to a combined acoustic energy flow single vector hydrophone positioning method. The method comprises the following steps: using the sound pressure and particle velocity signals received by the single vector hydrophone to calculate the multi-channel combined acoustic energy flow signal; using the obtained combined acoustic energy flow signal to construct a signal covariance matrix, calculate the weighted vector corresponding to the combined acoustic energy flow signal, and perform conventional beamforming calculations on the covariance matrix; construct a combined system point spread function with linear shift invariant characteristics; perform deconvolution operations on the conventional beamforming output and the point spread function, iteratively search for the target's position, and obtain underwater target positioning results. The method of the present invention can effectively reduce the beam width, improve the output beam main-sidelobe ratio, suppress background noise interference, and further improve the target positioning accuracy while retaining the robustness of conventional beamforming positioning performance, so that the single vector hydrophone is more suitable for underwater target positioning work scenarios with complex working conditions.

Description

Technical Field

[0001] The invention belongs to the field of measurement technology, and in particular relates to a combined acoustic energy flow single-vector hydrophone positioning method. Background Art

[0002] Vector hydrophones can pick up the sound pressure and three-dimensional particle velocity of the radiating sound source at the same point in three-dimensional space. Underwater target positioning can be completed using a single vector hydrophone. Compared with the traditional scalar hydrophone array, vector hydrophones have technical advantages such as small scale, frequency independence, and strong interference suppression ability. It is an emerging underwater target positioning technology. However, limited by the number of channels of a single vector hydrophone, when array signal processing theory is used for target positioning, the number of equivalent array elements is small, and the beam width obtained by conventional beamforming methods is usually wide, the main sidelobe ratio is low, and the adjacent target resolution ability is weak, which limits the multi-target resolution ability and target positioning accuracy of the single vector hydrophone. When high-resolution positioning methods such as multiple signal subspaces and minimum variance distortion-free are used, although the beam width can be effectively reduced, the algorithm has poor robustness and needs to rely on prior knowledge such as the number of known signal sources, high signal-to-noise ratio, or good working conditions, which cannot meet the target positioning requirements of single vector hydrophones in complex underwater environments. Summary of the invention

[0003] The present invention focuses on overcoming the shortcoming that the beam width of the conventional beamforming output of the single vector hydrophone is limited by the number of equivalent array elements. The sound pressure and particle velocity signals received by the single vector hydrophone are combined for processing, and a combined system point spread function with linear shift invariant characteristics is designed and constructed. The deconvolution theory is used to iteratively search for the target position, thereby obtaining a narrower main lobe width than conventional beamforming, a more robust positioning performance than a high-resolution beamforming algorithm, and a better background noise suppression capability, thereby effectively improving the underwater target positioning accuracy of the single vector hydrophone.

[0004] The present invention provides a combined acoustic energy flow single vector hydrophone positioning method, comprising the following steps:

[0005] Step (1): Calculate the multi-channel combined acoustic energy flow signal using the sound pressure and particle velocity signals received by the single vector hydrophone;

[0006] Step (2): construct a signal covariance matrix using the obtained combined acoustic energy flow signal, calculate the weighted vector corresponding to the combined acoustic energy flow signal, and perform conventional beamforming calculation on the covariance matrix;

[0007] Step (3): construct a point spread function of the combined system with linear shift invariant characteristics;

[0008] Step (4): Perform deconvolution operation on the conventional beamforming output and the point spread function, iteratively search for the target location, and obtain the underwater target positioning result.

[0009] Preferably, in step (1), the multi-channel combined acoustic energy flow signal is calculated as follows:

[0010] ;

[0011] in, Represents the time series signal of the sound pressure channel, express The time series signal of the particle velocity of the shaft, express The time series signal of the particle velocity of the shaft; express Axial acoustic energy flow, express Axial acoustic energy flow, express The second-order statistics of the axis-mass velocity, express The second-order statistics of the axis-mass velocity, express Axis and The second-order statistics of the cross-correlation of the axis-mass velocity;

[0012] Construct the combined acoustic energy flow signal matrix: .

[0013] Preferably, in step (2), the signal covariance matrix constructed is for:

[0014] ;

[0015] in, is the acoustic energy flow signal matrix The conjugate transpose of is the matrix conjugate transpose operation;

[0016] Calculate the weighted vector corresponding to the combined acoustic energy flow signal :

[0017] ;

[0018] in, is the preset scanning beam angle, is the matrix transpose operation;

[0019] Using the acoustic energy flow signal matrix and weight vector Get the output of the general beamforming calculation:

[0020] ;

[0021] in, is the number of combined array elements.

[0022] Preferably, in step (3), the point spread function of the constructed combined system with linear shift-invariant characteristics is:

[0023] ;

[0024] in, is the target's azimuth angle.

[0025] Preferably, in step (4), the deconvolution operation formula is:

[0026] ;

[0027] in, For the The estimated value of the target position obtained by the deconvolution operation, is scalar multiplication, is the cross-correlation operation, is the convolution operation;

[0028] For each iteration, The sequence is normalized:

[0029] ;

[0030] in, for The maximum value of the sequence;

[0031] Using the average acoustic energy flow right To perform weighting:

[0032] ;

[0033] ;

[0034] in, is the current signal snapshot period;

[0035] Repeat the above process. After deconvolution iterations, the estimated value of the target's position can be obtained.

[0036] The present invention proposes a combined acoustic energy flow single vector hydrophone positioning method, which combines and processes the sound pressure and particle velocity signals received by the single vector hydrophone, designs a combined system point spread function with linear shift invariance characteristics, and uses deconvolution theory to iteratively search for the target's position. The method of the present invention can overcome the shortcomings of the existing positioning method that the beam width is limited by the number of array elements to a certain extent. While retaining the robustness of conventional beamforming positioning performance, it effectively reduces the beam width, improves the output beam main-sidelobe ratio, suppresses background noise interference, and further improves the target positioning accuracy, making the single vector hydrophone more suitable for underwater target positioning work scenarios with complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A schematic flow chart of a combined acoustic energy flow single vector hydrophone positioning method provided in an embodiment of the present invention;

[0038] Figure 2 This is a schematic diagram of a single vector hydrophone receiving signals in three-dimensional space;

[0039] Figure 3 For a single vector hydrophone and a single target ( )Detection effect simulation comparison;

[0040] Figure 4 For a single vector hydrophone with two targets ( , )Detection effect simulation comparison;

[0041] Figure 5 This is a comparison of directivity diagrams of single-vector hydrophone arrays;

[0042] Figure 6 A point spread function array constructed in the implementation of the present invention;

[0043] Figure 7 It is the time-azimuth diagram of the stationary target for the CBF beamforming method;

[0044] Figure 8 It is the time-azimuth diagram of the stationary target for the MVDR beamforming method;

[0045] Fig. 9 It is the time-azimuth diagram of the stationary target of the MUSIC beamforming method;

[0046] Fig.10 It is the time-azimuth diagram of the stationary target;

[0047] Fig.11 Comparison diagram of different methods for moving target positioning effects; among them, (a) CBF; (b) MVDR; (c) MUSIC; (d) the method of the present invention;

[0048] Fig.12 Comparison diagram of beam output time slices. DETAILED DESCRIPTION

[0049] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0050] Embodiment 1 The combined acoustic energy flow single vector hydrophone positioning method proposed in the present invention has the following process: Figure 1 As shown, first, the received sound pressure and particle velocity channel signals are combined to calculate the first-order acoustic energy flow and the second-order combined acoustic energy flow, and then the signal covariance matrix of the combined acoustic energy flow signal is constructed; secondly, the weighted vector of the combined acoustic energy flow signal is calculated, and the beam output is obtained by processing it with the conventional beamforming algorithm; thirdly, a point spread function with linear shift invariant characteristics is constructed, and it is deconvolved iteratively with the conventional beamforming output result, and finally the target direction estimation result is obtained. The specific steps are as follows:

[0051] Step (1): Calculate the multi-channel combined acoustic energy flow signal using the sound pressure and particle velocity signals received by the single vector hydrophone.

[0052] The schematic diagram of a vector hydrophone receiving a sound source radiation signal in three-dimensional space is shown in Figure 2 As shown in the figure, a single vector hydrophone can pick up the sound pressure and three-dimensional particle velocity of the sound source target at the same point. For long-distance targets, the acoustic signal reception in the two-dimensional plane is generally considered, and the time series signal of the sound pressure channel is And particle velocity channel signal , It can be expressed as:

[0053] ;

[0054] in, Indicates There are incident signal sources, totaling An independent sound source signal is received by a single vector hydrophone; is the horizontal azimuth of the target, is the pitch angle of the target, , , are the noise of each channel respectively.

[0055] The multi-channel combined acoustic energy flow signal is calculated as follows: ;

[0056] in, Represents the time series signal of the sound pressure channel, express The time series signal of the particle velocity of the shaft, express The time series signal of the particle velocity of the shaft; express Axial acoustic energy flow, express Axial acoustic energy flow, express The second-order statistics of the axis-mass velocity, express The second-order statistics of the axis-mass velocity, express Axis and The second-order statistics of the cross-correlation of the axis-mass velocity;

[0057] Construct the combined acoustic energy flow signal matrix: .

[0058] Step (2): construct a signal covariance matrix using the obtained combined acoustic energy flow signal, calculate the weighted vector corresponding to the combined acoustic energy flow signal, and perform conventional beamforming calculation on the covariance matrix.

[0059] Constructed signal covariance matrix for:

[0060] ;

[0061] in, is the acoustic energy flow signal matrix The conjugate transpose of is the matrix conjugate transpose operation;

[0062] Calculate the weighted vector corresponding to the combined acoustic energy flow signal :

[0063] ;

[0064] in, is the preset scanning beam angle, is the matrix transpose operation;

[0065] Using the acoustic energy flow signal matrix and weight vector Get the output of the general beamforming calculation:

[0066] ;

[0067] in, is the number of combined array elements.

[0068] Step (3): Construct a point spread function of the combined system with linear shift invariant properties.

[0069] set up is the point spread function to be constructed, Indicates when the target is at the beam angle The target orientation distribution function at position, is the conventional beamforming output sequence. The conventional beamforming calculation can be equivalent to the convolution between the point spread function and the position of the target to be estimated, that is, when and When known, it can be obtained by the following positive convolution calculation :

[0070] .

[0071] Therefore, when and When it is known, is the target position distribution function to be solved.

[0072] First, construct the point spread function :

[0073] ;

[0074] Secondly, using the trigonometric formula To merge and sort:

[0075] ;

[0076] Finally, it can be found that , this point spread function has the linear shift invariant property.

[0077] Step (4): Perform deconvolution operation on the conventional beamforming output and the point spread function, iteratively search for the target location, and obtain the underwater target positioning estimation result.

[0078] Known conventional beamforming sequences and the point spread function with shift-invariant properties , use the following formula to search for the estimated value of the target's position through multiple iterations:

[0079] ;

[0080] in, For the The deconvolution estimation result is The estimated value of the target position obtained by the iteration is is scalar multiplication, is the cross-correlation operation, is the convolution operation;

[0081] For each iteration, The sequence is normalized:

[0082] ;

[0083] in, for The maximum value of the sequence;

[0084] Calculate the current signal snapshot period The average acoustic energy flow in:

[0085] ;

[0086] Using the average acoustic energy flow right Weighting is performed to achieve uniformity of background noise:

[0087] .

[0088] Repeat the above process. ( After (total number of iterations) deconvolution iterations, the estimated value of the target position can be obtained. , ;Will The time-azimuth diagram of the single-vector hydrophone target azimuth estimation can be obtained by traversing the time.

[0089] Example 2 The method proposed in the present invention is used to conduct a single target simulation experiment. The echo of the radiation noise of a single target ship is simulated, and the simulation parameters are set as follows: the harmonic order of the radiation noise of the receiving ship , number of propeller blades , propeller speed rpm. Radiated noise signal-to-noise ratio , sampling frequency , sampling time 1s, signal bandpass filter passband range 10Hz to 5kHz, target location ,exist The target scanning is performed within the range. The positioning capabilities of the conventional beamforming method (CBF), the minimum variance distortion-free method (MVDR) and the combined acoustic energy flow method proposed in this application for single target are compared. Figure 3As shown, the CBF method can accurately locate the angle of the target, but is limited by the array signal processing theory, the number of effective array elements is only 3, and the main lobe width of the beam is relatively large. Compared with the CBF method, the MVDR method can effectively reduce the beam width and accurately locate the position of a single target, but the sidelobe performance of its beam output is poor, and a certain beam broadening phenomenon occurs. The main lobe width of the combined acoustic energy flow method of the present application at -3dB is equivalent to that of the MVDR method, and its ability to maintain the main lobe width is better, and there is no obvious sidelobe and beam broadening phenomenon, which can effectively achieve high-resolution positioning of the target.

[0090] Example 3 The method of the present invention is used to conduct a dual-target simulation experiment, setting the echo of the dual-target ship radiation noise to be simulated, and the parameters of the simulation target 1 are set as follows: the harmonic order of the receiving ship radiation noise , number of propeller blades , propeller speed rpm, target location The parameters of simulation target 2 are set as: the harmonic frequency of the receiving ship radiation noise , number of propeller blades , propeller speed rpm, target location Radiated noise signal-to-noise ratio , sampling frequency , sampling time 1s, signal bandpass filter passband range 10Hz to 5kHz, The target scanning is performed within the range. The positioning capabilities of the CBF method, the MVDR method and the combined acoustic energy flow method proposed in this application are compared respectively. Figure 4 The basic theory of single vector hydrophone positioning limits the single vector hydrophone to only be able to distinguish two targets at most, and the target interval that can be distinguished is large. Therefore, the CBF method is not suitable for and The dual-target resolution is poor and the positioning result is only the composite azimuth of the dual targets, and the number and position of the targets cannot be resolved. The MVDR method can accurately distinguish the locations of the dual targets and maintain a good beam width and main-sidelobe ratio. The combined acoustic energy flow method proposed in this application can obtain a mainlobe width equivalent to that of the MVDR method at -3dB, with good discrimination between the dual targets. Compared with the MVDR method, its beam width widens more slowly, and the sidelobe level is lower, thus achieving better dual-target resolution.

[0091] The directivity diagram of the combined acoustic energy flow array with shift-invariant properties designed by the present invention is as follows: Figure 5 As shown, compared with the directivity diagram of the conventional method, the method of the present invention has a narrower main lobe width while maintaining the shift-invariant property, which is more conducive to high-resolution positioning of the target. Figure 6 It can be observed that the point spread function array designed by the combined acoustic energy flow method has good linear shift invariance characteristics within the principal value interval and can meet the prerequisites for deconvolution calculations.

[0092] Example 4 The application effect of the present invention is further demonstrated by combining measured data: the field experimental data of ship target positioning on a certain voyage is selected for algorithm processing, and the positioning effects of different positioning methods on the sound source target are compared. First, the data of a stationary target is selected for processing. The ship is in a stationary state and uses the transmitting transducer to continuously radiate into the water. The single-frequency signal of the underwater acoustic signal is received by a single vector hydrophone. The receiving time is 45 minutes and the signal snapshot time is 1 minute. The conventional beamforming method (CBF), the minimum variance distortion-free method (MVDR), the multiple signal subspace method (MUSIC) and the combined acoustic energy flow method proposed in this application are used to process the target with the beamforming algorithm, and the output time-azimuth diagram is observed. Figures 7 to 10 It can be observed that the CBF method can accurately locate the position of the target, but its main lobe width is large and the side lobe level is high, resulting in severe beam expansion on the angular axis and large background noise. Compared with the CBF method, the MVDR method can effectively reduce the beam width, improve the resolution of target positioning, and has a better suppression effect on background noise. The MUSIC method can further reduce the beam width compared to the MVDR method, but because the MUSIC method is more dependent on the signal-to-noise ratio of the signal, its background noise suppression effect is not good in underwater target positioning applications with complex working conditions. The combined acoustic energy flow method proposed in this application can obtain a narrower beam width, the energy output is more concentrated, and the background noise can be suppressed to a certain extent, thereby improving the positioning accuracy of the target.

[0093] Secondly, the echo data of the moving target is selected for processing. The ship target slowly passes in the transverse direction in front of the single vector hydrophone observation point, and the observation time is 120 seconds. The signals received by the single vector hydrophone are processed by positioning algorithms using CBF, MVDR, MUSIC and the combined acoustic energy flow method proposed in this application, and the movement trajectory of the ship target is continuously observed.

[0094] Depend on Fig.11It can be observed that the CBF method can effectively locate the target, but the target trajectory is coarse and the energy diffusion is more serious, the ability to suppress background noise is weak, and the main lobe width of the beam is large, which is not conducive to the precise positioning of the target. The MVDR method can effectively reduce the beam width, has better continuous tracking capability for the target position and can effectively suppress background noise. Although the MUSIC method can obtain a narrower beam width than the MVDR method, its ability to suppress background noise is limited. The method proposed in the present application can obtain a narrower beam width than the MUSIC method, better background noise suppression capability than the MVDR method, and less sidelobe interference, more concentrated energy output, and can effectively observe changes in the target's navigation trajectory.

[0095] In order to further observe the contrast effect of beam output, Fig.12 The time slice of the beam output is given, and it can be observed that the CBF method can accurately locate the target position, but the obtained beam is wide and the sidelobe level is high. The MVDR method can effectively reduce the beam width and lower the beam sidelobe level. Although the MUSIC method can further reduce the beam width, it is seriously interfered by noise, and the sidelobe level is comparable to the CBF method, and the ability to suppress interference from background noise needs to be improved. The combined acoustic energy flow method proposed in this application can obtain a better main-sidelobe ratio on the basis of further lowering the beam width, and effectively suppress the interference of background noise.

Claims

1. A combined acoustic energy flow single vector hydrophone positioning method, characterized in that: The following steps are involved: Step (1): Calculate the multi-channel combined acoustic energy flow signal using the sound pressure and particle velocity signals received by the single vector hydrophone; the multi-channel combined acoustic energy flow signal is calculated as follows: ； in, Represents the time series signal of the sound pressure channel, express The time series signal of the particle velocity of the shaft, express The time series signal of the particle velocity of the shaft; express Axial acoustic energy flow, express Axial acoustic energy flow, express The second-order statistics of the axis-mass velocity, express The second-order statistics of the axis-mass velocity, express Axis and The second-order statistics of the cross-correlation of the axis-mass velocity; Construct the combined acoustic energy flow signal matrix: ; Step (2): construct a signal covariance matrix using the obtained combined acoustic energy flow signal, calculate the weighted vector corresponding to the combined acoustic energy flow signal, and perform conventional beamforming calculation on the covariance matrix; the constructed signal covariance matrix for: ； in, is the acoustic energy flow signal matrix The conjugate transpose of is the matrix conjugate transpose operation; Calculate the weighted vector corresponding to the combined acoustic energy flow signal : ； in, is the preset scanning beam angle, is the matrix transpose operation; Using the acoustic energy flow signal matrix and weight vector Get the output of the general beamforming calculation: ； in, is the number of combined array elements; Step (3): Construct a point spread function of the combined system with linear shift invariance: ； in, is the target’s azimuth angle; Step (4): Perform deconvolution operation on the conventional beamforming output and the point spread function, iteratively search for the target location, and obtain the underwater target positioning result.

2. The combined acoustic energy flow single vector hydrophone positioning method according to claim 1, characterized in that: In step (4), the deconvolution operation formula is: ； in, For the The estimated value of the target position obtained by the deconvolution operation, is scalar multiplication, is the cross-correlation operation, is the convolution operation; For each iteration, The sequence is normalized: ； in, for The maximum value of the sequence; Using the average acoustic energy flow right To perform weighting: ； ； in, is the current signal snapshot period; Repeat the above process. After deconvolution iterations, the estimated value of the target's position can be obtained.

Images