Fusion method for multi-pitch-angle passive broadband warning

Abstract

The invention discloses a multi-pitch-angle passive broadband warning fusion method, which comprises the following steps of: 1, constructing a receiving beam space pointing definition, and setting beams of a sound array in a horizontal dimension and a pitch dimension; 2, processing a horizontal-dimension wave beam of a pitching wave beam; and step 3, carrying out fusion processing on a plurality of pitching dimension beams. The invention provides a fusion method for multi-pitch-angle passive broadband warning, and the method carries out the fusion of the azimuth-time history diagrams of all pitch angles through an information fusion technology, and enables an operator to search all underwater acoustic targets on one history diagram. Compared with an existing conventional method, an operator does not need to observe a plurality of process diagrams at the same time, the workload of the operator is greatly relieved, and the method has a good application prospect.

Description

Technical Field

[0001] The invention belongs to the field of underwater acoustic signal processing, and in particular relates to a fusion method for multi-pitch angle passive broadband warning. Background Art

[0002] When working in the deep sea, the underwater array will receive target signals propagated through three channels: direct waves, convergence zones, and seabed reflections. In winter, when there is a surface channel and the receiving array is located in the channel, it will also receive target signals propagated through this channel.

[0003] The grazing angle of sound waves that can propagate in the surface sound channel is extremely small (the maximum is generally not more than 2°, which is related to the thickness of the sound channel and the depth of the sound source, etc.). When this angle is exceeded, the sound line will penetrate downward into the deep ocean; the direct wave propagation distance is only a few thousand meters, and the incident angle of the received sound line is generally within ±10°; the convergence zone is located at a distance of 50km to 65km from the receiving sound array, with a width of about 3km to 5km. When the target is located in the convergence zone, the angle of the sound line entering the receiving sound array is generally within ±5°; seabed reflection propagation (acoustic shadow zone) means that the target's sound signal is reflected from the seabed and enters the receiving array. When the target is far away, the grazing angle of the sound line is small, and when the target is close, the grazing angle of the sound line is large. Considering the installation of the sound array and the sound lines mainly coming from the seabed, the grazing angle of the sound line that the sound array can receive is generally within 20° to -60° (positive upward).

[0004] From the above analysis, we can see that the surface sound channel has a long propagation distance but is unstable, the direct wave propagation has no blind spot, but the propagation distance is short, and the convergence zone has a long propagation distance, but can only detect targets within a 3km to 5km ring. The seabed reflection channel has no blind spot, so in the deep sea, it is of great significance to detect targets in the acoustic shadow zone through the seabed reflection channel.

[0005] The acoustic array usually has a certain height, and the receiving beam has a certain width in the vertical direction, which makes it difficult to cover the above range of 20° to -60°. Figure 1 As shown, the acoustic array sets a plurality of receiving beams with different pitch angles in the vertical direction to receive target signals propagating at different grazing angles.

[0006] Since the angle of the seabed reflected sound line of the target signal is unknown, the display and control console needs to open multiple "azimuth-time" history diagrams with different pitch angles at the same time. Different targets may be distributed in different history diagrams. The same target may appear in different pitch angles at different time periods due to movement. Therefore, the operator needs to observe multiple history diagrams at the same time, which brings a huge workload to the operator. Summary of the invention

[0007] In view of the shortcomings of the prior art, the present invention proposes a fusion method for multi-pitch angle passive broadband warning. The "azimuth-time" history diagrams of all pitch angles are fused through information fusion technology, so that the operator can search for all hydroacoustic targets on one history diagram, greatly reducing the operator's workload.

[0008] The technical solution of the present invention is:

[0009] A fusion method for multi-pitch angle passive broadband warning includes the following steps:

[0010] Step 1: Construct the spatial pointing definition of the receiving beam and set the beam of the acoustic array in the horizontal and elevation dimensions;

[0011] Step 2: Process the horizontal beam of a pitch beam;

[0012] Step 3: Fusion processing of multiple pitch-dimensional beams.

[0013] Preferably, the step 1 is to construct a receiving beam spatial pointing definition and set the beams of the acoustic array in the horizontal dimension and the pitch dimension, and the specific steps are:

[0014] Step 1.1: Construct the spatial pointing definition of the receiving beam, where O is the geometric center of the acoustic array, XOY is the horizontal plane of the acoustic array, X is the bow direction, Y points to the starboard side, and Z points to the top. is the beam azimuth (beam pointing to Projection on the horizontal plane The angle between the beam and the X axis is θ, and the beam pitch angle is Its projection on the horizontal plane The angle between ?

[0015] Step 1.2: Assume that the sound array forms N in the horizontal dimension h horizontal beams, each of which has a pitch dimension (vertical dimension) of N v beams, forming a total of N h N v beam b i,j (n)(i=0,1,...,N h -1; j = 0, 1, ..., N v -1).

[0016] Preferably, the step 2 is to process the horizontal beam of a pitch beam, and the specific steps are:

[0017] Step 2.1, solve the probability density distribution of the integrator background noise energy output:

[0018] If the background noise at the output of the hydrophone follows a normal distribution, then the beamformer output b i,j(n) has a mean of 0 and a variance of The normal distribution of x i,j (n) is subject to the mean The variance is The chi-squared value with 1 degree of freedom 2 distributed;

[0019] According to the central limit theorem, when N is large enough, the integrator output y i,j (n) approximates the mean The variance is The normal distribution of is:

[0020]

[0021] In formula (1), B is the noise bandwidth, T = Nf s is the integration time, N is the number of sampling times, f s is the sampling rate.

[0022] Step 2.2, mean and signal-to-noise ratio estimation:

[0023] At the same pitch angle, there are N h horizontal beam, for any elevation angle θ j , N h The energy integration output of the horizontal beam is y i , use OTA (Order Truncate Average) algorithm to find the average and define the space:

[0024] Ω i ={iM,...,i-1,i,i+1,...,i+M} (2),

[0025] In formula (2), i is the horizontal beam number, M is the warning beam interval, Ω i is the beam space involved in the mean calculation;

[0026] For i∈Ω i of y i , the output of median filtering is In y i Remove all the numbers greater than The value of , α is a constant (usually 0.9 to 1.1), and the remaining I values ​​are averaged, and the mean is estimated as:

[0027]

[0028] In formula (3), μ y for y i The mean value of y i Output data for the beam integrator;

[0029] The normalized output is:

[0030]

[0031] In formula (4), z i is the normalized output, y i is the beam integrator output data, μ y for y i The mean of .

[0032] z i The output signal-to-noise ratio is:

[0033]

[0034] In formula (5), SNR i is the output signal-to-noise ratio, B is the noise bandwidth, T = Nf s is the integration time, N is the number of sampling times, f s is the sampling rate, y i is the beam integrator output data, μ y for y i The mean of .

[0035] Preferably, the step 3 is a fusion process of the pitch-dimensional beam, and the specific steps are:

[0036] Step 3.1: Each horizontal beam number corresponds to N v The pitch beam and the horizontal beam The corresponding N v Elevation beam output SNR j (n), z j (n), the calculation formula is the same as step 2;

[0037] SNR j (j=0,1,...,N v -1) Sort by small to large, and take the median value Then the fusion processing algorithm of the pitch-dimensional beam is:

[0038]

[0039] In formula (6), P i is the i-th horizontal beam N v The fusion output of the pitch beam, z j is the normalized output of (4), SNR j is the output signal-to-noise ratio calculated by formula (5), and the value range of β is 1.2 to 1.3;

[0040] Step 3.2: Output P through equation (6) i The data is sent to the display to form an azimuth-time history diagram of the fused elevation beams.

[0041] Compared with the prior art, the beneficial effects of the present invention are: the "azimuth-time" history diagrams of all pitch angles are merged, so that the operator can search for all hydroacoustic targets on one history diagram, effectively reducing the operator's workload. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a schematic diagram of the pitch beam receiving the seabed reflected signal;

[0043] Figure 2 It is the beam coordinate system definition;

[0044] Figure 3 It is a schematic diagram of horizontal-vertical two-dimensional beam distribution;

[0045] Figure 4 is the horizontal dimension N h Schematic diagram of beam processing;

[0046] Figure 5 is the azimuth-time history diagram after different pitch angles and fusion, where Figure 5 (a) Figure 5 (b) Figure 5 (c) Figure 5 (d) Figure 5 (e) Figure 5 (f) Figure 5 (g) The corresponding pitch angle output results are 0°, -8°, -16°, -25°, -35°, -46° and -60° respectively. Figure 5 (h) is the output result after multi-pitch angle fusion using the method of the present invention. DETAILED DESCRIPTION

[0047] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0048] like Figures 1 to 4 As shown, the present invention provides a fusion method for multi-pitch angle passive broadband warning, and the specific steps are as follows:

[0049] Step 1: Construct the spatial pointing definition of the receiving beam and set the beam of the acoustic array in the horizontal and elevation dimensions, including:

[0050] Step 1.1: Construct the spatial pointing definition of the receiving beam, where O is the geometric center of the acoustic array, XOY is the horizontal plane of the acoustic array, X is the bow direction, Y points to the starboard side, and Z points to the top. is the beam azimuth (beam pointing to Projection on the horizontal plane The angle between the beam and the X axis is θ, and the beam pitch angle is Its projection on the horizontal plane The angle between ?

[0051] Step 1.2: Assume that the sound array forms N in the horizontal dimension h horizontal beams, each of which has a pitch dimension (vertical dimension) of N v beams, forming a total of N h N v beam b i,j (n)(i=0,1,...,N h -1; j = 0, 1, ..., N v -1).

[0052] Step 2: Process the horizontal beam of a pitch beam, including:

[0053] Step 2.1, solve the probability density distribution of the integrator background noise energy output:

[0054] If the background noise at the output of the hydrophone follows a normal distribution, then the beamformer output b i,j (n) has a mean of 0 and a variance of The normal distribution of x i,j (n) is subject to the mean The variance is χ with 1 degree of freedom 2 distributed;

[0055] According to the central limit theorem, when N is large enough, the integrator output y i,j (n) approximates the mean The variance is The normal distribution of is:

[0056]

[0057] In formula (1), B is the noise bandwidth, T = Nf s is the integration time, N is the number of sampling times, f s is the sampling rate. From formula (1), we can see that the variance of the integrated output data can be estimated by the mean.

[0058] Step 2.2, mean and signal-to-noise ratio estimation:

[0059] For any pitch angle θ j , N h The energy integration output of the horizontal beam is y i , use OTA algorithm to find the average and define the space:

[0060] Ω i ={iM,...,i-1,i,i+1,...,i+M} (2),

[0061] In formula (2), i is the horizontal beam number, M is the warning beam interval, Ω i is the beam space involved in the mean calculation;

[0062] For i∈Ω i of y i , the output of median filtering is In y i Remove all the numbers greater than The value of , α is a constant (usually 0.9 to 1.1), and the remaining I values ​​are averaged, and the mean is estimated as:

[0063]

[0064] In formula (3), μ y for y i The mean value of y i Output data for the beam integrator;

[0065] The normalized output is:

[0066]

[0067] In formula (4), z i is the normalized output, y i is the beam integrator output data, μ y for y i The mean of .

[0068] z i The output signal-to-noise ratio is:

[0069]

[0070] In formula (5), SNR i is the output signal-to-noise ratio, B is the noise bandwidth, T = Nf s is the integration time, N is the number of sampling times, f s is the sampling rate, y i is the beam integrator output data, μ y for y i The mean of .

[0071] Step 3: Fusion processing of multiple pitch-dimensional beams, including:

[0072] Step 3.1: Each horizontal beam number corresponds to N v The elevation beam and the horizontal beam The corresponding N v Elevation beam output SNR j (n), z j (n), the calculation formula is the same as step 2;

[0073] SNR j (j=0,1,...,N v -1) Sort by small to large, and take the median value Then the fusion processing algorithm of the pitch-dimensional beam is:

[0074]

[0075] In formula (6), P i is the i-th horizontal beam N v The fusion output of the pitch beam, z j is the normalized output of (4), SNR j is the output signal-to-noise ratio calculated by formula (5), and the value range of β is 1.2 to 1.3;

[0076] Step 3.2: Output P through equation (6) i The display forms an azimuth-time history diagram after the fusion processing of the pitch-dimensional beam.

[0077] In an application scenario of the present invention, the acoustic array is set to a cylindrical array with a diameter of 6m and a height of 1.5m. 128 hydrophones are evenly arranged in the horizontal dimension, 10 hydrophones are evenly arranged in the vertical dimension, a total of 1280 hydrophones, and the hydrophone spacing is 18.75cm. A total of 512 evenly distributed beams are formed in the horizontal dimension, and 7 beams are formed in the elevation dimension, with beam elevation angles of 0°, -8°, -16°, -25°, -35°, -46° and -60° respectively.

[0078] At the same time, two moving targets were set in the deep sea environment. The initial azimuth of target 1 was 90°, and the azimuth gradually increased as the distance increased from far to near. The initial azimuth of target 2 was 120°, and the azimuth gradually decreased as the distance increased from near to far. The azimuths of the two targets intersected at about 108°.

[0079] from Figure 5 The output result shows that, Figure 5 In (a), the beam with an elevation angle of 0° cannot detect the two target signals reflected from the seabed. Figure 5 (b) to Figure 5 In (e), the receiving beams with elevation angles of -8°, -16°, -25°, and -35° can only detect target #1 in part of the time period (target #1 is to the left). Figure 5 (c) to Figure 5 In (g), the receiving beams with elevation angles of -16°, -25°, -35°, -46°, and -60° can only detect target #2 (target #2 is to the right) in part of the time period. Figure 5(h) is the result of multi-pitch angle fusion of the present invention. It can be seen that after fusion, the target can be found in the entire time period. The operator only needs to observe the fused azimuth-time history diagram to complete the all-round target search.

[0080] It should be noted that the above embodiments are merely preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. All equivalent modifications made on the basis of the above embodiments belong to the protection scope of the present invention.

Claims

1. A fusion method for multi-pitch angle passive broadband warning, characterized in that: include: Step 1: Construct the spatial pointing definition of the receiving beam and set the beam of the acoustic array in the horizontal and elevation dimensions; Step 2: Process the horizontal beam of a pitch beam; Step 3: Fusion processing of multiple pitch-dimensional beams.

2. The fusion method of multi-pitch angle passive broadband warning according to claim 1 is characterized by: The step 1 is to construct the spatial pointing definition of the receiving beam and set the beams of the acoustic array in the horizontal dimension and the elevation dimension. The specific steps are: Step 1.1: Construct the spatial pointing definition of the receiving beam, where O is the geometric center of the acoustic array, XOY is the horizontal plane of the acoustic array, X is the bow direction, Y points to the starboard side, and Z points to the top. is the beam azimuth angle, θ is the beam elevation angle; Step 1.2: Assume that the sound array forms N in the horizontal dimension h horizontal beams, each of which has a pitch dimension of N v beams, forming a total of N h N v beam b i,j (n)(i=0,1,...,N h -1; j = 0, 1, ..., N v -1).

3. The fusion method of multi-pitch angle passive broadband warning according to claim 2 is characterized in that: The step 2 is to process the horizontal beam of a pitch beam, and the specific steps are as follows: Step 2.1, solve the probability density distribution of the integrator background noise energy output: If the background noise at the output of the hydrophone follows a normal distribution, then the beamformer output b i,j (n) has a mean of 0 and a variance of The normal distribution of x i,j (n) is subject to the mean The variance is χ with 1 degree of freedom 2 distributed; According to the central limit theorem, when N is large enough, the integrator output y i,j (n) approximates the mean The variance is The normal distribution of is: In formula (1), B is the noise bandwidth, T = Nf s is the integration time, N is the number of sampling times, f s is the sampling rate; Step 2.2, mean and signal-to-noise ratio estimation: For any pitch angle θ j , N h The energy integration output of the horizontal beam is y i , use OTA algorithm to find the average and define the space: Ω i ={i-M,...,i-1,i,i+1,...,i+M} (2), In formula (2), i is the horizontal beam number, M is the warning beam interval, Ω i is the beam space involved in the mean calculation; For i∈Ω i of y i , the output of median filtering is In y i Remove all the numbers greater than The value of , α is a constant, and the remaining I values ​​are averaged, then the mean estimate is: In formula (3), μ y for y i The mean value of y i Output data for the beam integrator; The normalized output is: In formula (4), z i is the normalized output, y i is the beam integrator output data, μ y for y i The mean of . z i The output signal-to-noise ratio is: In formula (5), SNR i is the output signal-to-noise ratio, B is the noise bandwidth, T = Nf s is the integration time, N is the number of sampling times, f s is the sampling rate, y i is the beam integrator output data, μ y for y i The mean of .

4. The fusion method of multi-pitch angle passive broadband warning according to claim 3 is characterized by: The step 3 is the fusion processing of the pitch-dimensional beam, and the specific steps are as follows: Step 3.1: Each horizontal beam number corresponds to N v The elevation beam and the horizontal beam The corresponding N v Elevation beam output SNR j (n), z j (n), for SNR j (j=0,1,...,N v -1) Sort by small to large, and take the median value Then the fusion processing algorithm of the pitch-dimensional beam is: In formula (6), P i is the i-th horizontal beam N v The fusion output of the pitch beam, z j is the normalized output of (4), SNR j is the output signal-to-noise ratio calculated by formula (5), and the value range of β is 1.2 to 1.3; Step 3.2: Output P through equation (6) i The display forms an azimuth-time history diagram after the fusion processing of the pitch-dimensional beam.

Images