Deep sea seabed planar array target positioning method based on multi-path coherent focusing
Technical Field
The invention belongs to the fields of underwater acoustic array signal processing, underwater acoustic detection, underwater acoustic positioning and the like, and relates to a deep sea seabed planar array target positioning method based on multi-path coherent focusing.
Background
The deep coast matrix is arranged on the sea floor, has high self safety and good concealment, can be formed into an array on a large scale, has excellent detection performance, and is one of the important means for passive detection of the underwater target at present. The target depth information is one of the currently known important criteria for classifying the underwater targets on the water surface, but the current deep coastal matrix has a plurality of difficulties in positioning the targets in the direct sound area in practical application. The matching field is one of classical means for solving the problem, but marine environment parameters are required to be known sufficiently, and further the acoustic field amplitude and phase of the receiving matrix are calculated by utilizing the underwater acoustic model to form a copying field vector, and the copying field vector is matched with acoustic data received by the matrix, so that the underwater target is positioned. The method needs to calculate the full field of the sound pressure field, and the calculated amount is quite large under the condition of three-dimensional space grid of the multi-frequency broadband signal. In practical application, even if the sound pressure copy field is calculated in advance and stored, the method is very sensitive to the mismatch of environmental parameters, so that the positioning error is large and even the method is invalid.
Deep sea sound fields typically have distinct, distinguishable, multi-path arrival structures, such as in the deep sea direct sound zone, where the signal received by the hydrophone is from the direct path, the sea surface primary reflection path, the sea bottom primary-sea surface primary reflection path, and so on. The existing method fully utilizes multi-path information to develop positioning research on a target, for example, china patent (patent No. ZL 202411119049.9) discloses an underwater sound source detection positioning method and device based on deep sea multi-path focusing, the method utilizes a space power spectrum estimation method to calculate wave beam output sound fields of sound pressure signals radiated by a target sound source in different frequency points and different azimuth angles within a set frequency range, performs phase compensation addition on the wave beam output sound fields of the frequency point in two different azimuth angles according to different assumed arrival time delays aiming at any frequency point to obtain a focusing result, adds the multi-path sound field focusing results of a plurality of frequency points to obtain a broadband multi-path three-dimensional sound field focusing result, compares two angles corresponding to the maximum value of the broadband focusing result and the arrival time delay with template values under different assumed target distances and depths, and determines estimated values of the target distances and the depth. The method realizes the coherent superposition detection of the deep sea multi-path signals, improves the detection positioning capability of the underwater weak sound source in the multi-path environment, and requires that the direct sound line reaching the array and the sea surface primary reflection sound line reach pitch angle to be distinguishable, but the equivalent vertical aperture of the submarine planar array is limited, and the pitch angles of the two paths reaching the array are usually very close and difficult to distinguish.
The paper (Mao Junjie, zhang Bo, etc., acoustic theory report, 2025, 1 st stage) discloses near field localization research of an air sound source by using sound rays with different paths, such as direct sound, submarine primary-sea primary reflection sound rays, submarine secondary-sea secondary reflection sound rays, and the like, which are incident into the deep sea from the air, to perform near field focusing beam formation, and to solve the problem of pseudo peaks of spatial spectrum, the weighting summation is performed on the spatial spectrums corresponding to multiple paths, so that the positioning of the air sound source in a near field Fresnel zone by the submarine horizontal array is realized. However, spatial spectrum weighted summation is incoherent summation of multi-path signal energy, so that the output signal-to-noise ratio cannot be improved, and weak target detection capability cannot be improved.
Disclosure of Invention
In order to solve the problems of large calculation amount, large storage amount of a copying field and the like in the prior art for passively positioning a target of a deep sea submarine planar array, the invention provides a target positioning method of the deep sea submarine planar array based on multi-path coherent focusing, which utilizes the physical characteristics that a sound source in a direct sound area reaches the array mainly through a direct path and a sea surface primary reflection path, by establishing three-dimensional grid points of possible sound source positions with different angles, distances and depths, further calculating propagation time and amplitude of a direct path D of each grid point reaching each array element under the condition of a given sound velocity profile and a sea surface primary reflection path S, constructing a new steering vector, carrying out beam forming to realize coherent accumulation of multipath sound energy, and finally obtaining the position of the grid point corresponding to the peak value of the ambiguity function as the position of the expected sound source. The method only needs to measure the sound velocity profile, is suitable for estimating the azimuth angle, the distance and the depth of the broadband and narrowband targets in the direct sound area, and has low calculation complexity. Meanwhile, the invention only needs to store the multi-path propagation time corresponding to the central frequency of the sound source in the search spaceSum amplitudeAs a parameter template required by the guide vector, the storage space is greatly reduced, and the hardware cost is greatly saved.
The method comprises the following specific steps:
Step 1, establishing a three-dimensional rectangular coordinate system and obtaining sound pressure signals of a submarine planar array;
Step 2, constructing three-dimensional search space grid points of all possible positions of a potential target, and calculating propagation time and amplitude corresponding to a direct path D and a sea surface primary reflection path S of each grid point reaching each array element by utilizing a ray model BELLHOP;
step 3, constructing a corresponding multi-path coherent focusing guide vector under each frequency point according to the searched sound source depth, horizontal distance and horizontal azimuth angle, and carrying out normalization processing;
And 4, carrying out beam forming on the given frequency point to obtain a ambiguity function so as to realize coherent accumulation of multipath acoustic energy, wherein the value of each point of the ambiguity function represents the multipath focusing energy of the acoustic source at the corresponding grid point, which propagates to the array along two paths, and the maximum value of the ambiguity function corresponds to the depth, horizontal distance and horizontal azimuth angle of the expected acoustic source.
The form of beamforming is not limited to conventional beamforming, but may also include MVDR (i.e., MINIMAL VARIANCE Distortion Response) methods, and the like. If the broadband signal is the broadband signal, further broadband beam energy synthesis can be performed to obtain a broadband ambiguity function.
The step 1 specifically comprises the following steps:
establishing a rectangular coordinate system by taking any position of the sea surface as an origin of coordinates O and taking east and north directions as an x axis and a y axis respectively, and setting a plane array to have M array elements in total, wherein the coordinates of the M array elements are as follows Wherein x m is the x-axis coordinate of the m-th array element, y m is the y-axis coordinate of the m-th array element, and H is the known sea depth. The sound source depth is z s, the horizontal distance to the origin of coordinates and the horizontal azimuth angle are r s and θ s, respectively, wherein the horizontal azimuth angle θ s is defined as the angle between the projection of the sound source onto the xoy plane and the x-axis, and takes on a value of 0 to 360 degrees.
Setting the frequency spectrum of the target sound source radiation signal received by the mth array element at the frequency point fWherein,、The lowest frequency and the highest frequency of the sound pressure signal, respectively. The sound pressure signals received by the M array elements are expressed as M multiplied by 1 dimension column vectors in a vector form:
(1)
the superscript T denotes the transpose of the vector.
The step 2 specifically comprises the following steps:
If the depth, horizontal distance and horizontal azimuth angle of the sound source in the search space are z, r and theta respectively, the coordinates of the sound source in the rectangular coordinate system can be written . When the sound source is located in the direct sound zone of the array, only the direct path D and the sea surface primary reflection path S are considered. Under the condition of known sound velocity profile and sea depth (the sound velocity profile and the sea depth can be measured on a sea test site through devices such as a sound velocity meter, a depth finder and the like), substituting the sound source position and the position of a receiver (here, the m-th array element of the array) into a ray model BELLHOP, and calculating to obtain the propagation time of the sound source reaching the m-th array element of the array through a direct path D and a sea surface primary reflection path SAmplitude and magnitude ofWhereinRespectively corresponding to the direct path D and the sea surface primary reflection path S.
When the bending of the sound ray is not considered, that is, straight line propagation is assumed, the propagation distance from the sound pressure signal sent by the actual sound source to the m-th array element through the first path is as follows:
(2)
Wherein the method comprises the steps of 。
At this time, the propagation time and amplitude of the sound pressure signal sent by the actual sound source reaching the mth array element through the first path can be approximated by the spherical wave propagation rule, namely:
Where c is the speed of sound in seawater. For the sake of brevity of the formula, the following will be made Abbreviated as,Abbreviated as;
Because sound velocity in the real marine environment presents certain distribution along the depth, as the distance between the sound source and the receiver increases, formulas (2) - (4) are no longer established, and at the moment, a ray model BELLHOP is adopted to calculate the propagation time of the sound source in the search space reaching the mth array element of the array through the direct path D and the sea surface primary reflection path SSum amplitudeMore accurate, wherein the calculation uses the signal center frequency. Once the computation is complete, the propagation time can be determinedSum amplitudeStored as templates for subsequent direct invocation. The matching field technology needs to calculate sound pressure copy fields aiming at all frequency points, and under the deep sea condition, the calculation amount is very large and is difficult to use for real-time calculation. The invention utilizes the advantage that the multi-path propagation time and the multi-path amplitude obtained by the ray model calculation are insensitive to the signal frequency, and the calculation amount is greatly reduced when the broadband signal is directly called. Meanwhile, the matching field needs to store broadband complex sound pressure fields received by a plurality of primitive arrays as a matching template, and the invention only stores the multi-path propagation time corresponding to the central frequency of the sound source in the search spaceSum amplitudeThe storage space is greatly reduced, and the hardware cost is greatly saved.
The step 3 specifically comprises the following steps:
given the frequency f of the acoustic signal, constructing a multi-path coherent focusing guide vector according to the signal amplitude and propagation time of the direct acoustic path D and the sea surface primary reflection acoustic path S under the actual sound source depth z, the horizontal distance r and the horizontal azimuth angle theta :
(5)
For guiding vectorNormalization is carried out:
(6)
Representing the normalized steering vector;
the step 4 comprises the following steps:
for sound pressure signals received by planar arrays Performing conventional beam forming processing to obtain beam domain signal pointing to target:
(7)
When the sound pressure signal is a narrowband signal, the energy of the output narrowband signal isAs a function of narrowband ambiguity. The processing has the greatest advantage that the coherent synthesis of the multi-path signals is realized similar to the matching field processing, and the signal to noise ratio is improved.
The form of beamforming in equation (7) is not limited to conventional beamforming, but may also include MVDR (i.e., MINIMAL VARIANCE Distortion Response) method, i.e.:
(8)
Wherein the method comprises the steps of In order to observe the data cross-spectral density matrix,Representing inverting the matrix.
When the sound pressure signal is a broadband signal, further broadband energy synthesis is performed to obtain a broadband ambiguity function as follows:
(9)
The value of each point of the narrowband or wideband ambiguity function represents the multi-path focused energy of the sound source propagating along two paths to the array at the corresponding grid point. The maximum value of the ambiguity function corresponds to the depth, horizontal distance and horizontal azimuth of the desired sound source.
Traversing and searching the maximum value of the ambiguity function, wherein the corresponding position is the estimated value of the horizontal azimuth angle of the expected sound sourceEstimated horizontal distance valueDepth estimation value:
。
Compared with the prior art, the invention has the advantages that:
When the deep sea planar array utilizes wave beam formation to estimate target parameters under a far-field plane wave model, the deep sea planar array is influenced by multi-path interference such as a shallow source direct path D and a sea surface primary reflection sound ray path S, and the target pitch angle, distance and depth are difficult to estimate. According to the deep sea seabed planar array target positioning method based on multi-path coherent focusing, the deep sea multi-path structure is utilized to improve the guide vector in the beam forming process, when the propagation time and the amplitude of a target reaching each array element of the planar array through a D, S path are calculated, three factors including the depth of a sound source, the horizontal distance and the horizontal azimuth angle are considered, and therefore the improved guide vector contains more target parameter information. After the multi-path wave beam focusing, the depth, the horizontal distance and the horizontal azimuth angle of the sound source can be estimated simultaneously by solving the position corresponding to the maximum value of the ambiguity function. Furthermore, the method further realizes the coherent accumulation of the energy of the deep sea multi-path signals, enhances the maximum value of the ambiguity function and is beneficial to weak target detection in a multi-path environment. In practical application, the construction of the guide vector only needs to calculate the multi-path propagation time and amplitude under a single frequency point, the calculated amount is small, and meanwhile, the template storage space is greatly reduced, so that the hardware cost is greatly saved.
Drawings
FIG. 1 is a coordinate system used in the embodiment, wherein FIG. 1 (a) is a top view, a solid circle is a circular array, and FIG. 1 (b) is a sound velocity profile;
FIG. 2 is a flow chart of a target positioning method for multi-pass coherent focusing by using a deep sea bottom horizontal array;
FIG. 3 is a three-dimensional broadband ambiguity function corresponding to a sound source, wherein FIG. 3 (a) is a three-dimensional broadband ambiguity function corresponding to a sound source 1, and FIG. 3 (b) is a three-dimensional broadband ambiguity function corresponding to a sound source 2;
FIG. 4 is a slice of the wideband ambiguity function for sound source 1, FIG. 4 (a) is a horizontal distance-angle slice of the wideband ambiguity function for sound source 1, FIG. 4 (b) is a depth-horizontal distance slice of the wideband ambiguity function for sound source 1;
FIG. 5 is a slice of the sound source 2 wideband ambiguity function, FIG. 5 (a) is a horizontal distance-angle slice of the sound source 2 wideband ambiguity function, FIG. 5 (b) is a depth-horizontal distance slice of the sound source 2 wideband ambiguity function;
Fig. 6 compares the result of coherent synthesis of multiple paths by the method of the sound source 2 with the result of conventional beam forming by a single path, where fig. 6 (a) is the result of the ambiguity function of the method on a distance-depth slice, fig. 6 (b) is the result of the ambiguity function considering only the direct path on a distance-depth slice, and fig. 6 (c) is the result of the ambiguity function considering only the sea surface primary reflection path on a distance-depth slice.
Detailed Description
The invention provides a target positioning method suitable for multi-path energy coherent focusing of a deep sea planar array, which comprises the steps of firstly collecting broadband sound pressure signals radiated by a target sound source through the deep sea planar array with horizontal and vertical apertures, then searching grid points of all possible positions of a potential target, calculating propagation time and amplitude corresponding to a direct path D of the sound source reaching each array element and a sea surface primary reflection path S of the sound source under different depths, target horizontal distances and target horizontal azimuth angles, constructing corresponding guide vectors and carrying out normalization processing, finally carrying out wave beam formation on each frequency point, carrying out broadband energy synthesis to obtain a broadband ambiguity function, and realizing positioning of a desired target according to the position corresponding to the maximum value of the ambiguity function.
The technical scheme of the invention is described in detail below with reference to the accompanying drawings and examples.
Step 1, establishing a three-dimensional rectangular space coordinate system and obtaining sound pressure signals of a submarine planar array:
In an embodiment, a deep sea seabed uniform circular array is considered without loss of generality. As shown in fig. 1 (a), the receiving array has 128 array elements, and the array elements are distributed on the sea floor, and the spacing between the array elements is d=5m. The projection of the origin of coordinates is at the center of the circular array, and array element 1 is located on the x-axis. The array locates the sound sources at two different positions, wherein the coordinates of the sound source 1 and the sound source 2 are (500 m ) and (1000 m,1000 m) respectively, and the depth is 80m, so that the horizontal distance from the two sound sources to the origin is 707m and 1414m respectively, and the horizontal azimuth angles relative to the x axis are 45 degrees. Both sound sources radiate broadband continuous signals with the frequency band range of 50 Hz-160 Hz. The selected marine environment is an incomplete deep sea sound channel environment, the sea depth is 1245m, and the sound velocity profile is shown in fig. 1 (b). The sound velocity at the sea floor was 1600m/s, the density was 1.6g/cm 3, and the absorption attenuation was 0.6 dB/lambda.
Step 2, constructing three-dimensional search space grid points of all possible positions of a potential target, and calculating propagation time and amplitude corresponding to a direct path D and a sea surface primary reflection path S of each grid point reaching each array element by using a ray model BELLHOP:
In the embodiment, the distance search range of the sound source is set to be 0.01km-8km, the distance search interval is set to be 0.01km, the depth search range of the sound source is set to be 1m-150m, the depth search interval is set to be 1m, the angle range of the sound source is set to be 0-360 degrees, and the angle interval is set to be 1 degree. Calculating the frequency as 105Hz of the center frequency of the sound source, and according to the ocean environment parameters and the array positions corresponding to the step 1, positioning each grid point in the search space The propagation time and the amplitude of the potential target reaching the m-th array element through the D, S path are calculated by combining the ray sound field calculation program BELLHOPAndWhereinAnd two paths corresponding to D, S respectively. The completion of the computation may be stored as a propagation time and amplitude template. The template is suitable for broadband signals, and can be directly called by different frequencies, so that the calculated amount of the method is greatly reduced.
Step 3, under each frequency point, constructing a corresponding multi-path coherent focusing guide vector according to the searched sound source depth, horizontal distance and horizontal azimuth angle, and carrying out normalization processing;
In an embodiment, the calculation in step 2 is used AndSelecting the frequency f of the concerned signal, substituting the frequency f into the formula (5) and the formula (6) to obtain a normalized steering vector。
And 4, carrying out beam forming on the given frequency point to obtain a narrow-band ambiguity function, wherein the step realizes the coherent accumulation of multipath acoustic energy, and the beam forming form is not limited to the conventional beam forming, and can also comprise an MVDR (i.e. MINIMAL VARIANCE Distortion Response) method and the like. If the broadband signal is the broadband signal, further broadband beam energy synthesis can be performed to obtain a broadband ambiguity function. The value of each point of the narrowband or wideband ambiguity function represents the multi-path focused energy of the sound source propagating along two paths to the array at the corresponding grid point. The maximum value of the ambiguity function corresponds to the depth, horizontal distance and horizontal azimuth of the desired sound source.
In an embodiment, sound source 1 and sound source 2 are positioned separately, and the implementation flow Cheng Ru is shown in fig. 2.
FIG. 3 (a) is a three-dimensional broadband ambiguity function corresponding to sound source 1, and FIG. 3 (b) is a three-dimensional broadband ambiguity function corresponding to sound source 2, broadband ambiguity functionsThe depth, the horizontal distance and the horizontal azimuth angle of the target corresponding to the maximum value in the range are the depth, the horizontal distance and the horizontal azimuth angle diagram of the expected sound source. 4 (a) and 4 (b) show horizontal distance-angle slices and depth-horizontal distance slices, respectively, of sound source 1, and the distance, depth and horizontal azimuth angles of the target can be estimated by the maximum values of the two sub-graph ambiguity functions at 700m, 80m and 45 degrees, respectively. Fig. 5 (a) and 5 (b) are horizontal distance-angle slices and depth-horizontal distance slices of the sound source 2, respectively, and the distance, depth and horizontal azimuth angles of the target can be estimated by the maximum values of the two sub-graph ambiguity functions, namely 1410m, 80m and 45 degrees. The estimated value and the actual value of the sound sources 1 and 2 are better matched, and the method is proved to be capable of realizing three-dimensional positioning of the underwater target. To further illustrate the advantages of the present method, we compare the result of the ambiguity function of the present method on the distance-depth slice (fig. 6 (a)) with the result of the ambiguity function on the distance-depth slice that only considers the direct path (fig. 6 (b)) or the sea surface primary reflection path (fig. 6 (c)), it can be seen that when only a single path is considered, both depth and distance estimates are blurred, and the distance of the sound source 2 can be accurately estimated with the present method. In addition, the maximum value of the method in fig. 6 (a) is-35.5 dB, and the maximum values respectively obtained by adopting only a single path in fig. 6 (b) and 6 (c) are-37.6 and-37.9 dB, so that the beam output maximum energy of the method is higher than that obtained by adopting only a single path by more than 2dB, and the advantage of multi-path energy coherent focusing is shown.