Autonomous detection of target line spectrum and motion element estimation method based on deep sea vertical array

Abstract

The application belongs to the technical field of underwater acoustic physics, and particularly relates to a target line spectrum autonomous detection and motion element estimation method based on a deep sea vertical array, which comprises the following steps: performing Fourier transform and beam forming processing on a time domain sound field collected by a vertical hydrophone array, and normalizing the sound field into a plurality of sub-band ranges; performing matching processing on a pre-calculated target sailing track to be detected and a measured sound field transformed into a beam-time domain; and estimating the frequency of a line spectrum by using a one-dimensional signal peak value estimation algorithm, so as to obtain the line spectrum frequency, sailing speed and nearest passing distance of the target to be detected. The line spectrum energy ratio relationship is used to autonomously identify and calculate the line spectrum frequency of an underwater acoustic target, and the average sailing speed and the nearest horizontal distance of the underwater acoustic target to the vertical array can be calculated.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of underwater acoustic physics, and particularly relates to a target line spectrum autonomous detection and motion element estimation method based on a deep sea vertical array. BACKGROUND

[0002] With the development of underwater target vibration reduction and noise reduction technology, the ship radiated noise source level is greatly reduced, and the performance of the traditional detection method based on wideband energy accumulation is limited. The ship radiated noise is a mixed spectrum composed of continuous spectrum and line spectrum, and the continuous spectrum contributes most of the energy of the wideband source level, while the low-frequency line spectrum usually has the characteristics of high intensity and good phase stability, which is a key feature for identifying underwater targets. The traditional line spectrum extraction method is through pre-processing, spectrum peak extraction and post-processing, etc., to extract the line spectrum feature under the condition of occasional random interference weak line spectrum. This line spectrum extraction method is feasible under the condition that the near field, environment and target are single. However, in the actual marine environment, the line spectrum interference of the environmental noise is also stable and continuous. Based on the characteristics of the underwater target-environment coupled sound field, the line spectrum feature of the underwater target is extracted, which has more advantages in the actual marine environment, and is also the premise for solving the key problem of underwater target monitoring in the typical deep sea environment. SUMMARY

[0003] The purpose of the present application is to overcome the defects of the prior art, and a target line spectrum autonomous detection and motion element estimation method based on a deep sea vertical array is proposed.

[0004] In order to achieve the above purpose, the present application proposes a target line spectrum autonomous detection and motion element estimation method based on a deep sea vertical array, which comprises:

[0005] Step 1) Fourier transform and beamforming processing are performed on the time domain sound field collected by the vertical hydrophone array, and are normalized to a plurality of sub-band ranges;

[0006] Step 2) The pre-calculated target track to be detected is matched with the measured sound field transformed into the beam-time domain;

[0007] Step 3) The frequency of the line spectrum is estimated by using a one-dimensional signal peak value estimation algorithm, and the line spectrum frequency, speed and closest approach distance of the target to be detected are obtained.

[0008] As an improvement of the above method, the vertical hydrophone array of step 1) is an N-element uniformly distributed vertical linear array with an array element spacing d.

[0009] As an improvement of the above method, step 1) specifically comprises:

[0010] The time domain signal p(t,z) collected by the vertical hydrophone array is subjected to Fourier transform and beamforming processing, and is normalized to a plurality of sub-band rangesn Perform a Fast Fourier Transform to obtain the frequency domain signal P(f,z). n ):

[0011]

[0012] Where f represents frequency, t represents time, and z n Indicates the signal amplitude;

[0013] Then P(f,z) consists of F frequencies and N array elements. n The received sound field matrix P(f,z) is formed. n );

[0014] Based on the depth z of the vertical linear array r The average speed of sound in seawater c(z) r The guiding vector w(f,θ) for the pointing angle θ of the vertical line array is obtained from the following formula:

[0015]

[0016] Where d is the element spacing of the vertical linear array, and T represents transpose;

[0017] This will simultaneously point to θ1, θ2, ..., θ L The L beams form the steering vector matrix A(f,θ):

[0018] A(f,θ)=[w(f,θ1),w(f,θ2),…,w(f,θ L )]

[0019] From the received sound field matrix P(f,z) n The beam response matrix B(f,θ) obtained from the guide vector matrix A(f,θ) and the frequency-grazing angle is:

[0020] B(f,θ)=|P(f,z n A(f,θ)| 2 ,

[0021] The matrix B(f,θ) has a size of F×N. θ N θ It represents the number of sweep angles;

[0022] Normalize matrix B(f,θ) to a certain frequency band using downsampling, resulting in B′(f) n ,θ m The standardized beam response matrix B′(f,θ) is composed of ) where the frequency band of the matrix is ​​the range of equally spaced frequency points (f1,f2,…,f n ,f n+1 ,…,f Nf), f n+1 -f n =Δf, B′(f) n ,θ m ) is (f n -Δf<f i <f n B(f) frequency band +Δf) i ,θ m The maximum value of )

[0023]

[0024] With time interval Δt, Δt = t n+1 -t n ,Will Total N t The matrices B′(f,θ) at each time step form a three-dimensional matrix S(f,θ,t), where the elements of the matrix...

[0025] As an improvement to the above method, step 2) specifically includes:

[0026] Based on the depth z of the vertical linear array r The average speed of sound in seawater c(z) r The depth z of the target to be measured s The average speed of sound in seawater c(z) s The glancing angle θ of the sound ray at the vertical line array. r The sound velocity profile c(z) is used to obtain the horizontal propagation distance r(θ) of the intrinsic sound ray according to the following formula. r )for:

[0027]

[0028] According to θ r The values ​​of θ1, θ2, ..., θ L The corresponding horizontal propagation distances r1, r2, ..., r are calculated using the above formula. L ;

[0029] The closest distance between the target and the position of the vertical line array is r. min Given a speed of v and a horizontal distance R between the target and the nearest point on the flight path at time t1, the horizontal projected distance r between the target and the vertical line array is obtained using the following formula. sr for:

[0030]

[0031] According to r(θ) r Combining this with the above formula, we obtain the grazing angle θ of the vertical linear array. rThe correspondence between the time t and the frequency f is:

[0032]

[0033] Thus, a two-dimensional curve is drawn, and the maximum trajectory energy E(f min ) of the frequency f i is obtained in the parameter range of v, R, and r i .

[0034]

[0035] As an improvement of the above method, the step 3) specifically comprises:

[0036] For the frequency of the underwater acoustic target line spectrum of the frequency range of N f frequencies, it is judged whether the energy of a frequency f i satisfies the following formula, and if yes, the frequency f i is the target line spectrum frequency:

[0037]

[0038] Wherein, E0 is a preset detection threshold;

[0039] The parameters of the maximum trajectory energy corresponding to the target line spectrum frequency are taken, so as to obtain the speed v and the closest passing distance R

[0040] Compared with the prior art, the advantages of the present application are:

[0041] 1. The line spectrum frequency of the underwater acoustic target can be autonomously identified and calculated by using the line spectrum energy ratio relationship of the present application; 2. The average speed and the closest horizontal distance of the underwater acoustic target to the vertical array can also be calculated by the present application. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 Fig. 1 is a schematic diagram of the underwater target motion trajectory and the vertical array position;

[0043] Figure 2 Fig. 5 is the estimation result of the target line spectrum frequency;

[0044] Figure 3 Fig. 8 is the estimation result of the speed and the closest passing distance of the target line spectrum frequency of 105 Hz;

[0045] Figure 4 Fig. 10 is the estimation result of the speed and the closest passing distance of the target line spectrum frequency of 126 Hz. ​Detailed Implementation

[0046] The purpose of this invention is to utilize a vertical array detection system to autonomously detect and identify underwater target information with line spectrum characteristics, including characteristic frequencies, speeds, and closest passing distances, thereby achieving long-term energy detection and autonomous line spectrum identification, and providing acoustic field characteristic information for underwater acoustic target depth estimation.

[0047] This invention presents a method for autonomous target line spectrum identification using a vertical array detection system. First, the time-domain sound field measured by the vertical hydrophone array is processed by Fourier transform and beamforming and normalized to multiple sub-frequency bands. Then, the pre-calculated target trajectory is matched with the measured sound field transformed into the beam-time domain. Finally, the frequency of the line spectrum is estimated using a one-dimensional signal peak estimation algorithm to obtain target information such as the target's line spectrum frequency, speed, and closest passing distance.

[0048] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0049] Example 1

[0050] To achieve the above objectives, Embodiment 1 of the present invention provides a method for autonomous target line spectrum detection and motion element estimation based on a deep-sea vertical array, comprising the following steps:

[0051] The following Fast Fourier Transform is used to transform the time-domain signal p(t,z) received by the vertical hydrophone array. n Transformed to the frequency domain P(f,z) n ),

[0052]

[0053] Take the depth z of the receiving array r The average speed of sound in seawater is c(z). r For an N-element uniformly distributed vertical linear array with element spacing d, the guiding vector w(f,θ) is:

[0054]

[0055] This will simultaneously form directions pointing to θ1, θ2, ..., θ L The L beams constitute the steering vector matrix A(f):

[0056] A(f,θ)=[w(θ1),w(θ2),…,w(θ L (3)

[0057] The beam response matrix B(f,θ) of frequency-grazing angle is then:

[0058] B(f,θ)=|P(f,z n)A(f,θ) 2 , (4)

[0059] where P(f,z n ) is the sound field matrix composed of the received sound field P(f,z n ) of F frequencies and N elements, B(f,θ) is F×N θ , N θ is the number of grazing angles.

[0060] Then, the matrix B(f,θ) is normalized to a certain frequency range by down-sampling, and the frequency range is equidistant frequency points f n+1 -f n = Δf, (n = 1, 2, …, N f -1), and the new matrix B'(f,θ) is composed of B'(f n ,θ m ). Where B'(f n ,θ m ) satisfies,

[0061]

[0062] That is, B'(f n ,θ m ) takes the maximum value of B(f n ,θ i ) in the frequency range (f n - Δf < f i < f m + Δf).

[0063] With a time interval Δt, the matrix B'(f,θ) of N t n+1 -t n = Δt, N t times is composed of a three-dimensional matrix S(f,θ,t), where

[0064]

[0065] Generally, given the grazing angle at the sound source, the horizontal propagation distance of the eigenray can be calculated,

[0066]

[0067] where θ r is the grazing angle of the sound ray at the receiving array, z s and z r are the sound source depth and receiving depth, respectively, and c(z) is the sound speed distribution with depth, i.e., the sound speed profile. Then, assuming θ rθ1, θ2, …, θ L The corresponding horizontal propagation distances r1, r2, …, r L may be obtained by numerical calculation.

[0068] Suppose that the motion state of the target is uniform linear motion, the nearest distance of the target from the receiving array is r min , the target speed is v, the current time is t1, the horizontal distance of the target position from the nearest point of the course at the current time is R, and the horizontal projection distance r sr of the underwater target from the vertical array can be expressed as

[0069]

[0070] According to the corresponding relationship between the propagation distance r and the grazing angle θ r of the receiving array depth, the corresponding relationship between the grazing angle θ r of the receiving array depth and the time t can be obtained by using the above formula. is a two-dimensional curve on a two-dimensional image, where v, R, r min are parameters to be assigned. In a certain parameter range, v, R, r min are assigned, for example: 0 < v < 20, -10 km < R < 10 km, 0 km ≤ r min < 10 km. Then, the trajectory energy under different assignments in the parameter range is calculated, and the maximum trajectory energy E(f i ) at the frequency f i is obtained.

[0071]

[0072] The frequency range of the underwater acoustic target line spectrum is determined as N f frequencies. If the energy i of the frequency point f satisfies the following relationship,

[0073]

[0074] where E0 is a detection threshold, which can be set according to experimental data processing experience and should generally be greater than 3 decibels. If the frequency point f i satisfies formula (10), the frequency point f i is the output target line spectrum frequency, and the parameters that obtain the maximum trajectory energy at the frequency are the average speed and the nearest passing distance of the target motion.

[0075] Example 2

[0076] Embodiment 2 of the present invention provides an autonomous target line spectrum detection and motion element estimation system based on a deep-sea vertical array, implemented based on the method of Embodiment 1. The system specifically includes: a transformation processing module, a matching processing module, and a detection estimation module; wherein,

[0077] The transformation processing module is used to perform Fourier transform and beamforming processing on the time-domain sound field acquired by the vertical hydrophone array, and normalize it to multiple sub-frequency bands.

[0078] The matching processing module is used to match the pre-calculated flight trajectory of the target under test with the measured sound field transformed into the beam-time domain.

[0079] The detection estimation module is used to estimate the frequency of the line spectrum using a one-dimensional signal peak estimation algorithm, and to obtain the line spectrum frequency, speed and closest passing distance of the target under test.

[0080] Simulation Examples

[0081] Figure 1 Let r be the target's trajectory and the vertical array position. Assume the target's motion is uniform linear motion, and the closest distance to the receiving array position is r. min =r0, the target speed is v, the current time is t1, and the horizontal distance between the target position and the nearest point on the route at the current time is R.

[0082] Figure 2 The target maximum trajectory energy E(f) calculated based on real data i The frequency range is 50Hz-150Hz. According to formula (10), let E0 = 5dB, the target characteristic spectrum frequencies can be estimated to be 105Hz and 126Hz.

[0083] Figure 3 This is a two-dimensional graph of S(f,θ,t) at f = 105 Hz, with the horizontal axis representing time and the vertical axis representing the horizontal distance from the vertical array. The dashed line in the graph represents the target trajectory estimated by this method, which is basically consistent with the target's maximum energy (the darker the color, the greater the energy). The estimated target speed is... The shortest distance traveled was

[0084] Figure 4 This is a two-dimensional graph of S(f,θ,t) at f = 126 Hz, with the horizontal axis representing time and the vertical axis representing the horizontal distance from the vertical array. The dashed line in the graph represents the target trajectory estimated by this method, which is basically consistent with the target's maximum energy (the darker the color, the greater the energy). The estimated target speed is... The shortest distance traveled was

[0085] Finally, it should be noted that the above examples are merely used to illustrate the technical solutions of the present application but not to limit the present application. Although the present application is explained in detail with reference to the examples, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, and all of these should be covered in the scope of the claims of the present application.

Claims

1. A method for target line spectrum autonomous detection and motion element estimation based on deep sea vertical array, the method comprising: Step 1) Fourier transform and beamforming processing are performed on the time domain sound field collected by the vertical hydrophone array, and are normalized into a plurality of sub-band ranges; Step 2) The pre-calculated target track to be detected is matched with the measured sound field transformed into the beam-time domain; Step 3) A one-dimensional signal peak estimation algorithm is used to estimate the frequency of the line spectrum, and the line spectrum frequency, speed and closest approach distance of the target to be detected are obtained; specifically comprising: For frequency range Total N f The frequencies of the underwater acoustic target line spectrum at each frequency point are used to determine the frequency f at a given frequency point. i energy If the following formula is satisfied, then the frequency point f is determined. i Target line spectrum frequency: Wherein, E0 is a preset detection threshold; Taking the parameter of the maximum track energy corresponding to the target line spectrum frequency, so as to obtain the speed of the target to be measured and the distance by the nearest 2. The method of claim 1, wherein, The vertical hydrophone array of the step 1) is an N-element uniformly distributed vertical linear array with an element spacing d.

3. The method of claim 2, wherein, The step 1) specifically comprises: The time-domain signal p(t,z) acquired by the vertical hydrophone array n Perform a Fast Fourier Transform to obtain the frequency domain signal P(f,z). n ): where f represents frequency, t represents time, z n represents signal amplitude; Then P(f,z) consists of F frequencies and N array elements. n The received sound field matrix P(f,z) is formed. n ); According to the depth z at which the vertical line array is located r , the average sound speed c(z r ) of the sea water, the steering vector w(f, θ) of the pointing angle θ of the vertical line array is obtained by the following formula: Wherein, d is the element spacing of the vertical linear array, and T represents transposition; This will simultaneously point to θ1, θ2, ..., θ L The L beams form the steering vector matrix A(f,θ): A(f, θ) = [w(f, θ1), w(f, θ2),..., w(f, θN) ]T L )] The frequency-swept angle beam response matrix B(f, θ) is obtained from the received sound field matrix P(f, z) and the steering vector matrix A(f, θ) as: n B(f, θ) = P(f, z)A(f, θ) B(f, θ) = |P(f, z n ) A(f, θ)| 2 , Wherein, the matrix B(f, θ) has a size of F × N θ , N θ is the number of grazing angles; Normalize matrix B(f,θ) to a certain frequency band using downsampling to obtain B′(f) n ,θ m The standardized beam response matrix B′(f,θ) is composed of , where the frequency band of the matrix is ​​the range of equally spaced frequency points. f n+1 -f n =Δf, B′(f) n ,θ m ) is (f n -Δf<f i <f n B(f) frequency band +Δf) i ,θ m The maximum value of ) at intervals Δt, Δt = t n+1 -t n , the N t matrices B'(f, θ) of N points in time form a three-dimensional matrix S(f, θ, t), wherein the elements of the matrix are 4. The method of claim 3, wherein, The step 2) specifically comprises: Based on the depth z of the vertical linear array r The average speed of sound in seawater c(z) r The depth z of the target to be measured s The average speed of sound in seawater c(z) s The glancing angle θ of the sound ray at the vertical line array. r The sound velocity profile c(z) is used to obtain the horizontal propagation distance r(θ) of the intrinsic sound ray according to the following formula. r )for: According to the value of θ r , the corresponding horizontal propagation distances r1, r2, …, r L are calculated by the above formula respectively. L ; The nearest distance of the target to be measured from the position of the vertical linear array is r min , the horizontal distance of the target to be measured from the nearest point of the course at the current time t1 is R, and the horizontal projection distance r of the target to be measured from the vertical linear array is obtained according to the following formula sr : According to r(θ r ), the grazing angle θ r of the vertical linear array is obtained by combining the above formula, and the corresponding relationship between the grazing angle θ r and time t is: Thus a two-dimensional curve is plotted, for v, R, r in the set parameter range min Assign values, and calculate the trajectory energy under different assignments respectively, from which the maximum trajectory energy E(f i ) at the frequency f i is obtained.

5. A system for deep-sea vertical array based target line spectrum autonomous detection and motion element estimation based on the method of claim 1, the system comprising: A transformation processing module, a matching processing module and a detection estimation module; wherein, The transformation processing module is used for Fourier transform and beamforming processing on the time domain sound field collected by the vertical hydrophone array, and normalizing into a plurality of sub-band ranges; The matching processing module is used for matching the pre-calculated target track to be detected with the measured sound field transformed into the beam-time domain; The detection estimation module is used for using a one-dimensional signal peak estimation algorithm to estimate the frequency of the line spectrum, and obtaining the line spectrum frequency, speed and closest approach distance of the target to be detected.

Images