A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone

Abstract

The present invention provides a method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Step 1: According to the sound intensity spectrogram of the signal received by the single hydrophone, extract the characteristic distribution of its interference structure in the frequency-mode Doppler frequency shift domain; Step 2: Based on the frequency-mode Doppler frequency shift domain characteristic distribution in Step 1, establish a corresponding matching objective function; Step 3: According to the objective function in Step 2, use the simulated annealing method to extract the coordinates of the maximum value of the objective function in the sound source radial velocity-depth domain, and use this as the estimated values of the sound source velocity and depth. The present invention is used to realize the estimation of the sound source velocity and depth.

Description

Technical Field

[0001] The present invention belongs to the field, and specifically relates to a method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Background Art

[0002] The analysis of radiated noise is one of the effective means for detecting underwater sound sources. The main progress direction of vibration reduction and noise reduction technology is to suppress the line spectrum characteristics in radiated noise, while the low-frequency continuous spectrum characteristics have not been effectively suppressed. In recent years, the processing methods based on broadband signals have received the attention of scholars.

[0003] Russian scholar Kuznetsov utilized the acoustic intensity interference structure of a single hydrophone in the time-frequency domain to achieve the estimation of the radial velocity of a moving sound source. In further research, they proposed an algorithm for estimating the motion parameters of a sound source using the time-frequency interference structure based on the waveguide invariant theory and the two-dimensional Fourier transform method.

[0004] However, the method of Kuznetsov et al. requires specifically determining the normal mode that makes the main contribution in the time-frequency domain acoustic intensity interference structure. The prior information of this feature includes ocean environmental parameters and the depth of the sound source. If the depth of the sound source is not prior information, it will seriously affect the application of this algorithm in processing the time-frequency interference structure. However, Nicolas et al. extracted the dispersion characteristics of the received signal in the frequency-wavenumber domain using a horizontal array, and this feature contains the depth information of the sound source. T.C. Yang obtained the modal characteristics of signals at different receiving depths using a vertical array and also obtained an accurate estimation of the sound source depth through modal matching. Thus, it can be seen that the estimation of the sound source depth can be achieved by using the normal mode characteristics of the received signal. Summary of the Invention

[0005] The present invention provides a method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone to achieve the estimation of the sound source velocity and depth.

[0006] The present invention is achieved through the following technical solutions:

[0007] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone, the method specifically includes the following steps:

[0008] Step 1: Extract the characteristic distribution of its interference structure in the frequency-modal Doppler frequency shift domain according to the acoustic intensity spectrogram of the signal received by the single hydrophone;

[0009] Step 2: Establish a corresponding matching objective function according to the frequency-modal Doppler frequency shift domain characteristic distribution in Step 1;

[0010] Step 3: According to the objective function in Step 2, use the simulated annealing method to extract the coordinates of the maximum value of the objective function in the radial velocity-depth domain of the sound source, and use these as the estimated values of the sound source velocity and depth.

[0011] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Specifically, Step 1 is as follows: The sound field under the far-field condition of a shallow-water horizontally invariant waveguide can be written in the form of a sum of normal modes:

[0012]

[0013] where f is the angular frequency, k rn (f) is the eigenvalue of the nth normal mode, ψ n (z) is the eigenfunction of the nth normal mode, z s and z r are the sound source depth and the receiving hydrophone depth respectively, r is the distance between the sound source and the receiving hydrophone, and ρ(z s ) is the seawater density at the sound source depth;

[0014] When analyzing the interference structure of the sound field, Equation (1) is abbreviated as:

[0015]

[0016] Based on Equation (2), the far-field sound field intensity is written in the form shown in the equation:

[0017]

[0018] where the difference in eigenvalues k rm -k rn is a physical quantity related to frequency. Therefore, the exp[·] part forms an interference structure in terms of distance r and frequency f.

[0019] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Write Equation (3) as the sum of the sound field intensities of each normal mode. Then, the expression for the interference sound intensity of each normal mode is:

[0020]

[0021] Assume that the sound source is in radial motion. The received distances at different times are written as:

[0022] r = r0 + v r t (5)

[0023] Then, the expression for the sound intensity of each normal mode in the time domain of the receiving field is given:

[0024]

[0025] Based on the formula-based interference structure and using the method of Fourier spectrum estimation, the characteristic distribution of the time-frequency interference structure in the frequency-modal Doppler frequency shift domain is obtained:

[0026]

[0027] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Specifically, in step two, according to the known ocean environmental parameters, the Kraken normal mode program is used to calculate the sound field result p rplc (z s , z r , r, f) generated by a sound source at different depths under broadband conditions at the receiving depth, and thus the sound field intensity I rplc (z s , z r , r, f) of the matched field is obtained, where the receiving depth z r and the frequency f are known;

[0028] According to the sensitivity characteristics of F(f, ν) to the radial velocity v r and the sound source depth z s and its insensitivity to the initial distance r0, finally, the sound intensity of the matched field is written in the form shown in formula (8):

[0029] I rplc (z s , z r , r, f) = I rplc (z r , t, f; z s , v r ) (8)

[0030] where z s and v r are the physical quantities to be matched.

[0031] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Based on the method of formula (7), the interference structure of the sound intensity of the matched field in the time-frequency domain obtained in formula (8) is subjected to Fourier analysis in the time domain to obtain the output result in the frequency-modal Doppler frequency shift domain of the matched field:

[0032]

[0033] Set the output result in the frequency-modal Doppler frequency shift domain after processing the observed field signal as F obs (ν, f); what is used in the matching process is F obs (ν, f) and F rplc (ν, f; z s , vr )'s modulus value. According to the least squares principle, the objective function is designed as shown in the formula:

[0034]

[0035] where N f and N ν are the number of data points on the f-axis and the ν-axis respectively; finally, the objective function in formula (11) is processed in decibels, as shown in the formula:

[0036]

[0037] where, J dB (z s , v r ) is the result of processing the objective function in decibels, and J(z s , v r ) is the value of the objective function.

[0038] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Specifically, in step three, the objective function is constructed on the ambiguity plane of (z s , v r ), and the coordinates of the maximum value point therein are set Its expression is:

[0039]

[0040] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone, wherein the simulated annealing method is an adaptive simplex simulated annealing algorithm.

[0041] The beneficial effects of the present invention are:

[0042] Based on the time-frequency interference spectrogram of sound intensity, the present invention gives the Fourier spectrum estimation of the time-domain interference structure, and obtains the energy distribution characteristics of the received signal in the frequency-modal Doppler frequency shift (f, ν) domain. On this basis, the influence of different sound source radial velocities and sound source depths on the energy distribution in the frequency-modal Doppler frequency shift domain is analyzed, the feasibility of realizing sound source velocity-depth estimation with this as the matched field feature is proved, and the objective function is designed accordingly. Subsequently, the simulated annealing algorithm is used to quickly find the maximum coordinate of the objective function in the sound source radial velocity-depth domain to be estimated, and this is used as the estimated value of the sound source velocity and depth. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is the sensitivity analysis diagram of the objective function of the present invention, wherein, (a) is the sensitivity analysis diagram of the radial velocity, (b) is the sensitivity analysis diagram of the sound source depth, and (c) is the sensitivity analysis diagram of the initial distance.

[0044] Figure 2 is the LOFAR spectrogram of the received signal intensity of the present invention. Among them, (a) is the LOFAR spectrogram of the received signal intensity of the example where the sound source is close to the hydrophone, and (b) is the LOFAR spectrogram of the received signal intensity of the example where the sound source is far from the hydrophone.

[0045] Figure 3 is the characteristic distribution of the time-frequency interference structure of the present invention in the frequency-modal Doppler frequency shift domain. Among them, (a) is the characteristic distribution of the received signal intensity of the example where the sound source is close to the hydrophone in the frequency-modal Doppler frequency shift domain, and (b) is the characteristic distribution of the received signal intensity of the example where the sound source is far from the hydrophone in the frequency-modal Doppler frequency shift domain.

[0046] Figure 4 is the ambiguity function of the objective function of the present invention in the sound source radial velocity - sound source depth domain. Among them, (a) is the ambiguity function of the objective function of the example where the sound source is close to the hydrophone in the sound source radial velocity - sound source depth domain, and (b) is the ambiguity function of the objective function of the example where the sound source is far from the hydrophone in the sound source radial velocity - sound source depth domain.

[0047] Figure 5 is the schematic diagram of the change of the objective function after the iteration of the simulated annealing algorithm of the present invention. Among them, (a) is the change of the objective function after the iteration of the simulated annealing algorithm of the example where the sound source is close to the hydrophone, and (b) is the change of the objective function after the iteration of the simulated annealing algorithm of the example where the sound source is far from the hydrophone.

[0048] Figure 6 is the flowchart of the method of the present invention. Detailed implementation manners

[0049] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0050] A method for estimating the radial velocity and depth of a moving sound source based on the received signal of a single hydrophone, the method specifically includes the following steps:

[0051] Step 1: Extract the characteristic distribution of its interference structure in the frequency-modal Doppler frequency shift domain according to the intensity spectrogram of the received signal of the single hydrophone;

[0052] Step 2: Establish a corresponding matching objective function according to the frequency-modal Doppler frequency shift domain characteristic distribution in Step 1;

[0053] Step 3: According to the objective function in Step 2, use the simulated annealing method to extract the coordinates of the maximum value of the objective function in the radial velocity-depth domain of the sound source, and use this as the estimated values of the sound source velocity and depth.

[0054] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Specifically, Step 1 is as follows: The sound field under the far-field condition of a shallow-water horizontally invariant waveguide can be written in the form of a sum of normal modes:

[0055]

[0056] where f is the angular frequency, k rn (f) is the eigenvalue of the nth normal mode, ψ n (z) is the eigenfunction of the nth normal mode, z s and z r are the sound source depth and the receiving hydrophone depth respectively, r is the distance between the sound source and the receiving hydrophone, and ρ(z s ) is the seawater density at the sound source depth;

[0057] When analyzing the interference structure of the sound field, Equation (1) is abbreviated as:

[0058]

[0059] Based on Equation (2), the far-field sound field intensity is written in the form shown in the equation:

[0060]

[0061] where the difference in eigenvalues k rm -k rn is a physical quantity related to frequency. Therefore, the exp[·] part forms an interference structure in terms of the distance r and frequency f.

[0062] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Write Equation (3) in the form of the sum of the sound field intensities of each normal mode. Then, the expression for the interference sound intensity of each normal mode is:

[0063]

[0064] Assume that the sound source is in radial motion. The received distances at different times are written as:

[0065] r = r0 + v r t (5)

[0066] Then, the expression for the sound intensity of each normal mode in the time domain of the receiving field is given:

[0067]

[0068] Based on the formula-based interference structure and using the method of Fourier spectrum estimation, the characteristic distribution of the time-frequency interference structure in the frequency-modal Doppler frequency shift domain is obtained:

[0069]

[0070] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Specifically, in step two, according to the known ocean environment parameters, the Kraken normal mode program is used to calculate the acoustic field result p rplc (z s , z r , r, f) generated by a sound source at different depths under broadband conditions at the receiving depth, and thus the acoustic field intensity I rplc (z s , z r , r, f) of the matched field is obtained, where the receiving depth z r and the frequency f are known;

[0071] According to the sensitivity characteristics of F(f, ν) to the radial velocity v r and the sound source depth z s and its insensitivity to the initial distance r0, finally, the matched field acoustic intensity is written in the form shown in formula (8):

[0072] I rplc (z s , z r , r, f) = I rplc (z r , t, f; z s , v r ) (8)

[0073] where z s and v r are the physical quantities to be matched.

[0074] A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone. Based on the method of formula (7), the interference structure of the matched field acoustic intensity in the time-frequency domain obtained in formula (8) is subjected to Fourier analysis in the time domain to obtain the output result in the frequency-modal Doppler frequency shift domain of the matched field:

[0075]

[0076] Set the output result in the frequency-modal Doppler frequency shift domain after processing the observed field signal as F obs (ν, f); what is used in the matching process is F obs (ν, f) and F rplc (ν, f; z s , vr )'s modulus value. According to the least squares principle, the objective function is designed as shown in the formula:

[0077]

[0078] where N f and N ν are the number of data points on the f-axis and the ν-axis respectively; finally, the objective function in formula (11) is processed in decibels, as shown in the formula:

[0079]

[0080] where J dB (z s , v r ) is the result of processing the objective function in decibels, and J(z s , v r ) is the objective function.

[0081] The sensitivity analysis of the objective function is as Figure 1 shown.

[0082] A method for estimating the radial velocity and depth of a moving sound source based on the received signal of a single hydrophone. Specifically, step three is to construct the objective function on the ambiguity plane of (z s , v r ) and set the coordinates of the maximum value point therein Its expression is:

[0083]

[0084] A method for estimating the radial velocity and depth of a moving sound source based on the received signal of a single hydrophone. The simulated annealing method is the adaptive simplex simulated annealing algorithm.

[0085] Set the sound speed of the seawater layer to be 1500 m / s and the depth of the seawater layer to be 50 m; the longitudinal wave speed c p in the semi-infinite space of the seabed is 1800 m / s, and the density ρ2 is 1.5 g / cm 3 , and the longitudinal wave propagation loss α p is 0.4 dB / λ. Set the sound source depth to be 10 m and the radial velocity to be 4 m / s. Set two examples: 1. The initial distance is 10 km and the sound source approaches the hydrophone; 2. The initial distance is 6.8 km and the sound source moves away from the hydrophone. Set the hydrophone receiving depth to be 50 m from the seabed. The resulting received signal intensity spectrogram is as Figure 2 shown.

[0086] According to the Fourier spectrum estimation of the formula, the characteristic distribution of the time-frequency interference structure in the frequency-mode Doppler frequency shift domain as shown in Figure 2 can be obtained, asFigure 3 As shown in:

[0087] The ambiguity function of the objective function calculated by the formula in the sound source radial velocity - sound source depth domain is as Figure 4 shown in:

[0088] Since the ambiguity function obtained by traversal is slow in calculation when calculating a large range and has low estimation accuracy, the simulated annealing method is used for estimation, and the change of the objective function after iteration is as Figure 5 shown in:

[0089] The results of the sound source radial velocity and depth estimated according to the simulated annealing algorithm are shown in Table 1:

[0090] Table 1 Estimation results of sound source radial velocity and depth

[0091]

[0092]

[0093] The method for estimating the sound source radial velocity - depth based on the characteristics in the frequency - mode Doppler frequency shift domain of the interference structure proposed by the present invention is introduced in detail above. In this article, the principles and implementation manners of the present invention are elaborated by using sound source calculation examples with different motion states.

Claims

1. A method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone, characterized in that The method specifically includes the following steps: Step 1: Extract the characteristic distribution of the interference structure in the frequency-modal Doppler frequency shift domain according to the sound intensity spectrogram of the signal received by a single hydrophone; Step 2: Establish a corresponding matching objective function according to the frequency-modal Doppler frequency shift domain characteristic distribution in Step 1; Step 3: According to the objective function in Step 2, use the simulated annealing method to extract the coordinates of the maximum value of the objective function in the source radial velocity-depth domain, and use this as the estimated values of the source velocity and depth; Specifically, in the second step, according to the known ocean environmental parameters, the Kraken normal mode program is used to calculate the sound field result p generated by sound sources at different depths under broadband conditions at the receiving depth. rplc (z s ,z r ,r,f), from which the sound field intensity I of the matched field is obtained. rplc (z s ,z r ,r,f), where the receiving depth z r and the frequency f are known. According to the sensitivity characteristics of the radial velocity v with respect to F(f, ν) r , the sound source depth z s , and the insensitivity characteristics to the initial distance r0, the matched-field sound intensity is finally written in the form shown in Equation (8): I rplc (z s ,z r ,r,f) = I rplc (z r ,t,f; z s ,v r )(8) where z s and v r are physical quantities to be matched; The specific content of Step 3 is that the objective function is constructed on the fuzzy plane of (z s , v r ), and the coordinates of the maximum value point are set The expression thereof is:

2. The method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone according to claim 1, characterized in that Specifically, Step 1 is as follows. The sound field under the far-field condition of a shallow-water horizontally invariant waveguide can be written as a sum of a set of normal modes: where f is the angular frequency, k rn (f) is the eigenvalue of the nth-order normal mode, and ψ n (z) is the eigenfunction of the nth-order normal mode, where z s and z r are the source depth and the receiving hydrophone depth respectively, r is the distance between the source and the receiving hydrophone, and ρ(z s ) is the seawater density at the source depth; When analyzing the interference structure of the sound field, Equation (1) is abbreviated as: Based on Equation (2), the far-field sound field intensity is written in the form shown in the equation: where the difference in eigenvalues k rm -k rn is a physical quantity related to frequency, so the exp[·] part forms an interference structure in terms of distance r and frequency f.

3. The method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone according to claim 2, wherein Writing Equation (3) as the sum of the sound field intensities of each order of normal modes, the expression for the interference sound intensity of each order of normal modes is: Assuming that the sound source is in radial motion, the received distance at different times is written as: r = r0 + v r t(5) Then the expression for the sound intensity of each order of normal modes in the time domain of the receiving field is given: Based on the interference structure of Equation (6), using the method of Fourier spectrum estimation, obtain the characteristic distribution of the time-frequency interference structure in the frequency-modal Doppler frequency shift domain:

4. The method for estimating the radial velocity and depth of a moving sound source based on the received signal of a single hydrophone according to claim 1, wherein Based on the method of Equation (7), perform Fourier analysis in the time domain on the matched-field sound intensity interference structure in the time-frequency domain obtained in Equation (8) to obtain the output result of the matched-field in the frequency-modal Doppler frequency shift domain: Set the frequency-modal Doppler frequency shift domain output result after the observation field signal processing as F obs (ν,f); F is used in the matching process obs (ν,f) and F rplc (ν,f; z s ,v r )'s modulus value. According to the least squares principle, design the objective function as shown in the formula: where N f and N ν are the number of data points on the f-axis and the ν-axis respectively; finally, the objective function in equation (10) is processed in decibels as shown in equation (11): Among them, J dB () is the decibel processing result of the objective function, and J is the objective function value.

5. The method for estimating the radial velocity and depth of a moving sound source based on the signal received by a single hydrophone according to claim 1, characterized in that The simulated annealing method is the adaptive simplex simulated annealing algorithm.

Images