Passive ranging method of sound source based on deep sea bottom return path

Abstract

A passive ranging method for sound sources based on deep seabed return paths is disclosed, belonging to the field of underwater acoustic signal processing and passive underwater target detection and localization. This invention addresses the problem that existing methods for deep-sea sound source localization neglect the internal structure of the seabed, affecting ranging accuracy. The method includes broadband beamforming processing of the acoustic signal to extract the observed time delay difference between the direct-to-surface path wave cluster and the deep seabed acoustic ejection path wave cluster; constructing an equivalent acoustic seabed model including sound velocity gradients; simulating the sound field data of the target frequency band using sound field calculation methods; performing the broadband beamforming processing on the sound field data to obtain the multipath time delay difference copy vector between the direct-to-surface path and the deep seabed path; and constructing a matching cost function to match the observed time delay difference with the multipath time delay difference copy vector to obtain the estimated distance of the surface sound source. This invention significantly improves the accuracy of sound source ranging.

Description

Technical Field

[0001] This invention relates to a passive ranging method for acoustic sources based on the return path of deep seabed, belonging to the field of underwater acoustic signal processing and passive detection and positioning of underwater targets. Background Technology

[0002] Distance estimation of passive acoustic sources in deep sea has always been a challenging and hot topic in the field of underwater acoustics. Traditional passive localization methods mostly employ matched-field localization, which involves simulating a copy field using known marine environmental parameters and matching it with the measured acoustic signal. Although matched-field localization has been proven effective in deep-sea environments, this method is susceptible to environmental mismatch. In recent years, with the development of underwater acoustic physics research in deep-sea environments, methods for localization using the characteristics of deep-sea sound fields have received widespread attention, leading to a series of localization methods utilizing multipath angle of arrival, multipath delay characteristics, and interference fringe characteristics. Since a relatively clear multipath signal arrival structure can be obtained using a receiving array, research focus has gradually shifted to passive localization methods based on multipath delay. Existing methods are as follows:

[0003] The first method, "Multipath Time Delay Difference Analysis of Deep-Sea Seabed Sound Reflection Zone and Near-Sea Surface Sound Source Localization Technology," derives the multipath arrival time delay structure of the deep-sea seabed sound reflection zone based on virtual source theory. It proposes a method to jointly estimate the distance and depth of near-sea surface sound sources by utilizing the time delay difference between multiple seabed reflection paths. However, it is limited to near-sea surface sound sources, and the time delay difference is greatly affected by seabed topography and seabed sediment parameters.

[0004] The second method, utilizing the time delay structure of seabed reflections, proposes a method for estimating the target distance and depth in deep-sea seabed reflection zones. This method extracts the time delay structure of characteristic seabed reflection multipaths to construct a nonlinear relationship between target distance and depth, thereby achieving parameter estimation. However, this method heavily relies on the assumption of seabed smoothness; mismatch in seabed parameters will significantly reduce estimation accuracy.

[0005] The third method is a passive sound source localization method that combines the angle of arrival and multipath delay difference of the direct wave and the first-order sea surface reflected wave. This method clearly points out that changes in seabed topography and boundary roughness will distort the multipath propagation path. If these underlying environmental parameters are ignored in theoretical modeling, a serious mismatch in the underwater acoustic environment model will occur, which will lead to the inability to effectively estimate the accurate location of the near-shore water surface sound source during actual matching.

[0006] The fourth method involves using dual hydrophones to passively locate sound sources in the deep-sea direct sound zone and the sound shadow zone (seabed reflection zone). This method is extremely sensitive to prior environmental parameters when generating multipath delays in the shadow zone. If the actual acoustic characteristics of the seabed are ignored in the theoretical calculations, the severe mismatch between the sound field and the seabed model will lead to a serious misalignment of the time delay difference matching function, ultimately causing the algorithm to be unable to effectively estimate the precise spatial location of near-surface sound sources.

[0007] The fifth method is to use matched filtering technology to separate multipath sound rays arriving at the vertical linear array, extract the multipath time delay difference between the peak values ​​of the compensation signal pulses, and then establish a relationship model between the time delay difference, the elevation angle of arrival and the target spatial position to realize the joint estimation method of sound source ranging and depth; this algorithm relies too much on the ideal shallow ocean waveguide approximation and lacks consideration for matching the real acoustic parameters of the deep seabed.

[0008] The sixth method involves using a single hydrophone to extract the multipath delay difference between the direct wave and the reflected wave from the sea surface, and then using this difference as the measurement input for an extended Kalman filter. This iterative filtering method estimates the spatial location of the sound source. This type of filtering algorithm heavily relies on accurate theoretical multipath delay differences as a benchmark. If the complex parameters of the actual seabed medium are ignored, the mismatch between the theoretical propagation model and the real deep-sea environment will directly lead to systematic deviations in the filter's observation equations, thus making it impossible to effectively estimate the location of near-surface sound sources.

[0009] In summary, existing passive positioning methods based on multipath delay often simplify the seabed into a semi-infinite space model, neglecting the actual impact of deep geological structures within the seabed (such as the sound velocity gradient increasing with depth in sediment layers) on multipath delay. Directly using simplified seabed models for calculations leads to severe model mismatch, which is the core reason why existing methods cannot effectively estimate the location of near-surface sound sources and have limited overall ranging accuracy. Summary of the Invention

[0010] To address the problem that existing methods for locating deep-sea sound sources ignore the internal structure of the seabed and affect ranging accuracy when using multipath delay, this invention provides a passive ranging method for sound sources based on the deep seabed return path.

[0011] The present invention provides a passive ranging method for sound sources based on deep seabed return paths, comprising:

[0012] The acoustic signals radiated by the surface sound source were received by a vertical array deployed on the deep seabed. The acoustic signals were processed by broadband beamforming to extract the time delay difference between the direct-to-surface path wave cluster and the deep-sea acoustic ejection path wave cluster.

[0013] Based on the actual sedimentation process of the deep seabed, an equivalent acoustic seabed model including sound velocity gradient is constructed. The equivalent acoustic seabed model equates the seabed to a three-layer structure of seawater, sedimentary layer, and hard seabed, and sets a continuous sound velocity gradient that gradually increases with depth in the sedimentary layer. Based on the measured seawater sound velocity profile of the observed sea area, the sound field data of the target frequency band is simulated using the equivalent acoustic seabed model and sound field calculation method. The sound field data is then processed by the broadband beamforming to obtain the multipath time delay difference copy vector of the direct-to-surface path and the deep seabed path at different horizontal search distances within the set distance search space.

[0014] A matching cost function is constructed to match the time delay difference observation value with the multipath time delay difference copy vector. The horizontal search distance corresponding to the global minimum point of the matching cost function is used as the estimated distance of the water surface sound source.

[0015] The passive ranging method for acoustic sources based on deep seabed return paths according to the present invention includes a method for broadband beamforming processing of acoustic signals, comprising:

[0016] The acoustic signal is subjected to a Fast Fourier Transform with a time window of 10 seconds to obtain the spectral information. Then, each frequency point in the spectral information is analyzed. Perform conventional beamforming processing.

[0017] According to the passive ranging method for sound sources based on deep seabed return paths of the present invention, for each frequency point in the spectrum information The method for performing conventional beamforming processing is as follows:

[0018] The array steering vector corresponding to the vertical array for:

[0019] ,

[0020] In the formula The pitch angle, The path difference phase of the nth element in the vertical array relative to the reference element at the array center. , The number of array elements in the vertical array;

[0021] ,

[0022] In the formula The speed of sound in water, For reference array element numbering, The distance between adjacent array elements;

[0023] From the spectrum information Each element at the frequency point The frequency domain data constitutes a column vector Calculate the pitch angle beam output power :

[0024] ;

[0025] The beam output power at each frequency point within the target frequency band is incoherently accumulated to obtain a broadband beam output. :

[0026] ,

[0027] In the formula For the i-th frequency point, This represents the total number of frequency points.

[0028] According to the passive ranging method for acoustic sources based on deep seabed return paths of the present invention, the method for extracting the time delay difference observation values ​​between the direct-to-surface path wave cluster and the deep seabed acoustic ejection path wave cluster is as follows:

[0029] Extract the spectrum of the direct-to-surface path and the deep-sea acoustic ejection path for each time window:

[0030] ,

[0031] ,

[0032] In the formula For the spectrum of the direct-to-sea-surface path, For the time window j, The pitch angle for the direct path to the sea surface. The spectrum of the acoustic ejection path in the deep seabed. The pitch angle for the deep-sea acoustic catapult path;

[0033] Construct a complete spectrum that satisfies conjugate symmetry :

[0034] ,

[0035] In the formula The time corresponding to the time window. for or The target frequency band is 20Hz-100Hz.

[0036] The time-domain signal was obtained through broadband Fourier synthesis. :

[0037] ;

[0038] In the formula The time domain moment within the time window;

[0039] Cross-correlation function between the direct-to-surface path and the deep-sea acoustic ejection path signal. for:

[0040] ,

[0041] In the formula For time delay, To correspond to the time-domain signal of the deep seabed acoustic ejection path, This corresponds to the time-domain signal of the direct-to-sea-surface path;

[0042] Based on cross-correlation function The time delay difference observation value was calculated.

[0043] According to the passive ranging method for sound sources based on deep seabed return paths of the present invention, a matching cost function is constructed as follows:

[0044] ,

[0045] In the formula For the matching cost function, To set the horizontal search distance within the search space, For the time delay difference observation value, This is the multipath delay difference copy vector.

[0046] According to the passive ranging method for sound sources based on deep seabed return paths of the present invention, the matching cost function values ​​under all horizontal search distances are traversed, and the horizontal search distance corresponding to the global minimum point of the matching cost function is selected. As an estimated distance for sound sources on the water surface.

[0047] According to the passive ranging method for sound sources based on deep seabed return paths of the present invention, The value can be 64 or 128.

[0048] According to the passive ranging method for sound sources based on the return path of deep seabed of the present invention, the sedimentary layer is provided with a continuous sound velocity gradient that gradually increases linearly with depth.

[0049] According to the passive ranging method for sound sources based on deep seabed return paths of the present invention, in the equivalent acoustic seabed model, the seawater depth is 4360m, the sound velocity in the sediment layer is 1550m / s-2050m / s, and the density is 1.1g / cm³. 3 -1.9g / cm 3 The attenuation coefficient is 0.1, and the thickness is 470m; the sound velocity on the hard seabed is 2100m / s, and the density is 2.1g / cm³. 3 The attenuation coefficient is 0.1.

[0050] According to the passive ranging method for sound sources based on the return path of deep seabed of the present invention, the sound field calculation method of the equivalent acoustic seabed model includes the normal mode method or the wavenumber integration method.

[0051] The beneficial effects of this invention are as follows: This invention addresses the problems of model mismatch and near-range ranging failure caused by existing deep-sea multipath time-delay positioning methods that simplify the seabed into a semi-infinite space and ignore deep geological sound velocity gradients. It innovatively constructs an equivalent acoustic seabed model incorporating continuous sound velocity gradients and transforms the measured high-energy deep-sea acoustic ejection phenomenon into reliable multipath time delay difference characteristics, which are then accurately matched with theoretical copy time delays. Simulation experiments and theoretical analysis results show that, compared to traditional semi-infinite seabed models, the equivalent acoustic seabed model constructed in this invention effectively eliminates the severe theoretical ranging bias caused by deep geological mismatch. This method exhibits superior distance estimation performance across the entire observation range, significantly improving theoretical ranging accuracy and successfully avoiding near-range ranging failures. It provides a practical engineering solution for passive ranging of surface targets using seabed ejection waves. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the ranging scenario of the passive ranging method for sound sources based on the deep seabed return path described in this invention.

[0053] Figure 2 This is a schematic diagram of an equivalent acoustic seabed model that includes a continuous sound velocity gradient.

[0054] Figure 3 This is a schematic diagram of the simulation results of broadband beamforming under an equivalent acoustic seabed model.

[0055] Figure 4 This is a schematic diagram of the cross-correlation time delay simulation results under the equivalent acoustic seabed model;

[0056] Figure 5 This is a comparison chart of the sound source estimation distance obtained using the layered seabed model (equivalent acoustic seabed model) of this invention and the existing semi-infinite seabed model;

[0057] Figure 6 This is a comparison chart of the estimation errors of ranging results using the layered seabed model (equivalent acoustic seabed model) of this invention and the existing semi-infinite seabed model. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Specific Implementation Method 1: Combination Figure 1 and Figure 2 As shown, this invention provides a passive ranging method for sound sources based on deep seabed return paths, including:

[0060] The acoustic signals radiated by the surface sound source were received by a vertical array deployed on the deep seabed. The acoustic signals were processed by broadband beamforming to extract the time delay difference between the direct-to-surface path wave cluster and the deep-sea acoustic ejection path wave cluster.

[0061] Based on the actual sedimentation process of the deep seabed, an equivalent acoustic seabed model including sound velocity gradient is constructed. The equivalent acoustic seabed model equates the seabed to a three-layer structure of seawater, sedimentary layer, and hard seabed, and sets a continuous sound velocity gradient that gradually increases with depth in the sedimentary layer. Based on the measured seawater sound velocity profile of the observed sea area, the sound field data of the target frequency band is simulated using the equivalent acoustic seabed model and sound field calculation method. The sound field data is then processed by the broadband beamforming to obtain the multipath time delay difference copy vector of the direct-to-surface path and the deep seabed path at different horizontal search distances within the set distance search space.

[0062] A matching cost function is constructed to match the time delay difference observation value with the multipath time delay difference copy vector. The horizontal search distance corresponding to the global minimum point of the matching cost function is used as the estimated distance of the water surface sound source.

[0063] Furthermore, assuming the target sound source is a moving target on the water surface (such as a ship), such as Figure 1 As shown, the surface target ship passes through the test sea area where the vertical array is located along a certain track. The acoustic signals collected by the vertical array are processed to extract the time delay difference between the deep seabed reflection path and the sea surface-direct path. Figure 1 The CPA exam is a recent development.

[0064] Methods for broadband beamforming processing of acoustic signals include:

[0065] The acoustic signal was subjected to a Fast Fourier Transform with a time window of 10 seconds to obtain the spectral information, corresponding to a frequency resolution of 0.1 Hz; then, for each frequency point in the spectral information... Perform conventional beamforming processing.

[0066] For each frequency point in the spectrum information The method for performing conventional beamforming processing is as follows:

[0067] The array steering vector corresponding to the vertical array for:

[0068] ,

[0069] In the formula The pitch angle, The path difference phase of the nth element in the vertical array relative to the reference element at the array center. , The number of array elements in the vertical array;

[0070] ,

[0071] In the formula The speed of sound in water, For reference array element numbering, The distance between adjacent array elements;

[0072] From the spectrum information Each element at the frequency point The frequency domain data constitutes a column vector At a single frequency point At this point, calculate the pitch angle. beam output power :

[0073] ;

[0074] The beam output power at each frequency point within the target frequency band is incoherently accumulated to obtain a broadband beam output. :

[0075] ,

[0076] In the formula For the i-th frequency point, This represents the total number of frequency points.

[0077] The method for extracting the time delay difference observations between the direct-to-sea-surface path wave cluster and the deep-sea-ejection path wave cluster is as follows:

[0078] Based on broadband beamforming processing, the spectra of the direct-to-surface path and the deep-sea acoustic ejection path are extracted for each time window:

[0079] ,

[0080] ,

[0081] In the formula For the spectrum of the direct-to-sea-surface path, For the time window j, The pitch angle for the direct path to the sea surface. The spectrum of the acoustic ejection path in the deep seabed. The pitch angle for the deep-sea acoustic catapult path;

[0082] After extracting the spectral information of each path, it is converted into a time-domain waveform using a wideband Fourier transform method. The target frequency band is selected as 20-100Hz. Before performing wideband Fourier transform, a complete spectrum satisfying conjugate symmetry needs to be constructed to ensure that a real waveform is obtained after the inverse transform. :

[0083] ,

[0084] In the formula The time corresponding to the time window. for or The target frequency band is 20Hz-100Hz. Indicates conjugate.

[0085] The time-domain signal was obtained through broadband Fourier synthesis. :

[0086] ;

[0087] In the formula The time domain moment within the time window;

[0088] To calculate the time delay between the two paths, the time-domain waveforms are normalized and cross-correlated. The cross-correlation function between the direct-to-surface path and the deep-sea acoustic catapult path signal is calculated. for:

[0089] ,

[0090] In the formula For time delay, To correspond to the time-domain signal of the deep seabed acoustic ejection path, This corresponds to the time-domain signal of the direct-to-sea-surface path;

[0091] Based on cross-correlation function The time delay difference observation value was calculated.

[0092] like Figure 2 As shown, during the experiment, the sound velocity profile of the seawater in the test area was measured using a CTD (Conductivity, Temperature, Depth) instrument. Combined with the seismic profile and deep-sea drilling (ODP / IODP) logging data of the same area, the seabed of the observed area was equivalently divided downwards into a three-layer medium structure: "seawater layer - sedimentary layer - hard substrate". In the sedimentary layer model, the sound velocity was set as a continuous sound velocity gradient profile that gradually increases linearly downwards with depth. Figure 2 middle , , These represent the speeds of sound in seawater, sedimentary layers, and hard seabed, respectively. , , ρ represents the density of seawater, sediment layer, and hard seabed, respectively; h represents the thickness of the sediment layer.

[0093] Based on the constructed equivalent acoustic seabed model, sound field data in the processing frequency band (20-100Hz) is simulated using normal mode or wavenumber integration methods, and then subjected to broadband beamforming, such as... Figure 3 As shown, by extracting the spectral information of the theoretically calculated direct-to-sea-surface path and the deep-sea-reflection path, and then performing cross-correlation and time delay processing on the two, the following can be obtained: Figure 4 The time delay simulation results are shown.

[0094] Furthermore, a matched field processing is performed on the time delay difference observations and the multipath time delay difference copy vector.

[0095] Construct the matching cost function:

[0096] ,

[0097] In the formula For the matching cost function, To set the horizontal search distance within the search space, For the time delay difference observation value, This is the multipath delay difference copy vector.

[0098] Within the defined distance search space, iterate through all matching cost function values ​​at all horizontal search distances, and find the horizontal search distance corresponding to the global minimum point of the matching cost function. As an estimated distance for sound sources on the water surface.

[0099] As an example, The value can be 64 or 128.

[0100] In this embodiment, the deposition layer is provided with a continuous sound velocity gradient that gradually increases linearly with depth.

[0101] As an example, in the equivalent acoustic seabed model, the seawater depth is 4360m, the sound velocity in the sediment layer is 1550m / s-2050m / s, and the density is 1.1g / cm³. 3 -1.9g / cm 3 The attenuation coefficient is 0.1, and the thickness is 470m; the sound velocity on the hard seabed is 2100m / s, and the density is 2.1g / cm³. 3 The attenuation coefficient is 0.1.

[0102] Following the above processing procedure, the measured multipath delay is matched and searched with the simulated multipath delay. The simulation results are as follows: Figure 5 and Figure 6 As shown, it can be observed that the estimation results of the continuous seabed in this invention are consistent with the actual seabed, while the distance estimation error of the semi-infinite seabed is extremely large at close range, gradually decreasing with increasing distance. When the distance is less than 4.2 km, the error is greater than 100%, and when the distance is greater than 14 km, the error stabilizes below 10%. Moreover, when the distance is greater than 16 km, the error is close to the actual distance. This simulation result fully demonstrates that even when facing close-range targets and complex seabed geological conditions, the signal processing and matching method based on the deep-sea continuous seabed proposed in this invention can still effectively overcome the fundamental mismatch error and accurately achieve distance estimation of surface sound sources.

[0103] As an example, the sound field calculation method for the equivalent acoustic seabed model includes the normal mode method or the wavenumber integration method.

[0104] This implementation constructs an equivalent seabed model incorporating a continuous sound velocity gradient, replacing the traditional semi-infinite simplified model. By precisely matching the measured multipath delay difference with the copy delay generated by this theoretical model, the problem of near-range ranging failure caused by seabed mismatch is completely solved, significantly improving the robustness and theoretical accuracy of passive ranging of deep-sea targets.

[0105] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A passive ranging method for sound sources based on deep seabed return paths, characterized in that... include, The acoustic signals radiated by the surface sound source were received by a vertical array deployed on the deep seabed. The acoustic signals were processed by broadband beamforming to extract the time delay difference between the direct-to-surface path wave cluster and the deep-sea acoustic ejection path wave cluster. Based on the actual sedimentation process of the deep seabed, an equivalent acoustic seabed model including sound velocity gradient is constructed. The equivalent acoustic seabed model equates the seabed to a three-layer structure of seawater, sedimentary layer, and hard seabed, and sets a continuous sound velocity gradient that gradually increases with depth in the sedimentary layer. Based on the measured seawater sound velocity profile of the observed sea area, the sound field data of the target frequency band is simulated using the equivalent acoustic seabed model and sound field calculation method. The sound field data is then processed by the broadband beamforming to obtain the multipath time delay difference copy vector of the direct-to-surface path and the deep seabed path at different horizontal search distances within the set distance search space. A matching cost function is constructed to match the time delay difference observation value with the multipath time delay difference copy vector. The horizontal search distance corresponding to the global minimum point of the matching cost function is used as the estimated distance of the water surface sound source.

2. The passive ranging method for sound sources based on deep seabed return paths according to claim 1, characterized in that, Methods for broadband beamforming processing of acoustic signals include: The acoustic signal is subjected to a Fast Fourier Transform with a time window of 10 seconds to obtain the spectral information. Then, each frequency point in the spectral information is analyzed. Perform conventional beamforming processing.

3. The passive ranging method for sound sources based on deep seabed return paths according to claim 2, characterized in that, For each frequency point in the spectrum information The method for performing conventional beamforming processing is as follows: The array steering vector corresponding to the vertical array for: ， In the formula The pitch angle, The path difference phase of the nth element in the vertical array relative to the reference element at the array center. , The number of array elements in the vertical array; ， In the formula The speed of sound in water, For reference array element numbering, The distance between adjacent array elements; From the spectrum information Each element at the frequency point The frequency domain data constitutes a column vector Calculate the pitch angle beam output power : ； The beam output power at each frequency point within the target frequency band is incoherently accumulated to obtain a broadband beam output. : ， In the formula For the i-th frequency point, This represents the total number of frequency points.

4. The passive ranging method for sound sources based on deep seabed return paths according to claim 3, characterized in that, The method for extracting the time delay difference observations between the direct-to-sea-surface path wave cluster and the deep-sea-ejection path wave cluster is as follows: Extract the spectrum of the direct-to-surface path and the deep-sea acoustic ejection path for each time window: ， ， In the formula For the spectrum of the direct-to-sea-surface path, For the time window j, The pitch angle for the direct path to the sea surface. The spectrum of the acoustic ejection path in the deep seabed. The pitch angle for the deep-sea acoustic catapult path; Construct a complete spectrum that satisfies conjugate symmetry : ， In the formula The time corresponding to the time window. for or The target frequency band is 20Hz-100Hz. The time-domain signal was obtained through broadband Fourier synthesis. : ； In the formula The time domain moment within the time window; Cross-correlation function between the direct-to-surface path and the deep-sea acoustic ejection path signal. for: ， In the formula For time delay, To correspond to the time-domain signal of the deep seabed acoustic ejection path, This corresponds to the time-domain signal of the direct-to-sea-surface path; Based on cross-correlation function The time delay difference observation value was calculated.

5. The passive ranging method for sound sources based on deep seabed return paths according to claim 4, characterized in that, Construct the matching cost function: ， In the formula For the matching cost function, To set the horizontal search distance within the search space, For the time delay difference observation value, This is the multipath delay difference copy vector.

6. The passive ranging method for sound sources based on deep seabed return paths according to claim 5, characterized in that, Iterate through all the matching cost function values ​​for each horizontal search distance, and find the horizontal search distance corresponding to the global minimum point of the matching cost function. As an estimated distance for sound sources on the water surface.

7. The passive ranging method for sound sources based on deep seabed return paths according to claim 3, characterized in that, The value can be 64 or 128.

8. The passive ranging method for sound sources based on deep seabed return paths according to claim 1, characterized in that, The deposition layer is configured with a continuous sound velocity gradient that gradually increases linearly with depth.

9. The passive ranging method for sound sources based on deep seabed return paths according to claim 8, characterized in that, In the equivalent acoustic seabed model, the seawater depth is 4360m, the sound velocity in the sediment layer is 1550m / s-2050m / s, and the density is 1.1g / cm³. 3 -1.9g / cm 3 The attenuation coefficient is 0.1, and the thickness is 470m; the sound velocity on the hard seabed is 2100m / s, and the density is 2.1g / cm³. 3 The attenuation coefficient is 0.

1.

10. The passive ranging method for sound sources based on deep seabed return paths according to claim 1, characterized in that, The sound field calculation methods for the equivalent acoustic seabed model include the normal mode method or the wavenumber integration method.

Images