Passive ranging method and system for bubble pulse sound source based on time-domain warping transform

Abstract

The present invention provides a passive ranging method and system for a bubble pulse sound source based on time-domain warping transformation, including: performing complex cepstrum analysis on a pulse signal to obtain a complex cepstrum signal; searching for peak points of the complex cepstrum signal, taking multiple points before and after the peak points for binomial fitting, and interpolating at the peak point according to the fitting result; returning the interpolated complex cepstrum signal to the time-domain signal; performing time-domain warping transformation on the time-domain signal to obtain (W hy )(t), and performing time-frequency analysis on (W hy )(t) to obtain (W hy )(f); extracting each order of single-frequency mode in (W hy )(f) through the Hough transform, and obtaining each mode time-frequency curve for each single-frequency mode through an inverse transformation function; comparing each mode time-frequency curve with the time-frequency analysis result of the time-domain signal; calculating to obtain each mode dispersion curve; substituting each mode time-frequency curve and each mode dispersion curve into a cost function to search for the minimum value. The present invention can achieve passive ranging of a bubble pulse sound source.

Description

Technical Field

[0001] The present invention relates to the technical field of passive ranging of pulsed sound sources, and in particular, to a method for passive ranging of pulsed sound sources containing bubble pulsation interference, and more particularly, to a method and system for passive ranging of bubble pulsed sound sources based on time-domain warping transformation. Background Technique

[0002] Due to the influence of sea surface and seabed acoustic reflections and medium inhomogeneity in shallow water waveguides, the acoustic signals propagating in the waveguide are typical multimodal and dispersive signals. The normal modes of each order are aliased together in the time-frequency plane, and traditional signal processing methods cannot effectively separate them. However, the dispersion characteristics of the normal modes of each order in the signal contain the position information of the sound source. Therefore, when studying the passive localization of underwater targets, how to effectively separate and utilize the dispersion characteristics of the normal modes is an important research content in the field of underwater acoustics.

[0003] The ideal pulse signal propagating in a shallow water waveguide can separate the normal modes of each order through warping transformation. However, during the generation of the pulse signal, a series of bubble pulsations often occur, which seriously interfere with the normal mode characteristics of the pulse signal, greatly reducing the ability of the warping transformation to separate the normal modes of this bubble pulse sound source, and the error of the ranging method based on the dispersion characteristics of the warping transformation will also increase significantly or even fail. Summary of the Invention

[0004] Aiming at the defects in the prior art, the present invention provides a method and system for passive ranging of bubble pulse sound sources based on time-domain warping transformation.

[0005] According to a method and system for passive ranging of bubble pulse sound sources based on time-domain warping transformation provided by the present invention, the scheme is as follows:

[0006] In the first aspect, a method for passive ranging of bubble pulse sound sources based on time-domain warping transformation is provided. The method includes:

[0007] Step S1: Perform cepstrum analysis on the pulse signal x(t) containing bubble pulsation to obtain the cepstrum signal

[0008] Step S2: Search for the peak points of the cepstrum signal , and take a symmetric plurality of points before and after the peak points for binomial fitting, and interpolate at the peak points according to the fitting results;

[0009] Step S3: Return the interpolated cepstrum signal to the time-domain signal

[0010] Step S4: Perform on the time-domain signal Perform a time-domain warping transform and resample in the time domain to make the phase a linear function of the resampling time. The signal after the time-domain warping transform is (W hy )(t). Perform time-frequency analysis on (W hy )(t) to obtain (W hy )(f);

[0011] Step S5: Extract each order of single-frequency mode in (W hy )(f) through the Hough transform, and obtain the time-frequency curve of each mode through the inverse transform function m is the order;

[0012] Step S6: Compare the modal dispersion similarity between the time-frequency curves of each mode and the time-frequency analysis result of the time-domain signal . If the trajectories are similar, go to Step S7; otherwise, repeat Step S5;

[0013] Step S7: Calculate the dispersion curves of each mode according to the known sound field parameters and the parameters to be estimated

[0014] Step S8: Substitute the time-frequency curves of each mode and the dispersion curves of each mode into the cost function to search for the minimum value, which is the distance of the bubble pulse sound source.

[0015] Preferably, the pulse signal x(t) is the convolution of the ideal pulse signal x0(t) and the impulse response h(t);

[0016] x(t) is expressed as:

[0017] x(t) = x0(t) * h(t)

[0018] where h(t) = δ(t) + a(δ(t - τ)), δ(t) is the impulse function; a represents the amplitude attenuation coefficient;

[0019] The complex cepstrum of x(t) is:

[0020]

[0021] where F and F - represent the Fourier transform and the inverse Fourier transform respectively; X0(ω) and H(ω) represent the Fourier transforms of x0(t) and h(t) respectively.

[0022] Preferably, the signal of the time-domain signal after the time-domain warping transform is:

[0023]

[0024] Among them, W hy represents the use of the time coordinate transformation function transformed signal; M represents the normal mode order; C m represents the amplitude of the m-th normal mode; j represents the imaginary unit; f cm represents the transformed single-mode frequency; t represents time.

[0025] Preferably, the step S7 includes: calculating each modal dispersion curve through the sound field propagation model according to the known sound field parameters and the parameters to be estimated

[0026] Among them, the known sound field parameters include: water depth d, sound speed c in water w , water density ρ w ;

[0027] The parameters to be estimated include: the distance of the bubble pulse sound source seabed sound speed time correction factor of the transceiver system

[0028] Preferably, the cost function in the step S8 includes:

[0029]

[0030] Among them, f represents the frequency of the dispersion curve.

[0031] In a second aspect, a passive ranging system for a bubble pulse sound source based on time-domain warping transformation is provided. The system includes:

[0032] Module M1: performing complex cepstrum analysis on the pulse signal x(t) containing bubble pulsation to obtain a complex cepstrum signal

[0033] Module M2: searching for the peak points of the complex cepstrum signal , taking a symmetric plurality of points before and after the peak points for binomial fitting, and interpolating at the peak points according to the fitting results;

[0034] Module M3: returning the interpolated complex cepstrum signal to the time-domain signal

[0035] Module M4: performing time-domain warping transformation on the time-domain signal , performing time-domain resampling to make the phase a linear function of the resampling time. The signal after time-domain warping transformation is (W hy )(t), and performing time-frequency analysis on (W hy )(t) to obtain (W hy )(f);

[0036] Module M5: Extract the single - frequency modes of each order in (W hy )(f) through the Hough transform, and obtain the time - frequency curves of each mode through the inverse transform function m is the order;

[0037] Module M6: Compare the modal dispersion similarity between the time - frequency curves of each mode and the time - frequency analysis result of the time - domain signal If the trajectories are similar, enter step S7; otherwise, repeat step S5;

[0038] Module M7: Calculate the dispersion curves of each mode according to the known sound - field parameters and the parameters to be estimated

[0039] Module M8: Substitute the time - frequency curves of each mode and the dispersion curves of each mode into the cost function to search for the minimum value, which is the distance of the bubble - pulse sound source.

[0040] Preferably, the pulse signal x(t) is the convolution of the ideal pulse signal x0(t) and the impulse response h(t);

[0041] x(t) is expressed as:

[0042] x(t)=x0(t)*h(t)

[0043] where h(t)=δ(t)+a(δ(t - τ)), δ(t) is the impulse function; a represents the amplitude attenuation coefficient;

[0044] The complex cepstrum of x(t) is:

[0045]

[0046] where F and F - respectively represent the Fourier transform and the inverse Fourier transform; X0(ω) and H(ω) respectively represent the Fourier transforms of x0(t) and h(t).

[0047] Preferably, the signal after the time - domain warping transformation of the time - domain signal is:

[0048]

[0049] where W hy represents the signal after transformation using the time - coordinate transformation function ; M represents the order of the normal mode; C m represents the amplitude of the m - th normal mode; j represents the imaginary unit; f cm represents the single - mode frequency after transformation; t represents time.

[0050] Preferably, the module M7 includes: calculating each modal dispersion curve through a sound field propagation model according to known sound field parameters and parameters to be estimated

[0051] Among them, the known sound field parameters include: water depth d, sound speed c in water w , water density ρ w ;

[0052] The parameters to be estimated include: the distance of the bubble pulse sound source Seabed sound speed Time correction factor of the transceiver system

[0053] Preferably, the cost function in the module M8 includes:

[0054]

[0055] Among them, f represents the frequency of the dispersion curve

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] 1. The present invention uses homomorphic filtering complex cepstrum analysis, and uses the processing method of searching for the spikes in the real part of the complex cepstrum to effectively eliminate the interference of bubble pulsation on the normal wave characteristics of the pulse signal and restore its normal wave characteristics;

[0058] 2. The present invention uses the time-domain warping transform combined with the Hough transform to effectively extract the modal dispersion curves of each order of the signal, and realizes the sound source ranging by matching and searching with the dispersion curve of the propagation model, thereby proving the effectiveness and accuracy of the present method for ranging the pulse signal with bubble pulsation interference. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] By reading the detailed description of the non-limiting embodiments with reference to the following drawings, other features, objects and advantages of the present invention will become more obvious:

[0060] Figure 1 is the principle block diagram of the present invention;

[0061] Figure 2 is the time-frequency domain ideal pulse signal diagram;

[0062] Figure 3 is the bubble pulse signal diagram;

[0063] Figure 4 is the time-frequency domain ideal pulse signal diagram after warping;

[0064] Figure 5 is the bubble pulse signal diagram after warping;

[0065] Figure 6 It is the real part of the complex cepstrum of the bubble pulse signal;

[0066] Figure 7 It is the time-frequency domain diagram of the signal after the bubble pulse signal returns to the time domain through interpolation in the complex cepstrum domain;

[0067] Figure 8 It is the time-frequency domain diagram after warping;

[0068] Figure 9 It is the diagram of the dispersion curve extracted from the bubble pulse signal and the optimal dispersion curve calculated by the propagation model;

[0069] Figure 10 It is the result diagram of the matching search distance of the cost function. Specific implementation mode

[0070] The present invention will be described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. It should be noted that those of ordinary skill in the art can make several changes and improvements without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0071] The embodiment of the present invention provides a passive ranging method for a bubble pulse sound source based on time-domain warping transformation. By using the matching search between the dispersion curve extracted after warping transformation with smoothing in the complex cepstrum domain and the dispersion curve calculated by the propagation model, passive ranging of the bubble pulse sound source is realized. It is applicable to the field of transient signal positioning such as single hydrophone and sonar for explosion, air gun sound source, etc., and can be applied to underwater warning air-dropped buoys and other underwater acoustic countermeasure processes. Refer to Figure 1 As shown, the method specifically includes:

[0072] Step S1: Perform complex cepstrum analysis on the pulse signal x(t) containing bubble pulsation to obtain the complex cepstrum signal

[0073] Among them, the pulse signal x(t) is the convolution of the ideal pulse signal x0(t) and the impulse response h(t);

[0074] x(t) is expressed as:

[0075] x(t) = x0(t) * h(t)

[0076] Among them, h(t) = δ(t) + a(δ(t - τ), δ(t) is the impulse function; a represents the amplitude attenuation coefficient;

[0077] The complex cepstrum of x(t) is:

[0078]

[0079] Among them, F and F - respectively represent the Fourier transform and the inverse Fourier transform; X0(ω) and H(ω) respectively represent the Fourier transforms of x0(t) and h(t).

[0080] Step S2: Take the real part for peak search, take 5 points before and after the search peak n = kn0 (a total of 10 points, that is, kn0 - 5,..., kn0 - 1, kn0 + 1,..., kn0 + 5) for binomial fitting, and then interpolate at the original peak point kn0 according to the fitting result.

[0081] Step S3: Return the interpolated complex cepstrum signal to the time-domain signal Step S3 is the inverse process of Step S1.

[0082] Step S4: Perform a time-domain warping transformation on the time-domain signal and perform time-domain resampling to make the phase a linear function of the resampling time. The signal after the time-domain warping transformation is (W hy )(t). Perform time-frequency analysis on (W hy )(t) to obtain (W hy )(f).

[0083] The time-domain signal after the time-domain warping transformation is:

[0084]

[0085] Among them, W hy represents the signal after transformation using the time coordinate transformation function ; M represents the normal mode order; C m represents the amplitude of the m-th normal mode; j represents the imaginary unit; f cm represents the single-mode frequency after transformation; t represents time. The signal after the warping transformation becomes a single-frequency signal with the cut-off frequency of this order of normal mode as the signal frequency.

[0086] Step S5: Extract each order of single-frequency mode in (W hy )(f) through the Hough transform, and obtain the time-frequency curve of each mode through the inverse transformation function which is the dispersion curve where m is the order.

[0087] Step S6: Compare the time-frequency analysis results of each mode time-frequency curve with the time-domain signal for modal dispersion similarity. If the trajectories are similar, enter Step S7; otherwise, repeat Step S5.

[0088] Step S7: According to the known sound field parameters and the parameters to be estimated, calculate the modal dispersion curves through the sound field propagation model Kraken

[0089] Among them, the known sound field parameters include: water depth d, sound speed c in water w , water density ρ w ; the parameters to be estimated include: the distance of the bubble pulse sound source seabed sound speed time correction factor of the transceiver system

[0090] Step S8: Substitute each modal time-frequency curve and each modal dispersion curve into the cost function to search for the minimum value, which is the distance of the bubble pulse sound source.

[0091] The cost function in Step S8 includes:

[0092]

[0093] Among them, f represents the frequency of the dispersion curve.

[0094] Next, the present invention will be described in more detail with reference to the accompanying drawings.

[0095] As Figure 2 , Figure 3 , Figure 4 , Figure 5 shown, it is a comparison diagram of the time-frequency domain analysis results after performing time-frequency domain analysis and warping transformation on the ideal pulse signal and the bubble pulse signal respectively. From the comparison of Figure 2 , Figure 4 , it can be seen that each order of mode can be clearly seen in the ideal pulse signal, while the Figure 3 and Figure 5 modal orders of the bubble pulse signal are mixed together, and the existing methods cannot effectively separate them.

[0096] Figure 6 It is the real part result diagram of performing cepstrum analysis on the bubble pulse signal x(t) . It can be seen from the figure that the bubble pulsation exists in the form of positive and negative alternating spikes, and the method of Step S2 is used to fit and interpolate at the spikes.

[0097] Figure 7 and Figure 8 are the signals after the fitting and interpolation are returned to the time domain The time-frequency analysis and the comparison diagram of the time-frequency analysis after warping transformation are respectively made. It can be seen from the figure that the modes of the signal after interpolation fitting are effectively separated, indicating that the processing method of the present invention can effectively remove the interference of bubbles on the normal wave propagation characteristics of the pulse signal and restore the normal wave characteristics of the signal.

[0098] Figure 9 The solid line in the figure is the dispersion curve of each order of normal wave extracted by using the Hough transform method and the inverse warping transform. Figure 10 The circles in the figure are the dispersion curves calculated by using the propagation model. In the figure, the one closest to has a high similarity, indicating that the propagation model designed by the present invention is relatively close to the actual sound field.

[0099] Figure 10 This is the search result of the cost function. It can be seen from the figure that the minimum value of the cost function, that is, the distance estimation value, is 7.4 km, the actual distance of the sound source is 8 km, and the ranging error is about 7.5%. This indicates that the method of the present invention can achieve passive ranging of the bubble pulse sound source with a small ranging error.

[0100] The embodiment of the present invention provides a method and system for passive ranging of a bubble pulse sound source based on time-domain warping transformation. Using homomorphic filtering cepstrum analysis and the method of searching and smoothing the spikes in the cepstrum domain to restore its normal wave characteristics, the sound source ranging is realized, and the ranging error is about 7.5%. This proves the effectiveness and accuracy of this method for ranging the pulse signal with bubble pulsation interference. Moreover, this method has a simple algorithm and is easy to implement, and is expected to be widely applied in the fields of single hydrophone and sonar transient signal positioning.

[0101] Those skilled in the art know that in addition to implementing the system and its various devices, modules, and units provided by the present invention in the form of pure computer-readable program code, the method steps can be logically programmed to make the system and its various devices, modules, and units provided by the present invention be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers to achieve the same function. Therefore, the system and its various devices, modules, and units provided by the present invention can be regarded as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be regarded as the structure within the hardware component; the devices, modules, and units for implementing various functions can also be regarded as both software modules for implementing the method and the structure within the hardware component.

[0102] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. Without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A passive ranging method for bubble pulse sound sources based on time-domain warping transformation, characterized in that, including: Step S1: Perform complex cepstrum analysis on the pulse signal x(t) containing bubble pulsation to obtain the complex cepstrum signal Step S2: Search for cepstrum signal for peak points. Take a plurality of symmetric points before and after the peak point for binomial fitting, and perform interpolation at the peak point according to the fitting result. Step S3: Return the interpolated complex cepstrum signal to the time-domain signal Step S4: For the time-domain signal perform a time-domain warping transformation, conduct time-domain resampling, so that the phase becomes a linear function of the resampling moment. The signal after the time-domain warping transformation is W hy (t). Perform time-frequency analysis on W hy (t) to obtain W hy (f); Step S5: Extract W through Hough transform hy For each single-frequency mode in (f), obtain the time-frequency curve of each mode through the inverse transform function m is the order; Step S6: Compare the time-frequency dispersion similarity of each modal time-frequency curve with the time-frequency analysis result of the time-domain signal to determine the modal dispersion similarity. If the trajectories are similar, proceed to Step S7; otherwise, repeat Step S5. Step S7: Calculate each modal dispersion curve based on the known sound field parameters and the parameters to be estimated Step S8: Substitute the time-frequency curves of each mode and the dispersion curves of each mode into the cost function to search for the minimum value, which is the distance of the bubble pulse sound source; The cost function in step S8 includes: where f represents the frequency of the dispersion curve.

2. The passive ranging method for bubble pulse sound source based on time-domain warping transform according to claim 1, characterized in that The pulse signal x(t) is the convolution of the ideal pulse signal x0(t) and the impulse response h(t); x(t) is expressed as: x(t) = x0(t) * h(t) where h(t) = δ(t) + a(δ(t - τ)), δ(t) is the impulse function; a represents the amplitude attenuation coefficient; The complex cepstrum of x(t) is: where F and F - denote the Fourier transform and the inverse Fourier transform, respectively; X0(ω) and H(ω) denote the Fourier transforms of x0(t) and h(t), respectively.

3. The passive ranging method for bubble pulse sound source based on time domain warping transformation according to claim 1, wherein The time-domain signal The signal after performing the time-domain warping transformation is: Among them, W hy represents the signal after using the time coordinate transformation function ; M represents the normal mode order; C m represents the amplitude of the m-th normal mode; j represents the imaginary unit; f cm represents the single-mode frequency after transformation; t represents time.

4. The passive ranging method for bubble pulse sound sources based on time-domain warping transformation according to claim 1, characterized in that The step S7 includes: calculating each modal dispersion curve through a sound field propagation model according to known sound field parameters and parameters to be estimated Among them, the known acoustic field parameters include: water depth d, sound speed c in water w , water density ρ w ; The parameters to be estimated include: the distance of the bubble pulse sound source Seabed sound speed Time correction factor of the transceiver system 5. A passive ranging system for bubble pulse sound sources based on time-domain warping transformation, characterized in that including: Module M1: Perform cepstrum analysis on the pulse signal x(t) containing bubble pulsation to obtain the cepstrum signal Module M2: Search for cepstrum signals Find the peak points, take multiple symmetric points before and after the peak points for binomial fitting, and interpolate at the peak points according to the fitting results; Module M3: Return the interpolated complex cepstrum signal to the time-domain signal Module M4: For the time-domain signal Perform a time-domain warping transformation on it, conduct time-domain resampling to make the phase a linear function of the resampling moment, and the signal after the time-domain warping transformation is W hy (t). Conduct time-frequency analysis on W hy (t) to obtain W hy (f); Module M5: Extract W through Hough transform hy For each single-frequency mode in (f), obtain the time-frequency curve of each mode through the inverse transform function m is the order; Module M6: Compare the time-frequency analysis results of the time-domain signal with the time-frequency curves of each modality to determine the modal dispersion similarity. If the trajectories are similar, proceed to step S7; otherwise, repeat step S5. with the time-domain signal to determine the modal dispersion similarity. If the trajectories are similar, proceed to step S7; otherwise, repeat step S5. Module M7: Calculate each modal dispersion curve based on known sound field parameters and parameters to be estimated Module M8: Substitute the time-frequency curves of each mode and the dispersion curves of each mode into the cost function and search for the minimum value, which is the distance of the bubble pulse sound source; The cost function in module M8 includes: where f represents the frequency of the dispersion curve.

6. The passive ranging system for bubble pulse sound sources based on time-domain warping transformation according to claim 5, characterized in that, The pulse signal x(t) is the convolution of the ideal pulse signal x0(t) and the impulse response h(t); x(t) is expressed as: x(t) = x0(t) * h(t) where h(t) = δ(t) + a(δ(t - τ)), δ(t) is the impulse function; a represents the amplitude attenuation coefficient; The complex cepstrum of x(t) is: where F and F - denote the Fourier transform and the inverse Fourier transform, respectively; X0(ω) and H(ω) denote the Fourier transforms of x0(t) and h(t), respectively.

7. The passive ranging system for bubble pulse sound source based on time-domain warping transform according to claim 5, characterized in that, The time-domain signal The signal after performing time-domain warping transformation is: Among them, W hy represents the transformed signal using the time coordinate transformation function ; M represents the normal mode order; C m represents the amplitude of the m-th normal mode; j represents the imaginary unit; f cm represents the single-mode frequency after transformation; t represents time.

8. The passive ranging system for bubble pulse sound sources based on time-domain warping transformation according to claim 5, characterized in that, The module M7 includes: calculating modal dispersion curves respectively through a sound field propagation model according to known sound field parameters and parameters to be estimated Among them, the known acoustic field parameters include: water depth d, sound speed c in water w , water density ρ w ; The parameters to be estimated include: the distance of the bubble pulse sound source The sound speed in the sea floor The time correction factor of the transceiver system

Images