Method and System for Extracting Gearbox Reduction Ratio Features of Ship Radiated Noise

Abstract

This invention discloses a method and system for extracting features of the gearbox reduction ratio of ship radiated noise. The method includes the following steps: acquiring ship radiated noise signals through passive sonar; performing frame-based preprocessing on the received signals to obtain framed signals; performing DEMON spectrum estimation on the framed signals to synthesize a ship radiated noise DEMON spectrum time-frequency map and extracting the target ship's propeller shaft frequency; performing LOFAR spectrum estimation on the framed signals to synthesize a ship radiated noise LOFAR spectrum time-frequency map and extracting the target ship's main engine ignition frequency; establishing a numerical relationship between the propeller shaft frequency, main engine ignition frequency, and gearbox reduction ratio through a mechanical coupling structure model; and calculating and extracting the target ship's gearbox reduction ratio features based on the numerical relationship. This method can provide stable, reliable, and interpretable feature parameters for ship target identification under complex operating conditions at different speeds.

Description

Technical Field

[0001] This invention belongs to the field of underwater acoustic target recognition technology, and particularly relates to a method, system, computer equipment, and readable storage medium for extracting features of the reduction ratio of a ship's radiated noise gearbox. Background Technology

[0002] Ship radiated noise identification is an important research area in underwater acoustic signal processing, attracting widespread attention and in-depth discussion from researchers both domestically and internationally. Especially with the development of deep learning theory, more and more algorithm models are being applied to underwater acoustic target identification. However, using deep learning for target identification suffers from poor physical interpretability in the network feature extraction and mapping process, and requires a large amount of data. How to extract stable features reflecting the essential characteristics of targets under complex operating conditions, and how to achieve accurate classification and identification of ship targets using a limited number of features under small sample conditions, is a very challenging problem in this field.

[0003] Ship targets, such as main engine speed, number of cylinders, shaft frequency, and number of propeller blades, usually exist in the DEMON spectrum of ship radiated noise signals in the form of line spectra. These parameters are often related to mechanical structure and operating speed, which not only directly affect the ship's operating conditions such as speed, but also have clear physical meaning. In some cases, the target type can be identified by just one or a few of these parameters.

[0004] DEMON spectrum is the demodulated spectrum of cavitation noise from ship propellers. Regarding propeller cavitation noise spectrum, there are two relatively mature and reliable mathematical models with theoretical backgrounds: pulse sequence theory and cavitation group theory. Furthermore, with the development and progress of technologies and algorithms such as DEMON spectrum estimation, propeller shaft frequency extraction, and blade number identification, the characteristic information contained in the DEMON harmonic line spectrum cluster is being extracted and utilized more effectively.

[0005] The LOFAR spectrum, representing the low-frequency portion of the short-time Fourier transform spectrum of ship radiated noise, is widely used as input to convolutional network models in ship target recognition. The abundant characteristic information it carries primarily originates from the periodic operation of mechanical equipment and the low-frequency vibrations of the ship's structure. Actual data processing of measured ship radiated noise data shows that the LOFAR spectrum line frequencies also exhibit certain multiple relationships. However, due to environmental noise interference, the signal-to-noise ratio is typically low, and line spectrum defects are present. Enhancement algorithms are required before extracting the line spectrum.

[0006] However, the correspondence between LOFAR line spectra and operating cycles and vibration frequencies is not yet clear, posing a challenge to research on ship identification characteristics and extraction of inherent features. Furthermore, the limited number of existing inherent feature parameters restricts the ability to identify radiated noise from ships operating under complex conditions such as different speeds.

[0007] In summary, in the identification of ship radiated noise under complex operating conditions such as different speeds, there is an urgent need for a stable and inherent feature with clear physical meaning at different speeds to improve the reliability and accuracy of ship target identification systems. Summary of the Invention

[0008] To address the aforementioned problems, the present invention aims to provide a method and system for extracting features of the gearbox reduction ratio of ship radiated noise, which can provide stable, reliable, and interpretable feature parameters for ship target identification under complex operating conditions at different speeds.

[0009] To achieve the above objectives, the technical solution of the present invention is as follows: a method for extracting the gearbox reduction ratio feature of ship radiated noise, comprising the following steps: acquiring ship radiated noise signals through passive sonar; performing frame preprocessing on the received signals to obtain framed signals; performing DEMON spectrum estimation on the framed signals to synthesize a ship radiated noise DEMON spectrum time-frequency diagram and extracting the target ship propeller shaft frequency value; performing LOFAR spectrum estimation on the framed signals to synthesize a ship radiated noise LOFAR spectrum time-frequency diagram and extracting the target ship main engine ignition frequency value; establishing a numerical relationship between propeller shaft frequency, main engine ignition frequency, and gearbox reduction ratio through a mechanical coupling structure model; and calculating and extracting the gearbox reduction ratio feature of the target ship based on the numerical relationship.

[0010] Preferably, acquiring ship radiated noise signals via passive sonar and performing frame preprocessing on the received signals to obtain framed signals further includes: receiving the ship radiated noise time-domain signal x(n) acquired by the passive sonar system: x(n)=[x(1),x(2),…,x(f s [),…,x(M)], where f s M represents the signal sampling rate, and M represents the number of time-domain sampling points.

[0011] The passive sonar received signal was preprocessed by framing to obtain:

[0012] Where L is the number of signal sampling points per frame, K is the number of signal frames, a 5-second signal is taken as one frame, and the duration of the overlapping signal between two adjacent frames is 1 second.

[0013] Preferably, performing DEMON spectrum estimation on the framed signals to synthesize a DEMON spectrum time-frequency diagram of the ship's radiated noise, and extracting the target ship's propeller shaft frequency value further includes: performing variational mode decomposition on the framed signals; selecting the (k+1)th frame signal X from the framed signals. k Empirical mode decomposition is performed on X to obtain an adaptive decomposition order j. k Performing j-th order variational mode decomposition yields j intrinsic mode components: Calculate the correlation coefficient of the narrowband signal for X.k The j intrinsic mode components are correlated pairwise using narrowband envelope correlation to obtain the correlation coefficient matrix: Among them, R j Let be the correlation coefficient between the j-th intrinsic mode component and other intrinsic mode components of different orders; for Square detection and power spectrum analysis are performed on each natural mode to obtain the envelope spectrum corresponding to the j natural mode components, and R is calculated according to the narrowband envelope correlation coefficient. kj Weighted fusion yields the DEMON spectrum P of the k-th frame signal. k : By combining the DEMON spectra of the K-frame signals, the time-frequency diagram of the DEMON spectrum of the ship's radiated noise received signal is obtained. Extract the fundamental frequency value f from the harmonic line spectrum cluster of the DEMON spectrum time-frequency diagram of the ship's radiated noise signal. z This frequency value is the ship's propeller shaft frequency.

[0014] Preferably, performing LOFAR spectrum estimation on the framed signals, synthesizing the ship's radiated noise LOFAR spectrum time-frequency map, and extracting the target ship's main engine ignition frequency value further includes: for each frame signal Power spectral density estimation is performed to obtain the power spectrum of the signal in that frame. Where f = [1 / f0, 2 / f0, ..., f s f0 is the power spectrum frequency resolution; the power spectrum of each frame of signal is synthesized into a time-frequency power spectrum of ship radiated noise. This process is equivalent to short-time Fourier transform. The 0-300Hz frequency band of the time-frequency power spectrum is taken as the LOFAR spectrum of ship radiated noise. Extracting the fundamental frequency value f from the harmonic line spectrum cluster of the LOFAR spectrum time-frequency diagram of the ship's radiated noise signal. d The fundamental frequency line spectrum value is the ignition frequency of a single cylinder of the ship's main engine; wherein, in the case of missing line spectrum, the fundamental frequency line spectrum value is calculated based on the frequency difference between two adjacent higher harmonic line spectra.

[0015] Preferably, establishing the numerical relationship between propeller shaft frequency, main engine ignition frequency, and gearbox reduction ratio through a mechanically coupled structural model further includes: establishing the single-cylinder ignition frequency f of the ship's main engine cylinder. d The relationship between crankshaft speed and cylinder rotation speed is as follows: For an i-stroke diesel engine, for each ignition, the crankshaft rotates 1 / 2 revolutions. d With crankshaft speed f q satisfy Introducing the gearbox reduction ratio σ, the ship's propeller main shaft speed f is established. z Relationship with main engine crankshaft speed f q :f q =σ·f zThe paper presents the frequency mapping relationship between the ship's radiated noise modulation spectrum and the low-frequency line spectrum, as well as the relationship between the propeller shaft speed and the ignition frequency of a single main engine cylinder, and uses this relationship for gearbox reduction ratio calculation.

[0016] Preferably, the calculation and extraction of the target ship's gearbox reduction ratio feature based on the numerical relationship further includes: inputting the extracted ship propeller shaft frequency and ship main engine single-cylinder ignition frequency into the established ship radiated noise modulation spectrum and low-frequency line spectrum frequency mapping relationship model to calculate the gearbox reduction ratio; and verifying the stability and separability of the gearbox reduction ratio feature at different speeds based on the radiated noise data of the ship target at different speeds.

[0017] Based on the same concept, this invention also provides a ship radiated noise gearbox reduction ratio feature extraction system, comprising: an acquisition module for acquiring ship radiated noise signals via passive sonar and performing frame-based preprocessing on the received signals to obtain framed signals; a DEMON spectrum estimation module for performing DEMON spectrum estimation on the framed signals, synthesizing a ship radiated noise DEMON spectrum time-frequency diagram, and extracting the target ship propeller shaft frequency value; a LOFAR spectrum estimation module for performing LOFAR spectrum estimation on the framed signals, synthesizing a ship radiated noise LOFAR spectrum time-frequency diagram, and extracting the target ship main engine ignition frequency value; a numerical relationship establishment module for establishing a numerical relationship between propeller shaft frequency, main engine ignition frequency, and gearbox reduction ratio through a mechanical coupling structure model; and a calculation module for calculating and extracting the target ship gearbox reduction ratio feature based on the numerical relationship.

[0018] Preferably, it also includes a preprocessing module for receiving the ship radiated noise time-domain signal x(n) collected by the passive sonar system: x(n)=[x(1),x(2),…,x(f s [),…,x(M)], where f s M represents the signal sampling rate, and M represents the number of time-domain sampling points.

[0019] The passive sonar received signal was preprocessed by framing to obtain:

[0020] Where L is the number of signal sampling points per frame, K is the number of signal frames, a 5-second signal is taken as one frame, and the duration of the overlapping signal between two adjacent frames is 1 second.

[0021] Based on the same concept, the present invention also provides a computer device, comprising: a memory for storing a processing program; and a processor, wherein the processor, when executing the processing program, implements the ship radiated noise gearbox reduction ratio feature extraction method described in any one of the above embodiments.

[0022] Based on the same concept, the present invention also provides a readable storage medium storing a processing program, which, when executed by a processor, implements the ship radiated noise gearbox reduction ratio feature extraction method described above.

[0023] Because the present invention adopts the above technical solution, it has the following advantages and positive effects compared with the prior art:

[0024] 1. This invention provides a new inherent feature for ship target identification. This feature corresponds to the ship's own mechanical structure, has a clear physical meaning, and exhibits good stability under complex operating conditions such as different speeds. Since different ship targets have different gearbox reduction ratios, and even for the same type or model of ship targets, the gearbox reduction ratio can vary due to manufacturing errors and other factors, this feature can be used to distinguish ship targets.

[0025] 2. In the estimation of ship radiated noise DEMON spectrum, this invention improves the traditional DEMON analysis technique by introducing variational mode decomposition to replace the traditional bandpass filter and adopting narrowband envelope correlation multi-subband weighted fusion technology to adaptively and fully extract the modulation information of ship radiated noise signal, which is beneficial to improving the signal-to-noise ratio of ship radiated noise DEMON line spectrum.

[0026] 3. Based on existing research on the LOFAR spectrum characteristics of ship radiated noise, this invention clarifies for the first time the actual physical meaning of the LOFAR line spectrum, proposes a method for extracting the ignition frequency of a single cylinder of the ship's main engine from the LOFAR line spectrum, and a technique for extracting this parameter by utilizing the frequency difference of adjacent higher harmonic line spectra when the line spectrum is missing.

[0027] 4. Based on the mechanical structure model, this invention establishes the frequency mapping relationship between the main engine ignition frequency, crankshaft speed, and main shaft speed, and provides the modulation spectrum and low-frequency line spectrum of ship radiated noise. It also provides a method for calculating the reduction ratio of ship target gearbox, which can achieve fast and accurate feature extraction and target recognition with small sample size. The computational load is small, which is beneficial to the actual deployment of ship target recognition system. Attached Figure Description

[0028] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, wherein:

[0029] Figure 1 This is an overall flowchart of the feature extraction of ship radiated noise gearbox reduction ratio characteristics according to the present invention;

[0030] Figure 2 This is a flowchart of the DEMON spectrum estimation process;

[0031] Figure 3 The DEMON spectrum of target A at different speeds; Figure 3 (a) Speed ​​4.9 knots, shaft frequency 3.0 Hz; Figure 3 (b) Speed ​​6.8 knots, shaft frequency 4.2 Hz; Figure 3 (c) Speed ​​8.6 knots, shaft frequency 3.0 Hz; Figure 3 (d) Speed ​​10.5 knots, shaft frequency 7.0 Hz;

[0032] Figure 4 The DEMON spectrum of target B at different speeds; Figure 4 (a) Speed ​​5.7 knots, shaft frequency 6.2 Hz; Figure 4 (b) Speed ​​6.6 knots, shaft frequency 4.2 Hz; Figure 4 (c) Speed ​​7.2 knots, shaft frequency 3.0 Hz; Figure 4 (d) Speed ​​7.6 knots, shaft frequency 7.1 Hz;

[0033] Figure 5 The LOFAR spectrum of target A at different speeds; Figure 5 (a) Speed ​​5.7 knots, main engine single-cylinder ignition frequency 6.3 Hz; Figure 5 (b) Speed ​​6.6 knots, main engine single-cylinder ignition frequency 8.9 Hz; Figure 5 (c) Speed ​​7.2kn, main engine single cylinder ignition frequency 11.8Hz; Figure 5 (d) Speed ​​7.6 knots, main engine single-cylinder ignition frequency 14.8 Hz;

[0034] Figure 6 The LOFAR spectrum of target B at different speeds; Figure 6 (a) Speed ​​5.7 knots, main engine single-cylinder ignition frequency 6.6 Hz; Figure 6 (b) Speed ​​6.6 knots, main engine single-cylinder ignition frequency 7.5 Hz; Figure 6 (c) Speed ​​7.2 knots, main engine single-cylinder ignition frequency 8.3 Hz; Figure 6 (d) Speed ​​7.6 knots, main engine single-cylinder ignition frequency 9.1 Hz;

[0035] Figure 7 This is a schematic diagram of the ship's mechanical structure model;

[0036] Figure 8 The present invention is used to extract the gearbox reduction ratios of five ships. Detailed Implementation

[0037] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will become clearer from the following description and claims. It should be noted that the drawings are all in a very simplified form and use non-precise ratios, and are only used to facilitate and clarify the illustration of the embodiments of the present invention.

[0038] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0039] like Figure 1 As shown, embodiments of the present invention propose a method for extracting the feature of the gearbox reduction ratio of ship radiated noise, the implementation of which includes the following steps:

[0040] Step 1: Acquire ship-radiated noise signals using passive sonar, and perform frame-based preprocessing on the received signals. The specific implementation is as follows:

[0041] Step 101: Receive the ship's radiated noise time-domain signal x(n) collected by the passive sonar system:

[0042] x(n) = [x(1), x(2), ..., x(f)] s ),…,x(M)]

[0043] Among them, f s Where is the signal sampling rate, and M is the number of time-domain sampling points.

[0044] Step 102: Perform frame preprocessing on the passive sonar received signal to obtain...

[0045]

[0046] Where L is the number of signal sampling points per frame, and K is the number of signal frames. Generally, a 5-second signal is taken as one frame, and the overlapping signal duration between two adjacent frames is 1 second.

[0047] Step 2: Perform DEMON spectrum estimation on the framed signal, synthesize the DEMON spectrum time-frequency diagram of the ship's radiated noise, and extract the propeller shaft frequency value of the target ship.

[0048] according to Figure 2 The process shown is implemented as follows;

[0049] Step 201: Perform variational mode decomposition on the framed signal.

[0050] Select the (k+1)th frame signal X from the framed signal k Empirical Mode Decomposition (EMD) is performed on X to obtain an adaptive decomposition order j. k Performing j-th order variational mode decomposition (VMD) yields j intrinsic mode components (IMFs), namely...

[0051]

[0052] Step 202: Calculate the correlation coefficient of the narrowband signal for X. k The j intrinsic mode components are correlated pairwise using narrowband envelope correlation to obtain the correlation coefficient matrix:

[0053]

[0054] Among them, R j is the correlation coefficient between the j-th intrinsic mode component and other intrinsic mode components of different orders.

[0055] Step 203: For Square detection and power spectrum analysis are performed on each natural mode to obtain the envelope spectrum corresponding to the j natural mode components, and R is calculated according to the narrowband envelope correlation coefficient. kj Weighted fusion yields the DEMON spectrum P of the k-th frame signal. k ,Right now

[0056]

[0057] Step 204: Combine the DEMON spectra of the K-frame signals to obtain the DEMON spectrum time-frequency diagram P of the ship's radiated noise received signal, i.e.

[0058]

[0059] Step 205: Extract the fundamental frequency value f from the harmonic line spectrum cluster of the DEMON spectrum time-frequency diagram of the ship's radiated noise signal. z This frequency value is the ship's propeller shaft frequency.

[0060] Figure 3 The DEMON spectrum time-frequency diagram for target A. Figure 4 The DEMON spectrum time-frequency diagram for target B.

[0061] Step 3: Perform LOFAR spectrum estimation on the framed signal, synthesize the ship's radiated noise LOFAR spectrum time-frequency map, and extract the target ship's main engine ignition frequency value.

[0062] The specific implementation is as follows:

[0063] Step 301: For each frame of signal Perform power spectral density estimation (PSD) to obtain the power spectrum S of the signal for that frame. k (f), that is

[0064]

[0065] Where f = [1 / f0, 2 / f0, ..., f s f0 is the power spectrum frequency resolution.

[0066] Step 302: Synthesize the power spectrum of each frame of signal into a time-frequency power spectrum of ship radiated noise. This process is equivalent to short-time Fourier transform (STFT). Take the 0-300Hz frequency band of the time-frequency power spectrum as the LOFAR spectrum S(f) of the ship radiated noise, i.e.

[0067]

[0068] Step 303: Extract the fundamental frequency value f from the harmonic line spectrum cluster of the LOFAR spectrum time-frequency diagram of the ship's radiated noise signal. d This frequency value is the ignition frequency of a single cylinder in the ship's main engine. In the absence of a line spectrum, this frequency value can be calculated based on the frequency difference between two adjacent higher harmonic line spectra.

[0069] Figure 5 The LOFAR spectrum time-frequency diagram of target A. Figure 6 The LOFAR spectrum time-frequency diagram of target B.

[0070] Step 4: According to Figure 7 The mechanical coupling structure model shown establishes the numerical relationship between the propeller shaft frequency, the main engine ignition frequency, and the reducer gear ratio.

[0071] The specific implementation is as follows:

[0072] Step 401: Establish the single-cylinder ignition frequency f of the ship's main engine cylinder d The relationship with crankshaft speed. For an i-stroke diesel engine, for each ignition, the crankshaft rotates i / 2 revolutions, therefore f d With crankshaft speed f q satisfy

[0073]

[0074] Step 402: Introduce the gearbox reduction ratio σ to establish the ship propeller main shaft speed (shaft frequency) f. z Relationship with main engine crankshaft speed (frequency) f q .Right now

[0075] f q =σ·f z

[0076] Step 403: Provide the frequency mapping relationship between the ship's radiated noise modulation spectrum and the low-frequency line spectrum, that is, the relationship between the propeller shaft speed (shaft frequency) and the ignition frequency of a single cylinder of the main engine cylinder, and use it for gearbox reduction ratio calculation.

[0077]

[0078] Step 5: Combine steps 1 through 4 to calculate and extract the gearbox reduction ratio characteristics of the target ship; the specific implementation is as follows:

[0079] Step 501: Input the ship propeller shaft frequency extracted in Step 2 and the ship main engine single-cylinder ignition frequency extracted in Step 3 into the ship radiated noise modulation spectrum and low-frequency line spectrum frequency mapping relationship model established in Step 4, and calculate the gearbox reduction ratio.

[0080] Step 502: Using radiated noise data of the ship target at different speeds, verify the stability and separability of this feature at different speeds.

[0081] The method and feature extraction effect of the present invention will be further explained in detail below with reference to the implementation example diagram.

[0082] The data comes from five target ships in a certain waterway. Target A has a reduction ratio of 2.5:1, Target B has a reduction ratio of 4.2:1, Target C has a reduction ratio of 3.5:1, Target D has a reduction ratio of 3.0:1, and Target E has a reduction ratio of 4.5:1. The data for each ship includes at least three different speeds. One hundred samples were randomly selected from each target, and the gearbox reduction ratio features were extracted according to this invention. The results are as follows. Figure 8 As shown, the gearbox reduction ratio feature extracted by this invention exhibits good stability at different speeds, and this parameter alone is sufficient to effectively identify the aforementioned five targets.

[0083] Based on the same inventive concept, this invention also provides a ship radiated noise gearbox reduction ratio feature extraction system, comprising: an acquisition module for acquiring ship radiated noise signals via passive sonar and performing frame-by-frame preprocessing on the received signals to obtain framed signals; a DEMON spectrum estimation module for performing DEMON spectrum estimation on the framed signals, synthesizing a ship radiated noise DEMON spectrum time-frequency diagram, and extracting the target ship propeller shaft frequency value; a LOFAR spectrum estimation module for performing LOFAR spectrum estimation on the framed signals, synthesizing a ship radiated noise LOFAR spectrum time-frequency diagram, and extracting the target ship main engine ignition frequency value; a numerical relationship establishment module for establishing a numerical relationship between propeller shaft frequency, main engine ignition frequency, and gearbox reduction ratio through a mechanical coupling structure model; and a calculation module for calculating and extracting the target ship gearbox reduction ratio feature based on the numerical relationship.

[0084] Preferably, it also includes a preprocessing module for receiving the ship radiated noise time-domain signal x(n) collected by the passive sonar system: x(n)=[x(1),x(2),…,x(f s [),…,x(M)], where f s M represents the signal sampling rate, and M represents the number of time-domain sampling points.

[0085] The passive sonar received signal was preprocessed by framing to obtain:

[0086] Where L is the number of signal sampling points per frame, K is the number of signal frames, a 5-second signal is taken as one frame, and the duration of the overlapping signal between two adjacent frames is 1 second.

[0087] Based on the same inventive concept, the present invention also provides a computer device, comprising: a memory for storing a processing program; and a processor, wherein the processor, when executing the processing program, implements the ship radiated noise gearbox reduction ratio feature extraction method described in any one of the present invention.

[0088] Based on the same inventive concept, the present invention also provides a readable storage medium storing a processing program, which, when executed by a processor, implements the ship radiated noise gearbox reduction ratio feature extraction method described in any one of the claims.

[0089] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0090] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, if these changes fall within the scope of the claims of the present invention and their equivalents, they shall still fall within the protection scope of the present invention.

Claims

1. A method for extracting the reduction ratio features of a ship's radiated noise gearbox, characterized in that, Includes the following steps: The ship's radiated noise signal is collected by passive sonar, and the received signal is preprocessed by framing to obtain the framed signal. Perform DEMON spectrum estimation on the framed signal, synthesize the DEMON spectrum time-frequency diagram of the ship's radiated noise, and extract the shaft frequency value of the target ship's propeller. LOFAR spectrum estimation is performed on the framed signal to synthesize the ship's radiated noise LOFAR spectrum time-frequency map and extract the target ship's main engine ignition frequency value. The numerical relationship between propeller shaft frequency, main engine ignition frequency and gearbox reduction ratio is established by using a mechanically coupled structural model. Based on the numerical relationship, the reduction ratio characteristics of the target ship's gearbox are calculated and extracted; This includes performing DEMON spectrum estimation on the framed signals, synthesizing the DEMON spectrum time-frequency map of the ship's radiated noise, and extracting the target ship's propeller shaft frequency value, which further includes: Variational mode decomposition is performed on the framed signal; Select the (k+1)th frame signal X from the framed signal k Empirical mode decomposition is performed on X to obtain an adaptive decomposition order j. k Performing j-th order variational mode decomposition yields j intrinsic mode components: ， Calculate the correlation coefficient of the narrowband signal for X. k The j intrinsic mode components are correlated pairwise using narrowband envelope correlation to obtain the correlation coefficient matrix: ， Among them, R j Let be the correlation coefficient between the j-th intrinsic mode component and other intrinsic mode components of other orders; right Square detection and power spectrum analysis are performed on each natural mode to obtain the envelope spectrum corresponding to the j natural mode components, and R is calculated according to the narrowband envelope correlation coefficient. kj Weighted fusion yields the DEMON spectrum P of the k-th frame signal. k : ， By combining the DEMON spectra of the K-frame signals, we obtain the time-frequency diagram P of the DEMON spectrum of the ship's radiated noise received signal: ， Extract the fundamental frequency value f from the harmonic line spectrum cluster of the DEMON spectrum time-frequency diagram of the ship's radiated noise signal. z This frequency value is the ship's propeller shaft frequency.

2. The method for extracting the reduction ratio feature of ship radiated noise gearboxes according to claim 1, characterized in that, The process of acquiring ship-radiated noise signals using passive sonar and performing frame-based preprocessing on the received signals to obtain framed signals further includes: Receive the time-domain signal of ship radiated noise collected by the passive sonar system : ， in, The signal sampling rate, This represents the number of time-domain sampling points. The passive sonar received signal was preprocessed by framing to obtain: ， Where L is the number of signal sampling points per frame, K is the number of signal frames, a signal with a duration of 5 seconds is taken as one frame, and the duration of the overlapping signal between two adjacent frames is 1 second.

3. The method for extracting the reduction ratio feature of ship radiated noise gearboxes according to claim 1, characterized in that, LOFAR spectrum estimation is performed on the framed signals to synthesize the ship's radiated noise LOFAR spectrum time-frequency map, and the target ship's main engine ignition frequency value is extracted, further including: For each frame of signal Power spectral density estimation is performed to obtain the power spectrum of the signal in that frame. : ， in, , For power spectrum frequency resolution; The power spectra of each frame of signal are synthesized into a time-frequency power spectrum of ship radiated noise. This process is equivalent to short-time Fourier transform. The 0-300 Hz frequency band of the time-frequency power spectrum is taken as the LOFAR spectrum of ship radiated noise. : ， Extracting the fundamental frequency line spectrum value from the harmonic line spectrum cluster of the LOFAR spectrum time-frequency diagram of the ship's radiated noise signal. The fundamental frequency line spectrum value is the ignition frequency of a single cylinder of the ship's main engine; In the case of missing line spectrum, the fundamental frequency line spectrum value is calculated based on the frequency difference between two adjacent higher harmonic line spectra.

4. The method for extracting the reduction ratio feature of ship radiated noise gearboxes according to claim 1, characterized in that, The numerical relationship between propeller shaft frequency, main engine ignition frequency, and gearbox reduction ratio is further established through a mechanically coupled structural model, including: Establish the single-cylinder ignition frequency of ship main engine cylinders The relationship between crankshaft speed and cylinder rotation speed is as follows: For an i-stroke diesel engine, the crankshaft rotates once per ignition cycle. lock up, With crankshaft speed satisfy ; Introducing gearbox reduction ratio Establish the ship's propeller main shaft speed Relationship with engine crankshaft speed : ; The relationship between the modulation spectrum of ship radiated noise and the frequency mapping of low-frequency line spectrum, and the relationship between the propeller shaft speed and the ignition frequency of a single cylinder of the main engine are given, and used for gearbox reduction ratio calculation. .

5. The method for extracting the reduction ratio feature of ship radiated noise gearboxes according to claim 1, characterized in that, The calculation and extraction of the reduction ratio characteristics of the target ship's gearbox based on the aforementioned numerical relationship further includes: The extracted ship propeller shaft frequency and ship main engine single cylinder ignition frequency are input into the established ship radiated noise modulation spectrum and low frequency line spectrum frequency mapping relationship model to calculate the gearbox reduction ratio. Based on the radiated noise data of the ship target at different speeds, the stability and separability of the gearbox reduction ratio characteristics at different speeds are verified.

6. A system for extracting the feature of gearbox reduction ratio in ship radiated noise, characterized in that, include: The acquisition module is used to collect ship radiated noise signals through passive sonar and perform frame preprocessing on the received signals to obtain framed signals. The DEMON spectrum estimation module is used to perform DEMON spectrum estimation on the framed signal, synthesize the DEMON spectrum time-frequency diagram of the ship's radiated noise, and extract the target ship's propeller shaft frequency value. The LOFAR spectrum estimation module is used to perform LOFAR spectrum estimation on the framed signal, synthesize the ship's radiated noise LOFAR spectrum time-frequency map, and extract the target ship's main engine ignition frequency value. The numerical relationship establishment module is used to establish the numerical relationship between propeller shaft frequency, main engine ignition frequency, and gearbox reduction ratio through a mechanically coupled structural model. The calculation module is used to calculate and extract the gearbox reduction ratio characteristics of the target ship based on the numerical relationship; The DEMON spectrum estimation module performs DEMON spectrum estimation on the framed signal, synthesizes the ship's radiated noise DEMON spectrum time-frequency map, and extracts the target ship's propeller shaft frequency value, which further includes: Variational mode decomposition is performed on the framed signal; Select the (k+1)th frame signal X from the framed signal k Empirical mode decomposition is performed on X to obtain an adaptive decomposition order j. k Performing j-th order variational mode decomposition yields j intrinsic mode components: ， Calculate the correlation coefficient of the narrowband signal for X. k The j intrinsic mode components are correlated pairwise using narrowband envelope correlation to obtain the correlation coefficient matrix: ， Among them, R j Let be the correlation coefficient between the j-th intrinsic mode component and other intrinsic mode components of other orders; right Square detection and power spectrum analysis are performed on each natural mode to obtain the envelope spectrum corresponding to the j natural mode components, and R is calculated according to the narrowband envelope correlation coefficient. kj Weighted fusion yields the DEMON spectrum P of the k-th frame signal. k : ， By combining the DEMON spectra of the K-frame signals, we obtain the time-frequency diagram P of the DEMON spectrum of the ship's radiated noise received signal: ， Extract the fundamental frequency value f from the harmonic line spectrum cluster of the DEMON spectrum time-frequency diagram of the ship's radiated noise signal. z This frequency value is the ship's propeller shaft frequency.

7. The ship radiated noise gearbox reduction ratio feature extraction system according to claim 6, characterized in that, It also includes a preprocessing module for receiving the time-domain signal of ship radiated noise collected by the passive sonar system. : ， in, The signal sampling rate, This represents the number of time-domain sampling points. The passive sonar received signal was preprocessed by framing to obtain: ， in, The number of sampling points per frame of signal. The signal frame number is denoted as 5 seconds. Each frame consists of a 5-second signal and the overlapping signal duration between two adjacent frames is 1 second.

8. A computer device, characterized in that, include: The memory is used to store the processing program; A processor, which, when executing the processing program, implements the ship radiated noise gearbox reduction ratio feature extraction method as described in any one of claims 1 to 5.

9. A readable storage medium, characterized in that, The readable storage medium stores a processing program, which, when executed by a processor, implements the ship radiated noise gearbox reduction ratio feature extraction method as described in any one of claims 1 to 5.

Images