Internal wave acoustic feature extraction method based on energy jitter

Abstract

The invention discloses an internal wave acoustic feature extraction method based on energy jitter, and the method comprises the following steps: S100, internal wave region acoustic signal preprocessing: sequentially carrying out the filtering processing and noise cancellation processing of a multi-channel acoustic signal obtained by a hydrophone, and obtaining an acoustic signal time-frequency diagram through windowed short-time Fourier transform; and S200, extracting internal wave sound energy jitter characteristics: acquiring data in a specific time period based on the acoustic signal time-frequency diagram, and processing and analyzing frequency point energy to obtain an energy jitter spectrum. The method is high in accuracy, low-frequency and high-frequency signals, carrier noise and background noise are effectively suppressed by comprehensively preprocessing the multi-channel acoustic signals, and subsequently extracted internal wave acoustic features are more accurate and reliable. The method is high in timeliness, the internal wave sound energy jitter characteristics can be extracted from the time-frequency image efficiently, and the internal wave information processing speed is increased to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of ocean information acquisition, and particularly to a method for extracting internal wave acoustic features based on energy jitter. Background Technique

[0002] In the research of ocean internal wave problems, the interaction theory between the sea surface wave field and ocean internal waves and the acquisition of detectable features of internal waves formed thereby are important topics in the forefront of ocean science research. The interaction between the sea surface wave field and ocean internal waves involves complex wave-current interaction theories and wave-wave interaction theories, and the content of this theory is also directly related to the flow near the sea surface of the sea surface wind field, so it is also an important part of ocean dynamics research.

[0003] The interaction between the sea surface wave field and ocean internal waves affects the local characteristics of the sea wave spectrum. The analysis of its spectral characteristics cannot currently be fully processed by methods such as the fully developed sea wave field spectrum, and even the source term of the sea wave spectrum may need to be modified, thus affecting the quantification of remote sensing in the internal wave area. In ocean remote sensing research, for the wave breaking formed by the interaction between sea surface waves and internal waves, the currently common method is to calculate the sea wave foam coverage rate, but this processing method also has some technical difficulties that are not easy to overcome due to the uneven distribution of sea wave foam in the internal wave area.

[0004] In addition, in the related research on realizing the feature detection of internal waves, there are still difficulties, especially in the detection of internal waves by underwater platforms. Due to the restrictions of real-time and efficient underwater communication and other technical problems, although internal wave observations can be achieved through various means, such as realizing on-site observation research of internal waves by adding CTDs with appropriate positions and densities in submersible buoys, or observing the flow field in the internal wave area through ADCPs, or using optical or microwave (SAR) high-resolution satellites to widely observe ocean internal waves, etc. Although these methods have solved the internal wave monitoring problem to a certain extent, due to the restrictions of traditional underwater communication technology, it is still difficult for underwater moving platforms to obtain internal wave information in real time, and the existing observation means cannot fully and effectively achieve high timeliness of internal wave information processing. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for extracting internal wave acoustic features based on energy jitter, which is used for underwater operation platforms to detect and warn ocean internal waves in real time, and solve the problems of the prior art in terms of extracting internal wave acoustic features and the timeliness of information processing.

[0006] The present invention is implemented as follows: A method for extracting internal wave acoustic features based on energy jitter, as Figure 10 shown, includes the following steps: S100. Preprocessing of acoustic signals in the internal wave region: Filter and cancel noise for the multi-channel acoustic signals obtained by the hydrophone in sequence, and then obtain the time-frequency diagram of the acoustic signals through windowed short-time Fourier transform; S200. Extraction of the jitter characteristics of internal wave acoustic energy: Obtain the data in a specific time period based on the time-frequency diagram of the acoustic signals, process and analyze the energy at the frequency points to obtain the energy jitter spectrum.

[0007] Further, in the step S100, the filtering process is to use a FIR band-pass filter designed based on the Kaiser window to suppress the low-frequency and high-frequency signals of the multi-channel acoustic signals.

[0008] Further, the Kaiser window function is shown as the following formula: where is the window length. is an adjustable parameter that can control the shape of and can directly affect the trade-off between sidelobe attenuation and main lobe width. is the modified Bessel function of the first kind of order zero and can reflect the smooth transition characteristic of the window function. The empirical formula of the parameter is as follows The window length can be estimated by the following formula: where is the normalized transition bandwidth, is the stopband attenuation; The steps of designing a FIR band-pass filter based on the Kaiser window include: S111. Specify the filter specifications, including the stopband frequency, passband frequency, stopband attenuation, and ripple parameters; S112. Calculate : According to use the above empirical formula; S113. Determine : Calculate the required window length using the transition bandwidth formula; S114. Generate the window function: Substitute into the Kaiser window formula to generate the time-domain window; S115. Multiply the window function by the impulse response of the ideal filter.

[0009] Further, in the step 100, the noise cancellation process is to use an LMS adaptive filter for two cancellation filtrations. For the first time, use the vehicle self-noise channel as the reference channel to perform adaptive processing on the forward and lateral channel data to suppress the vehicle noise. For the second time, use the processed lateral channel signal as the reference channel to filter the forward channel to cancel the background noise.

[0010] Further, the LMS cancellation processing algorithm of the internal wave signal model is as follows: where , representing the forward and lateral channels of the hydrophone on the vehicle respectively; Denote the target signal output laterally as the reference signal, and denote the target signal output by the forward channel as the new signal to be processed, and apply the LMS cancellation processing algorithm again; Here, is the self-noise measurement value of the vehicle, is the output signal of the filter, is the forward channel signal, is the target signal after denoising the forward channel signal and denoted; is the weight vector, is the step size parameter, is the adjustment parameter; The process of filtering by the LMS adaptive filter includes the following steps: S121. Initialize the filter parameters: Before starting the adaptive filtering, initialize the filter parameters and set the initial value of the filter to 0.

[0011] S122. Read the collected multi-channel signals: including the noise signal , the forward and lateral channel signals and the signal of channel 2 .

[0012] S123. Estimate the signal and calculate the error signal in the way of weighted sum of filter coefficients, and thus adjust and update the filter parameters.

[0013] S124. Repeat the above steps, and continuously update the filter parameters through multiple iterations until the error of the filter output signal reaches the preset requirement.

[0014] Furthermore, the mathematical expression of the short-time Fourier transform is: where is the original signal, is the window function; the steps of the windowed short-time Fourier transform include: S131. Frame segmentation and windowing processing: Segment the filtered signal of the forward channel into short-time frames with a length of , and the frame shift is ; In order to reduce spectral leakage, Hamming window processing is further adopted here; the formula is ; S132. Short-time Fourier transform: Perform a fast Fourier transform (FFT) on each frame of the signal to obtain the complex spectrum ; Calculate the power spectrum and convert it to dB representation: ; S133. Time-frequency matrix splicing: Arrange the spectra of each frame in chronological order to form a time-frequency matrix, where the vertical axis is time and the time points are calculated based on the frame shift, and the horizontal axis is frequency, and the actual frequency values are calculated based on the sampling rate. . Further, in the step S200, the acquisition of data for a specific time period is to take the data from the M1th to M2th seconds in the two-dimensional acoustic energy time-frequency image after FIR filtering, where 30 ≤ M1 ≤ 90; 270 ≤ M2 ≤ 360.

[0015] Further, the processing and analysis of the frequency point energy includes the following steps: S210. Calculate the average of the energy values of all frequency points along the matrix columns to obtain an average energy sequence with respect to frequency; S220. Further process the average energy sequence with a moving average with a window size of M points and remove the trend, where M is any one of 7, 9, 11, 13, and 15; S230. According to the processed frequency energy sequence, use a parameter with a set width of threshold 10 - 20 to find the peak value and its corresponding frequency point; S240. For each frequency point, determine the energy fluctuation curve by finding the maximum value within a certain frequency range according to the fluctuation situation; S250. Filter, de-tilt, and remove the DC component from the energy fluctuation sequence to obtain a new fluctuation sequence, and then perform spectral analysis on this sequence to obtain the energy jitter spectrum after normalization.

[0016] Further, the method for removing the trend is to remove the trend term by linear fitting.

[0017] Further, when filtering the energy fluctuation sequence, a low-pass filter with an adjustable cut-off frequency is used.

[0018] Compared with the prior art, the beneficial effects of the present invention are: 1. The present invention has high accuracy: By comprehensively preprocessing multi-channel acoustic signals, low-frequency, high-frequency signals, vehicle noise, and background noise are effectively suppressed, making the subsequent extracted internal wave acoustic features more accurate and reliable, and reducing the influence of noise interference on the detection results.

[0019] 2. The present invention has strong timeliness: The present invention can efficiently extract the internal wave acoustic energy jitter features from the time-frequency image. Compared with traditional methods, it improves the speed of internal wave information processing to a certain extent, helps the underwater operation platform to obtain internal wave information in a timely manner, realizes real-time detection and early warning, and meets the requirements for timeliness in practical applications. Description of the Drawings

[0020] Figure 1is the acoustic time-frequency diagram with or without internal waves (the upper figure shows the case without internal waves, and the lower figure shows the case with internal waves); Figure 2 is the frequency response diagram of the low-frequency noise filter for test measurement data; Figure 3 are the results before and after the time-domain FIR filtering process for three channels; Figure 4 is the time-frequency domain analysis result after the FIR processing of the signal in Channel 1; Figure 5 is the internal wave noise signal model and the adaptive processing flow chart; Figure 6 is the algorithm test result of noise cancellation based on LMS; Figure 7 is the time-frequency distribution image of the filtered signal in the forward channel (when there is no internal wave); Figure 8 is the analysis result of the peak value and frequency point of the time-frequency data; Figure 9 is the analysis result of the energy jitter spectrum obtained after normalization of Channel 1 (at the frequency point 379); Figure 10 is the flow chart of the method for extracting internal wave acoustic characteristics based on energy jitter proposed by the present invention. Detailed implementation manners

[0021] In the present invention, unless otherwise clearly specified and defined, terms such as "installation", "connection", "connection", "fixation", etc. shall be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral body; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the internal communication of two components or the interaction relationship between two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0022] The following is a further description in combination with the accompanying drawings and specific embodiments: 1. Proposal of the acoustic energy change characteristics in the internal wave area 1.1 Proposal of the acoustic energy change characteristics

[0023] Internal ocean waves refer to a type of non-linear internal ocean waves, also known as internal solitary waves or internal solitary wave trains (hereinafter referred to as internal waves or internal solitary waves). Such internal waves mostly occur in specific sea areas with undulating ridges or seamounts and are excited by strong tidal currents. After generation, they are a type of fixed-form waves that propagate forward in the open sea. Such internal waves propagate relatively fast in the open sea and have stable waveform characteristics, but may experience polarity reversal at the shelf break during propagation, and disperse into multiple wave trains, and finally dissipate energy in the shallower shelf area and disappear. Research shows that due to the generally large amplitude and strong jet characteristics of such internal ocean waves underwater, they have varying degrees of impact on the safety of fixed or moving underwater operation platforms. In addition to the large amplitude and strong jet underwater, such internal waves also have a strong sea surface effect. They interact with the waves on the sea surface and have obvious dynamic and acoustic characteristics. For example, their sea surface flow field can cause wave breaking, forming characteristics similar to the disturbance of the sea surface by the wind field, and can change the optical and microwave scattering characteristics of the sea surface. Therefore, such internal ocean waves can also be observed using optical remote sensors or synthetic aperture radars for microwave imaging. In previous studies, people mostly focused on the dynamic characteristics of internal ocean waves on the sea surface or underwater, and there was less research on the specific internal wave acoustic characteristics formed by the interaction between internal waves and the sea surface wave field. The present invention focuses on the analysis and extraction technology of passive acoustic characteristics in the internal wave area caused by internal wave dynamics.

[0024] 1.2 Analysis of acoustic energy change characteristics

[0025] Based on the analysis of signals obtained by various hydrophones at different positions and of different types in the internal wave area, the present invention refers to the "jitter" phenomenon of internal wave acoustic energy presented near multiple frequency points in the acoustic energy time-frequency analysis image as the passive acoustic "energy jitter" characteristic of internal waves under the influence of internal waves. Signal analysis shows that the internal wave energy and jitter characteristics formed under conditions such as different internal wave intensities, hydrophone measurement positions, and other channel configurations are not exactly the same. Figure 1 A set of acoustic time-frequency diagrams with and without internal waves is given, where the upper sub-diagram is the acoustic time-frequency diagram without internal waves, and the lower sub-diagram is the acoustic time-frequency diagram representation when internal waves are present.

[0026] Analysis shows that when internal waves exist, the time-frequency image of their passive acoustic energy mainly has the following three manifestations: Manifestation 1: The frequency points where acoustic energy jitters occur are uncertain. This is related to the background noise and vehicle mechanical noise in the ocean environment. For example, the noise obtained near the ship's side has obvious multiple fixed-frequency noises generated during the operation of the ship's main and auxiliary engines. However, the ship itself has little impact on the noise measured by the hydrophone carried in an autonomous underwater vehicle (AUV). But we also notice that the frequency points where jitters occur will also change. Manifestation 2: The intensity of energy jitters at different frequency points is different. This feature in the time-frequency analysis diagram is related to the time-frequency processing method used on the one hand. For example, using the short-time Fourier transform or wavelet transform will be significantly different. But the fundamental reason is related to the energy of the signal itself. In fact, when transforming the acquired acoustic signal into an energy time-frequency image, it can be seen that the result obtained after transformation itself also reflects the energy of the acoustic signal (such as the Fourier transform of the signal autocorrelation). This is also one of the reasons why we call the passive acoustic feature of internal waves "energy jitter". The third manifestation: When analyzing the time-frequency image of internal wave acoustic energy jitters in different time periods, at the same moment but different frequency points, the change directions of energy jitters are consistent. Therefore, the change law of jitters over time can be obtained by analyzing this consistency. Here, this change law is called the spectrum of energy jitter - "jitter spectrum". In order to effectively obtain a specific method for describing the characteristics of internal wave energy jitters, it is necessary to preprocess the signal, such as the signal truncation length, filter parameters, window size, and normalization method, etc. These preprocessings can actually highlight the characteristics of energy jitters in the time-frequency image.

[0027] 1.3 Definition of Acoustic Energy Change Characteristics

[0028] Based on the passive acoustic signals obtained in the ocean internal wave area, the present invention proposes the concept of internal wave acoustic energy jitter characteristics for obtaining the existence of internal waves, that is, internal wave acoustic energy jitter characteristics, and calls the characteristics that change with time in the jitter characteristics "jitter spectrum". Hereinafter, we design a preprocessing process such as noise cancellation and filtering for the acoustic signal, and give a specific extraction method and technology for obtaining such ocean internal wave acoustic energy jitter characteristics. This technical method can be used for real-time detection and early warning of the existence of such internal waves in an underwater operation platform.

[0029] As Figure 10 shown, a method for extracting internal wave acoustic characteristics based on energy jitter proposed by the present invention mainly includes the following steps: S100. Preprocessing of acoustic signals in the internal wave area: Filter and cancel the noise of the multi-channel acoustic signals obtained by the hydrophone in sequence, and then obtain the time-frequency diagram of the acoustic signal through windowed short-time Fourier transform; S200. Extraction of the characteristics of internal wave acoustic energy jitter: Obtain data for a specific time period based on the time-frequency diagram of the acoustic signal, process and analyze the energy at each frequency point to obtain the energy jitter spectrum.

[0030] 2 (Step S100). Preprocessing of the acoustic signal in the internal wave area The acoustic signals of the hydrophones in the internal wave area involved in the present invention are mainly the signals of three main channels (similarly processed when there are more than three channels), namely the forward measurement channel, the lateral measurement channel, and the vehicle self-noise channel (here, the forward and lateral directions refer to the front and right directions of the vehicle). The preprocessing of the acoustic signal is to obtain a new forward measurement channel signal after being canceled by a general filter and two LMS adaptive filters and obtain the energy time-frequency image. It mainly includes the following three processing processes: First, use a general FIR band-pass filter to filter the low-frequency and high-frequency signals. This step requires processing the signals of all channels. Second, use the LMS adaptive filter for two cancellation filtering processes. The first filtering uses the measured vehicle self-noise channel as the reference channel (the first reference channel) to adaptively process the data of the forward and lateral channels, thereby suppressing the vehicle noise through cancellation filtering. The second filtering uses the processed lateral channel signal as the reference channel again to filter the forward channel, further canceling the background noise. Third, use the windowed short-time Fourier transform to convert the filtered signal into the time-frequency diagram of the acoustic signal.

[0031] 2.1 General FIR band-pass filter and filtering process

[0032] Here, windowed FIR band-pass filtering is selected. An FIR band-pass filter designed with a Kaiser window, which can provide better sidelobe suppression, is used to suppress the low-frequency and high-frequency signals. The Kaiser window function is shown as follows: where, is the window length. is an adjustable parameter that can control the shape of and can directly affect the trade-off between sidelobe attenuation and main lobe width. is the modified Bessel function of the first kind of order zero and can reflect the smooth transition characteristic of the window function. The empirical formula of the parameter is as follows The window length can be estimated by the following formula: where, is the normalized transition bandwidth (the difference between the stopband cut-off frequency and the passband cut-off frequency, divided by the sampling rate). is the stopband attenuation. This formula can ensure the minimum transition bandwidth under a given attenuation.

[0033] The steps for designing the FIR band-pass filter based on the Kaiser window include: S111. Given filter specifications, including stopband frequency, passband frequency, stopband attenuation, and ripple parameters; S112. Calculate : According to Use the above empirical formula; S113. Determine : Calculate the required window length using the transition bandwidth formula; S114. Generate a window function: Substitute into the Kaiser window formula to generate a time-domain window; S115. Multiply the window function by the ideal filter impulse response.

[0034] Table 1 gives parameters such as the filter stopband frequency, passband frequency, stopband attenuation, and passband ripple. The frequency response of this filter is as Figure 2 shown. Figure 3 Gives an example of the processing result of a certain channel signal (selecting a section with a length of 1 second, the results of the time-domain processing of three channels before and after). All measurement data was analyzed and processed using this method. From the time-frequency domain analysis results of Channel 1 ( Figure 4 ), the general FIR band-pass filter can suppress low-frequency and high-frequency signals.

[0035] Table 1 General Filter Parameter Table Serial number Parameter Parameter value Description 1 Stopband frequency 0.002 Can be used to calculate the transition bandwidth 2 Passband frequency 0.2 Can be used to calculate the transition bandwidth 3 Stopband attenuation 60 dB - 4 Passband ripple 1 dB - 2.2 Adaptive Processing of Signal Noise Cancellation for Different Channels Considering that the signals received by the hydrophones on the vehicle in the lateral and forward directions may be affected by the noise of the vehicle itself, it is necessary to use the correlation between signals and adopt LMS adaptive filtering to suppress the influence of the vehicle's own noise. At the same time, aiming at the different influences of internal wave noise on the lateral and forward directions, it is assumed that the forward channel is more affected by internal waves, while the lateral channel is less affected, and they are both affected by environmental background noise. Therefore, LMS adaptive filtering can be further used to suppress the background noise to obtain enhanced internal wave influence information. Here, a time-domain internal wave signal model and an adaptive processing flow using the LMS adaptive filtering method are established, and finally the suppressed time-domain filtered signal is obtained. The LMS cancellation processing algorithm for the internal wave signal model is as follows: Where , respectively represent the forward and lateral channels of the hydrophones on the vehicle.

[0036] Record the target signal output laterally as the reference signal, and record the target signal output from the forward channel as the new signal to be processed, and apply the LMS cancellation processing algorithm again.

[0037]

[0038] Here, is the self-noise measurement value of the vehicle, is the output signal of the filter, is the forward channel signal, is the target signal after denoising the forward channel signal and is denoted as is the weight vector, is the step size parameter, is the adjustment parameter. The processing flow is mainly divided into the following steps: S121. Initialize the filter parameters: Before starting adaptive filtering, initialize the filter parameters and set the initial value of the filter to 0.

[0039] S122. Read the collected multi-channel signals: including the noise signal , the forward and lateral channel signals and the signal of channel 2 .

[0040] S123. Estimate the signal and calculate the error signal in the way of weighted sum of filter coefficients, and thus adjust and update the filter parameters.

[0041] S124. Repeat the above steps, and continuously update the filter parameters through multiple iterations until the error of the filter output signal reaches the preset requirement.

[0042] Figure 6 is the internal wave signal model diagram. Figure 7 are the processing results of the three-channel measurement signals implemented by this method. For the sake of intuition, the processing process of 2.3 is used. Among them, (a) is the processing result of channel 1 canceling channel 3; (b) is the processing result of channel 1 canceling channel 3; (c) is the direct processing result of channel 3; (d) is the result after channel 1 is processed and then cancels the processing result of channel 2, and it can be intuitively seen that there are obvious jitter characteristics at frequencies of 1.5 kHz and 3.8 kHz.

[0043] 2.3 Acoustic energy time-frequency processing based on short-time Fourier transform Time-frequency analysis is one of the important tools for processing acoustic signals. It can transform the time-domain signal that changes with time into a two-dimensional signal on the frequency and time axes, and can visually analyze the frequency change characteristics of the signal at different times in the frequency domain. The signal change characteristics at different frequencies also reflect the change of signal energy. Here, on the basis of the above forward-channel filtering process, a general short-time Fourier transform (STFT) processing process is further adopted, and finally the time-frequency distribution image of the internal wave field energy showing the change of signal energy on the frequency and time axes is obtained. During the processing, the long-time non-stationary signal forward-channel filtering signal is segmented into short-time stationary signal segments (windowing and framing) through the short-time Fourier transform, and the Fourier transform is performed segment by segment to obtain the joint distribution map of the signal in time and frequency - the time-frequency distribution image. The mathematical expression of the short-time Fourier transform is: where is the original signal, is the window function. The processing steps are as follows: S131. Framing and windowing processing: The forward-channel filtering signal is segmented into short-time frames of length , and the frame shift is ; To reduce spectral leakage, Hamming window processing is further adopted here; The formula is

[0044] S132. Short-time Fourier transform: Perform a fast Fourier transform (FFT) on each frame of the signal to obtain the complex spectrum ; Calculate the power spectrum and convert it to dB representation: ; S133. Time-frequency matrix splicing: Arrange the spectra of each frame in time order to form a time-frequency matrix. The vertical axis is time and the time points are calculated according to the frame shift, and the horizontal axis is frequency, and the actual frequency values are calculated according to the sampling rate Figure 7 Figure shows the time-frequency distribution image after the forward channel is filtered by FIR and LMS. This result is the case without internal waves, and it is not obvious to visually see the obvious characteristics of jitter.

[0045] 3 (Step S200). Extraction of the jitter characteristics of internal wave sound energy

[0046] Compared with the received signal before the internal wave arrives, the breaking sound formed by the interaction between the internal wave and the sea surface wave significantly increases the sound level in the area. However, this increase in noise level cannot be used as an acoustic feature of the internal wave. If the time-frequency diagrams of the acoustic energy before and when the internal wave arrives are compared, it is not difficult to find that the energy of the original frequency at multiple frequency points in the diagram shows obvious left-right shifts or fluctuations, that is, the time-frequency diagram when the internal wave arrives shows the phenomenon of "jitter" of energy near a certain frequency point over time. To extract the characteristics of the "jitter" phenomenon of the signal in the time-frequency diagram, the time dimension and frequency dimension are separated and spectral analysis techniques are used to obtain the "jitter spectrum" diagram of the internal wave acoustic energy at multiple frequency points. The processing steps are as follows: S210: Read the time-frequency energy data Based on the two-dimensional acoustic energy time-frequency image obtained after FIR filtering, the data from the 60th to 300th seconds is taken as the input for subsequent processing (actually a matrix). The purpose of selecting the data sequence from the 60th to 300th seconds is to exclude the situation of unstable convergence in the initial about 60-second sequence due to adaptive filtering.

[0047] S220: Energy sequence signal of frequency The energy values at all frequency points are averaged along the matrix columns to obtain the average energy sequence regarding frequency; this energy sequence is further processed with a moving average with a window size of 11 points and the trend is removed.

[0048] S230: Calculate the peak value and frequency point of the time-frequency data According to the processed frequency energy sequence, using the parameter with a set width of the threshold of 10 - 20, the peak value and its corresponding frequency point are obtained; S240: Analyze the fluctuation of the energy at each frequency point over time For each frequency point, according to the fluctuation situation, by finding the maximum value within a certain frequency range, the energy fluctuation curve is determined.

[0049] S250: Analyze the fluctuation of the acoustic energy at the main frequency points The energy fluctuation sequence is filtered, detrended, and de-dc processed to obtain a new fluctuation sequence, and then spectral analysis is performed on this sequence. After normalization, the energy jitter spectrum is obtained - that is, the spectral characteristics of the acoustic energy jitter caused by the internal wave.

[0050] Figure 8 is the analysis result of the peak value and frequency point of the time-frequency data; Figure 9 is the analysis result of the energy jitter spectrum obtained after normalization. This result can be used as the main basis for the presence or absence of internal waves.

[0051] In summary, the present invention proposes the acoustic energy change characteristics caused by the interaction between the local ocean internal wave field and the sea surface wave field - the acoustic energy jitter characteristics of internal waves, and presents the extraction techniques and methods for such characteristics. It includes the proposal of the acoustic energy change characteristics, that is, the energy jitter characteristics of internal waves - spectral characteristics; the realization of filtering processing for specific channels for the three-channel measurement signals; the extraction methods and techniques for such energy change characteristics based on the acoustic energy time-frequency image, etc.

[0052] The present invention has high accuracy: By comprehensively preprocessing the multi-channel acoustic signals, the low-frequency, high-frequency signals, vehicle noise, and background noise are effectively suppressed, making the subsequently extracted internal wave acoustic characteristics more accurate and reliable, and reducing the influence of noise interference on the detection results.

[0053] The present invention has strong timeliness: The present invention can relatively efficiently extract the internal wave acoustic energy jitter characteristics from the time-frequency image. Compared with the traditional methods, it improves the speed of internal wave information processing to a certain extent, helps the underwater operation platform to obtain internal wave information in a timely manner, realizes real-time detection and early warning, and meets the requirements for timeliness in practical applications.

[0054] The above is only the preferred embodiment of the present invention and is not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for extracting the acoustic characteristics of internal waves based on energy jitter, characterized in that, It includes the following steps: S100. Preprocessing of acoustic signals in the internal wave zone: Filter and cancel noise for the multi-channel acoustic signals obtained by the hydrophone in sequence, and then obtain the time-frequency diagram of the acoustic signals through windowed short-time Fourier transform; the filtering process is to use an FIR band-pass filter designed based on the Kaiser window to suppress low-frequency and high-frequency signals of the multi-channel acoustic signals; S200. Extraction of the jitter characteristics of internal wave acoustic energy: Obtain data for a specific time period based on the time-frequency diagram of the acoustic signals, process and analyze the energy of frequency points to obtain the energy jitter spectrum.

2. The extraction method of internal wave acoustic characteristics based on energy jitter according to claim 1, wherein The Kaiser window function is shown as follows: where is the window length, is an adjustable parameter that can control the shape of, and can directly affect the trade-off between sidelobe attenuation and main lobe width, is the modified Bessel function of the first kind of order zero and can reflect the smooth transition characteristics of the window function. The empirical formula for the parameter is as follows The window length can be estimated by the following formula: where is the normalized transition bandwidth, is the stopband attenuation; The steps for designing the FIR band-pass filter based on the Kaiser window include: S111. Specify the filter specifications, including stopband frequency, passband frequency, stopband attenuation, and ripple parameters; S112. Calculate : According to Use the above empirical formula; S113. Determine : Calculate the required window length using the transition bandwidth formula; S114. Generate the window function: Substitute into the Kaiser window formula to generate the time-domain window; S115. Multiply the window function by the impulse response of the ideal filter.

3. The extraction method of internal wave acoustic characteristics based on energy jitter according to claim 2, characterized in that In step 100, the noise cancellation process is to perform two cancellation filters using an LMS adaptive filter. The first time, the self-noise channel of the vehicle is used as the reference channel to perform adaptive processing on the forward and lateral channel data to suppress vehicle noise. The second time, the processed lateral channel signal is used as the reference channel to filter the forward channel to cancel background noise.

4. The extraction method of internal wave acoustic characteristics based on energy jitter according to claim 3, characterized in that, The LMS cancellation processing algorithm for the internal wave signal model is as follows: where , respectively represent the forward and lateral channels of the hydrophone on the vehicle; The target signal output laterally is denoted as the reference signal, and the target signal output from the forward channel is denoted as the new signal to be processed, and the LMS cancellation processing algorithm is applied again; Here, is the measured value of the vehicle's self-noise, is the output signal of the filter, is the forward channel signal, is the target signal after denoising the forward channel signal and is denoted; is the weight vector, is the step size parameter, is the adjustment parameter; The process of filtering by the LMS adaptive filter includes the following steps: S121. Initialize the filter parameters: Before starting the adaptive filtering, initialize the filter parameters and set the initial value of the filter to 0; S122. Read the multi-channel signals collected: including noise signals , forward and lateral channel signals and the signal of channel 2 ; S123. Estimate the signal and calculate the error signal in the way of weighted sum of filter coefficients, and then adjust and update the filter parameters accordingly; S124. Repeat the above steps, and continuously update the filter parameters through multiple iterations until the error of the filter output signal reaches the preset requirement.

5. The extraction method of internal wave acoustic characteristics based on energy jitter according to claim 1, characterized in that The mathematical expression of the short-time Fourier transform is as follows: where is the original signal, is the window function; The steps of the windowed short-time Fourier transform include: S131. Frame segmentation and windowing processing: Segment the forward channel filtering signal into short-time frames with a length of , and the frame shift is ; To reduce spectral leakage, Hamming window processing is further adopted here; the formula is ; S132, Short-time Fourier transform: Perform a fast Fourier transform (FFT) on each frame of the signal to obtain a complex spectrum ; Calculate the power spectrum and convert it to dB representation: ; S133. Time-frequency matrix splicing: Arrange the spectra of each frame in chronological order to form a time-frequency matrix. The vertical axis represents time and the time points are calculated based on the frame shift. The horizontal axis represents frequency and the actual frequency values are calculated based on the sampling rate. .

6. The extraction method of internal wave acoustic characteristics based on energy jitter according to claim 1, characterized in that In step S200, the acquisition of data for a specific time period is to take the data from the M1th to M2th seconds in the two-dimensional acoustic energy time-frequency image after FIR filtering, where 30 ≤ M1 ≤ 90; 270 ≤ M2 ≤ 360.

7. The extraction method of internal wave acoustic characteristics based on energy jitter according to claim 1, characterized in that The processing and analysis of the energy of frequency points include the following steps: S210. Calculate the average of the energy values of all frequency points along the matrix columns to obtain the average energy sequence with respect to frequency; S220. Further process the average energy sequence with a moving average with a moving window size of M points and remove the trend, where M is any one of 7, 9, 11, 13, and 15; S230. According to the processed frequency energy sequence, use a parameter with a set width of threshold 10 - 20 to find the peak value and its corresponding frequency point; S240. For each frequency point, determine the energy fluctuation curve by finding the maximum value within a certain frequency range according to the fluctuation situation; S250. Filter, de-trend, and remove the DC component from the energy fluctuation sequence to obtain a new fluctuation sequence, and then perform spectral analysis on this sequence and normalize it to obtain the energy jitter spectrum.

8. A method for extracting the acoustic characteristics of internal waves based on energy jitter according to claim 7, characterized in that, The method for removing the trend is to remove the trend term by linear fitting.

9. The extraction method of internal wave acoustic characteristics based on energy jitter according to claim 7, characterized in that When filtering the energy fluctuation sequence, a low-pass filter with an adjustable cut-off frequency is used.

Images