阅读说明：本技术 一种拖线阵湍流边界层噪声测试数据线谱净化方法 (Line spectrum purification method for noise test data of drag linear array turbulent boundary layer ) 是由 张海生 罗斌 邵杨梦 梁清 于 2021-06-16 设计创作，主要内容包括：本发明公开了一种拖线阵湍流边界层噪声测试数据线谱净化方法。通过滤波提取湍流边界层噪声谱的趋势,将其从原始谱中剔除,再设置合适的线谱阈值,提取线谱频率；对线谱频率进行扩展,并对原始噪声谱进行线谱净化；对净化后的湍流边界层噪声谱进行指定频段的平均,得到湍流边界层噪声测量结果。本发明可自动实现线谱的提取与剔除,在技术上消除了平台噪声中的线谱成分对湍流边界层噪声谱测量结果的不利影响,有效提升湍流边界层噪声谱测量结果的准确性。(The invention discloses a line spectrum purification method for noise test data of a drag linear array turbulent boundary layer. Extracting the trend of a turbulent flow boundary layer noise spectrum through filtering, removing the trend from an original spectrum, setting a proper line spectrum threshold value, and extracting line spectrum frequency; expanding the line spectrum frequency and performing line spectrum purification on the original noise spectrum; and averaging the noise spectrum of the purified turbulent boundary layer in the specified frequency band to obtain a noise measurement result of the turbulent boundary layer. The invention can automatically extract and eliminate the line spectrum, technically eliminates the adverse effect of line spectrum components in platform noise on the noise spectrum measurement result of the turbulent boundary layer, and effectively improves the accuracy of the noise spectrum measurement result of the turbulent boundary layer.)

1. A line spectrum purification method for noise test data of a drag linear array turbulent boundary layer is characterized by comprising the following steps: the method comprises the following steps:

step one, turbulent boundary layer noise time domain signals s (t) are collected and subjected to spectrum analysis to obtain original turbulent boundary layer noise spectrums P (F), and the frequency set is recorded as F ═ F | F_L≤f≤f_HIn which f_LAnd f_HRespectively, an upper frequency limit and a lower frequency limit;

step two, performing line spectrum extraction on the original turbulent flow boundary layer noise spectrum P (f) to obtain the center frequency f of n line spectra_ci(i is 1,2,3, …, n), and the center frequency set is denoted as C { f ═ f_c1,f_c2,…,f_cn}；

Step three, each central frequency f in the set C_ciExpanding to obtain n centersFrequency spreading subset K_i＝{f|f_ci-△f_i1≤f≤f_ci+△f_i2Where Δ f_i1And Δ f_i2Are respectively f_ciThe left frequency spreading amount and the right frequency spreading amount of the frequency domain to obtain a central frequency spreading set K ═ K₁∪K₂∪…∪K_n；

Removing the central frequency expansion set K from the frequency set F, namely calculating a difference set F-K of the central frequency expansion set K and the frequency expansion set F-K, marking the difference set as an effective frequency set X, and marking a set formed by original turbulent boundary layer noise spectrums corresponding to the frequencies in the effective frequency set as an effective turbulent boundary layer noise spectrum set Y ═ P (F) F ∈ X };

step five, calculating a turbulent boundary layer noise spectrum P' (f) avg ({ P (f) | f ∈ X and w is equal to W after line spectrum interference is eliminated_L(f)≤f≤w_H(f) H), where avg (·) represents the average of all elements in the computation set, w_L(f) And w_H(f) Respectively an upper frequency window limit and a lower frequency window limit which vary with frequency.

2. The towed linear array turbulent boundary layer noise test data line spectrum purification method of claim 1, which is characterized in that: and the line spectrum extraction method comprises the steps of firstly extracting a trend spectrum of the turbulent flow boundary layer noise spectrum through filtering, then removing the trend spectrum from the original turbulent flow boundary layer noise spectrum, and then setting a proper line spectrum threshold value to obtain the central frequency of the line spectrum.

3. The towed linear array turbulent boundary layer noise test data line spectrum purification method of claim 1, which is characterized in that: step three, the left frequency expansion quantity delta f_i1And right frequency spread Δ f_i2Equality, i.e. Δ f_i1＝△f_i2。

4. The towed linear array turbulent boundary layer noise test data line spectrum purification method of claim 1, which is characterized in that: step five, the lower limit w of the frequency window_L(f) And an upper frequency window limit w_H(f) Selecting according to an octave relation;for the center frequency f₀When 1/k octave is selected, the lower limit of the frequency window is w_L(f)＝f₀*2^-1/(2K)Upper limit of frequency window is w_H(f)＝f₀*2^1/(2K)。

Technical Field

The invention relates to the field of sonar of electronic equipment, in particular to a line spectrum purification method for noise test data of a turbulent boundary layer of a towed linear array.

Background

The towed linear array sonar has the advantages of being far away from a carrying platform, adjustable in working depth and free of limitation of the length of a base array by the space size of the platform, and is widely applied to remote and ultra-remote warning by navies of various countries. The towed self-noise is one of the main factors for restricting the detection distance of towed linear array sonar, and mainly comprises array vibration noise and turbulent flow boundary layer noise (linear array towed self-noise measurement and analysis, span, acoustic and electronic engineering, 2000 (3)). The former is that the towing cable is dragged along the motion direction with a certain inclination angle, the formed vortex causes the towing cable to shake, and the tail rope of the towing line array shakes under the action of uncertain wake flow, and after the two kinds of vibrations pass through the head and tail vibration isolation sections, partial vibration energy is still transmitted to the hydrophones in the towing line array, so that vibration noise of the towing line array is formed. The latter is due to the pulsating pressure fluctuation in the turbulent boundary layer on the outer surface of the drag line array jacket, which is directly transmitted to the hydrophone through the jacket, or through the jacket coupling excitation to generate reradiation (reviewed in several key technologies of drag line array, juliengqing, acoustic and electronic engineering, 2020 (2)).

The method can accurately measure the noise of the turbulent boundary layer of the towed linear array, and is an important reference for towed linear array sonar sheath material selection, in-array structure optimization, array forming process research and the like. In order to accurately obtain the noise of the turbulent boundary layer of the linear array, the vibration noise of the array is suppressed as much as possible in the test process.

The noise measurement system of the turbulent boundary layer of the towed linear array of the underwater counter-pulling winch is shown in figure 1 and mainly comprises two underwater winches, cables, a vibration isolation section, an acoustic section and the like. When the two underwater winches rotate clockwise as shown by arrows in the figure, the cable drives the vibration isolation section and the acoustic section to move rightwards, and the movement speed is controlled by the rotating speed of the winches. On one hand, in the process of the movement of the acoustic section, the head end and the tail end of the array section are restrained by the tensile force of the vibration isolation section, so that the jitter amplitude of the array section can be effectively reduced; on the other hand, the mechanical vibration in the operation process of the underwater winch is greatly inhibited by the vibration reduction and isolation function of the vibration isolation section. Therefore, the underwater counter-pulling winch measuring system can effectively inhibit vibration noise of the array. Under the conditions that a test water area is quiet and the electronic self-noise of an array section can be ignored, the test result of the underwater counter-pulling winch measuring system is mainly the turbulent boundary layer noise of a drag line array. However, in the operation process of the underwater winch, the reciprocating motion of the mechanical structure can excite a series of single-frequency sound waves, platform noise formed by the sound waves is diffused in water and is propagated to the acoustic section, and the measurement result of the noise of the turbulent boundary layer of the towing line array can be adversely affected.

Disclosure of Invention

Aiming at the problem that a noise measurement system of a towed linear array turbulent boundary layer of an underwater counter-pulling winch is interfered by platform noise, the invention provides a towed linear array turbulent boundary layer noise test data line spectrum purification method aiming at line spectrum components in platform noise, which can automatically realize the extraction and elimination of line spectrum, technically eliminate the adverse effect of the line spectrum components in the platform noise on the noise spectrum measurement result of the turbulent boundary layer, and effectively improve the accuracy of the noise spectrum measurement result of the turbulent boundary layer.

The object of the present invention is achieved by the following technical means. A line spectrum purification method for noise test data of a drag linear array turbulent boundary layer comprises the following steps:

step one, turbulent boundary layer noise time domain signals s (t) are collected and subjected to spectrum analysis to obtain original turbulent boundary layer noise spectrums P (F), and the frequency set is recorded as F ═ F | F_L≤f≤f_HIn which f_LAnd f_HRespectively, an upper frequency limit and a lower frequency limit;

step two, performing line spectrum extraction on the original turbulent flow boundary layer noise spectrum P (f) to obtain the center frequency f of n line spectra_ci(i is 1,2,3, …, n), and the center frequency set is denoted as C { f ═ f_c1,f_c2,…,f_cn}；

Step three, each central frequency f in the set C_ciExpanding to obtain n central frequency expansion subsets K_i＝{f|f_ci-△f_i1≤f≤f_ci+△f_i2Where Δ f_i1And Δ f_i2Are respectively f_ciThe left frequency spreading amount and the right frequency spreading amount of the frequency domain to obtain a central frequency spreading set K ═ K₁∪K₂∪…∪K_n；

Removing the central frequency expansion set K from the frequency set F, namely calculating a difference set F-K of the central frequency expansion set K and the frequency expansion set F-K, marking the difference set as an effective frequency set X, and marking a set formed by original turbulent boundary layer noise spectrums corresponding to the frequencies in the effective frequency set as an effective turbulent boundary layer noise spectrum set Y ═ P (F) F ∈ X };

step five, calculating a turbulent boundary layer noise spectrum P' (f) avg ({ P (f) | f ∈ X and w is equal to W after line spectrum interference is eliminated_L(f)≤f≤w_H(f) H), where avg (·) represents the average of all elements in the computation set, w_L(f) And w_H(f) Respectively an upper frequency window limit and a lower frequency window limit which vary with frequency.

And the line spectrum extraction method comprises the steps of firstly extracting a trend spectrum of the turbulent flow boundary layer noise spectrum through filtering, then removing the trend spectrum from the original turbulent flow boundary layer noise spectrum, and then setting a proper line spectrum threshold value to obtain the central frequency of the line spectrum.

Step three, the left frequency expansion quantity delta f_i1And right frequency spread Δ f_i2Equality, i.e. Δ f_i1＝△f_i2。

Step five, the lower limit w of the frequency window_L(f) And an upper frequency window limit w_H(f) And selecting according to the octave relation. For the center frequency f₀When 1/k octave is selected, the lower limit of the frequency window is w_L(f)＝f₀*2^-1/(2K)Upper limit of frequency window is w_H(f)＝f₀*2^{1 /(2K)}。

The invention has the beneficial effects that: the invention can automatically realize the extraction and elimination of line spectrum components in platform noise, technically eliminates the adverse effect of the line spectrum components in the platform noise on the noise spectrum measurement result of the turbulent boundary layer, and effectively improves the accuracy of the noise spectrum measurement result of the turbulent boundary layer.

Drawings

FIG. 1 is a schematic diagram of a noise measurement system for a turbulent boundary layer of a linear array of underwater counter-pulling winches, which mainly comprises two underwater winches, cables, a vibration isolation section, an acoustic section and the like.

Fig. 2 shows a turbulent boundary layer noise spectrum and its 1/3 octave average result during radio spectrum interference, where the abscissa is frequency, the ordinate is the turbulent boundary layer noise spectrum, the solid line represents the turbulent boundary layer noise spectrum with a frequency resolution of 1Hz, and the dotted line represents the 1/3 octave average result. It can be seen that the noise spectrum of the turbulent boundary layer after the 1/3 octave averaging has the same overall trend as the noise spectrum before averaging.

Fig. 3 shows the turbulent boundary layer noise spectrum with spectral interference and its 1/3 octave average result, where the abscissa is frequency, the ordinate is the turbulent boundary layer noise spectrum, the solid line represents the turbulent boundary layer noise spectrum with ambient spectral interference, and the dotted line represents the result after 1/3 octave average. It can be seen that the line spectrum gives rise to 1/3 octave averaging results in a step-wise bulge, deviating from the actual turbulent boundary layer noise spectrum.

Fig. 4 is a 1/3 octave average spectrum comparing turbulent boundary layer noise spectrum in the case of wire spectrum interference and wireless spectrum interference, where the abscissa is frequency, the ordinate is turbulent boundary layer noise spectrum, the dotted line represents the result of the turbulent boundary layer noise spectrum averaged over 1/3 octaves in the case of wireless spectrum interference, and the solid line represents the result of the turbulent boundary layer noise spectrum averaged over 1/3 octaves in the case of wire spectrum interference. It is obvious that line spectrum causes 1/3 octave averaging results to produce step-up, so that turbulent boundary layer noise spectrum of frequency bands near the line spectrum deviates from the true turbulent boundary layer noise spectrum.

FIG. 5 is a turbulent boundary layer noise spectral trend curve extraction result;

FIG. 6 is the result of the difference between the noise spectrum line spectrum of the turbulent boundary layer and the trend curve, and the line spectrum position can be effectively extracted by setting a proper threshold value;

FIG. 7 is a turbulent boundary layer noise line spectrum extraction result, the line spectrum position is marked by a square;

FIG. 8 is a turbulent boundary layer noise spectrum after removal of line spectral interference;

FIG. 9 is a comparison of the 1/3 octave average results of turbulent boundary layer noise spectrum after wireless spectrum interference, wired spectrum interference and line spectrum rejection.

Detailed Description

In order to make the objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the invention without inventive step, such as for example embodiments relating to the basic concept only with a changed use and without changing the claims, belong to the protective scope of the invention.

The invention discloses a line spectrum purification method for noise test data of a drag linear array turbulent boundary layer, which is characterized in that the trend of a noise spectrum of the turbulent boundary layer is extracted through filtering, the noise spectrum is removed from an original spectrum, a proper line spectrum threshold value is set, and line spectrum frequency is extracted; expanding the line spectrum frequency and performing line spectrum purification on the original noise spectrum; and averaging the noise spectrum of the purified turbulent boundary layer in the specified frequency band to obtain a noise measurement result of the turbulent boundary layer.

The method specifically comprises the following steps:

step one, turbulent boundary layer noise time domain signals s (t) are collected and subjected to spectrum analysis to obtain original turbulent boundary layer noise spectrums P (F), and the frequency set is recorded as F ═ F | F_L≤f≤f_HIn which f_LAnd f_HRespectively, an upper frequency limit and a lower frequency limit;

step two, performing line spectrum extraction on the original turbulent flow boundary layer noise spectrum P (f) to obtain the center frequency f of n line spectra_ci(i is 1,2,3, …, n), and the center frequency set is denoted as C { f ═ f_c1,f_c2,…,f_cn}; the method specifically comprises the following steps: firstly, a trend spectrum of a turbulent flow boundary layer noise spectrum is extracted through filtering, then the trend spectrum is removed from an original turbulent flow boundary layer noise spectrum, and a proper line spectrum threshold value is set to obtain the center frequency of the line spectrum.

Step three, each central frequency f in the set C_ciExpanding to obtain n central frequency expansion subsets K_i＝{f|f_ci-△f_i1≤f≤f_ci+△f_i2Where Δ f_i1And Δ f_i2Are respectively f_ciThe left frequency spreading amount and the right frequency spreading amount of the frequency domain to obtain a central frequency spreading set K ═ K₁∪K₂∪…∪K_n(ii) a Preferably, the left frequency spreading amount Δ f_i1And right frequency spread Δ f_i2Equality, i.e. Δ f_i1＝△f_i2。

Removing the central frequency expansion set K from the frequency set F, namely calculating a difference set F-K of the central frequency expansion set K and the frequency expansion set F-K, marking the difference set as an effective frequency set X, and marking a set formed by original turbulent boundary layer noise spectrums corresponding to the frequencies in the effective frequency set as an effective turbulent boundary layer noise spectrum set Y ═ P (F) F ∈ X };

step five, calculating a turbulent boundary layer noise spectrum P' (f) avg ({ P (f) | f ∈ X and w is equal to W after line spectrum interference is eliminated_L(f)≤f≤w_H(f) H), where avg (·) represents the average of all elements in the computation set, w_L(f) And w_H(f) Respectively an upper frequency window limit and a lower frequency window limit which vary with frequency. The lower limit w of the frequency window_L(f) And an upper frequency window limit w_H(f) For the center frequency f, selected according to an octave relationship₀When 1/k octave is selected, the lower limit of the frequency window is w_L(f)＝f₀*2^-1/(2K)Upper limit of frequency window is w_H(f)＝f₀*2^1/(2K)。

In order to eliminate the influence of random noise on the turbulent boundary layer noise measurement result, the average noise intensity in a frequency band is generally used to represent the turbulent boundary layer noise of a formulated frequency point. When the 1/3 octaves are averaged, the averaged result is shown by the dashed line in fig. 2, and the trend is the same as the turbulent boundary layer noise result in the 1Hz bandwidth, and the value is similar. Line spectrum interference in platform noise causes peaks in the turbulent boundary layer noise spectrum as shown by the solid line in fig. 3, where the result averaged by 1/3 octaves has a step-like protrusion, so that the averaged result deviates from the true value as shown in fig. 4.

The principle and process of the invention are as follows:

the trend of the turbulent flow boundary layer noise spectrum is extracted through a filtering method, the trend is removed from the turbulent flow boundary layer noise spectrum, and then a line spectrum position can be obtained through setting a proper line spectrum detection threshold value. After the noise spectrum of the turbulent boundary layer subjected to the line spectrum interference is removed, the interference of the line spectrum on the 1/3 octave average result can be eliminated.

Example (b):

the present embodiment is verified by a simulation experiment, and the initial conditions of the experiment are as follows: turbulent boundary layer noise sampling frequency is 6kHz, turbulent boundary layer noise spectrum subjected to line spectrum interference is shown as a solid line in FIG. 3, and extraction of turbulent boundary layer noise spectrum trend by using an alpha filter is shown as FIG. 5. The line spectral position shown in fig. 7 can be obtained by trending out the original turbulent boundary layer noise spectrum shown in solid lines in fig. 3 and setting the line spectral detection preset to 6 dB. Selecting left frequency spread quantity delta f_i1And right frequency spread Δ f_i2Equal,. DELTA.f_i1＝△f_i2Noise data within 10Hz to the left and right of the line spectrum is culled as shown in fig. 8. 1/3 octave averaging is carried out on the purified turbulent flow boundary layer noise spectrum, and the result is shown in fig. 9, so that the method can obviously reduce the influence of a line spectrum on the turbulent flow boundary layer noise measurement result, and the result after 1/3 octave averaging is basically overlapped with the result after noise spectrum averaging of wireless spectrum interference.

The foregoing detailed description is intended to illustrate and not limit the invention, which is intended to be within the spirit and scope of the appended claims, and any changes and modifications that fall within the true spirit and scope of the invention are intended to be covered by the following claims.