阅读说明：本技术 一种主被动联合处理的声纳目标高精度线谱提取方法 (Active and passive combined processing sonar target high-precision line spectrum extraction method ) 是由 孙微 王方勇 唐浩 査继林 谷新禹 于 2021-05-31 设计创作，主要内容包括：本发明公开了一种主被动联合处理的声纳目标高精度线谱提取方法,该方法利用主动声纳估计目标相对运动速度,映射为多普勒频偏后对被动声纳估计的目标功率谱进行修正,修正后的功率谱时间相关性和准确性将得到极大改善,再对补偿后的各时间快拍功率谱进行累积处理,利用累积后的功率谱进行特征提取,获得高精度高信噪比的线谱特征。本发明的有益效果为：该方法较常规方法提取到的线谱特征位置更为准确,不受多普勒频偏影响,同时具有更高的信噪比,大幅提高了线谱检测能力,能够对声纳目标识别性能起到改善提高作用,可应用于目前我国大部分主被动声纳设备。(The invention discloses a sonar target high-precision line spectrum extraction method based on active and passive combined processing, which comprises the steps of estimating a target relative motion speed by using an active sonar, correcting a target power spectrum estimated by a passive sonar after mapping the target relative motion speed to Doppler frequency offset, greatly improving the time correlation and accuracy of the corrected power spectrum, accumulating the compensated snapshot power spectrum at each time, and extracting features by using the accumulated power spectrum to obtain line spectrum features with high precision and high signal-to-noise ratio. The invention has the beneficial effects that: compared with the conventional method, the method has the advantages that the extracted line spectrum characteristic position is more accurate, the influence of Doppler frequency offset is avoided, meanwhile, the signal to noise ratio is higher, the line spectrum detection capability is greatly improved, the sonar target identification performance can be improved, and the method can be applied to most active and passive sonar equipment in China at present.)

1. A sonar target high-precision line spectrum extraction method based on active and passive combined processing is characterized by comprising the following steps: the method comprises the steps of estimating the relative motion speed of a target by using an active sonar, correcting the power spectrum of the target estimated by the passive sonar after mapping the target to Doppler frequency offset, accumulating the compensated snapshot power spectrum at each time, and extracting features by using the accumulated power spectrum to obtain the line spectrum features with high precision and high signal-to-noise ratio.

2. The active-passive joint processing sonar target high-precision line spectrum extraction method according to claim 1, wherein: the method specifically comprises the following steps:

the method comprises the following steps: acquiring target noise data, and performing segmentation processing on the noise data within T time to obtain N pieces of snapshot data;

X_n＝W_n*X_T n＝1,2,…N

whereinAs a function of a rectangular window, X_TInput data in a T time period;

step two: obtaining target echo data, processing each batch of echoes within T time, and obtaining target Doppler frequency offset through maximum value search and threshold screening

To obtain an estimate of the relative motion state of the object, i.e.

Step three: calculating each snapshot power spectrum, calculating each frequency Doppler frequency offset by using the relative motion state of the target, and correcting;

calculating power spectrum of each snapshot

Y_n(f)＝|FFT(X_n)|²

Calculating Doppler frequency offset of each frequency by using the following formula

Correcting each frequency to obtain a corrected power spectrum

Y_n(f)′＝Y_n(f+Δf)

Step four: accumulating the corrected power spectrums at all the moments, and extracting the characteristics of the accumulated power spectrums to obtain the characteristics of a high signal-to-noise ratio line spectrum;

for Y at different time_n(f) ' carry out accumulation processing to obtain the accumulated power spectrum

Obtaining line spectrum characteristics by matching maximum value search mode with signal-to-noise ratio threshold screening

F＝extract(Y_T(f)′)＝{f₁′,f₂′,...,f_l′}

Wherein f is_l' is the l-th line spectrum in the feature set.

Technical Field

The invention belongs to the field of acoustic signal processing target identification, and mainly relates to an active and passive joint processing sonar target high-precision line spectrum extraction method.

Background

With the continuous increase of the working distance of active and passive sonar detection equipment in China, the number of targets which can be found by sonar rises linearly, most of the targets are interference targets, so that the identification of real targets becomes extremely difficult, and the technical performance of target identification needs to be improved so as to deal with increasingly complex underwater acoustic environments. At present, target line spectrum characteristics are often used in sonar identification technology, and the principle is that for artificial underwater acoustic targets (such as submergence vehicles, naval vessels, commercial vessels and the like), the radiation noise spectrum reflects inherent characteristics of various aspects such as target structures, power, working conditions, states and the like, so that the line spectrum characteristics extracted from the radiation noise spectrum can represent target types or even target states, and the purpose of passive target identification is achieved. However, in practical situations, because the target and the sonar are mostly in a relative motion state, the generated doppler effect causes a non-negligible shift of the frequency spectrum, which results in reduced coherence, reduced line spectrum signal-to-noise ratio, and difficult feature extraction. At present, most sonar equipment acquires stable line spectrum characteristics in a long integration time processing mode, but due to the fact that the relative motion state of the ship and a target is not constant, line spectrum characteristics extracted by snapshot power spectrums at each time have certain difference, certain coherent gain is lost during accumulation, and the identification effect is poor. The PCW signal is utilized by the active sonar to realize the estimation of the relative speed of the target, so that the Doppler frequency offset of the target is estimated by utilizing the active sonar, then the target power spectrum acquired by the passive sonar is corrected, and then the line spectrum characteristic with high target precision and high signal-to-noise ratio can be obtained through integration time coherent processing, and the target identification effect is greatly improved.

Disclosure of Invention

The invention aims to provide a sonar target high-precision line spectrum extraction method based on active and passive combined processing aiming at the problem of low accuracy of underwater sound target line spectrum characteristics caused by Doppler frequency offset.

The object of the present invention is achieved by the following technical means. An active and passive combined processing sonar target high-precision line spectrum extraction method comprises the steps of estimating a target relative motion speed by using an active sonar, correcting a target power spectrum estimated by a passive sonar after mapping the target relative motion speed into Doppler frequency offset, greatly improving the time correlation and accuracy of the corrected power spectrum, accumulating the compensated power spectrum at each time, and extracting features by using the accumulated power spectrum to obtain line spectrum features with high precision and high signal-to-noise ratio.

Further, the method specifically comprises the following steps:

the method comprises the following steps: acquiring target noise data, and performing segmentation processing on the noise data within T time to obtain N pieces of snapshot data;

X_n＝W_n*X_T n＝1,2,…N

whereinAs a function of a rectangular window, X_TInput data in a T time period;

step two: obtaining target echo data, processing each batch of echoes within T time, and obtaining target Doppler frequency offset through maximum value search and threshold screening

To obtain an estimate of the relative motion state of the object, i.e.

Step three: calculating each snapshot power spectrum, calculating each frequency Doppler frequency offset by using the relative motion state of the target, and correcting;

calculating power spectrum of each snapshot

Y_n(f)＝|FFT(X_n)|²

Calculating Doppler frequency offset of each frequency by using the following formula

Correcting each frequency to obtain a corrected power spectrum

Y_n(f)′＝Y_n(f+Δf)

Step four: accumulating the corrected power spectrums at all the moments, and extracting the characteristics of the accumulated power spectrums to obtain the characteristics of a high signal-to-noise ratio line spectrum;

for Y at different time_n(f) ' carry out accumulation processing to obtain the accumulated power spectrum

Obtaining line spectrum characteristics by matching maximum value search mode with signal-to-noise ratio threshold screening

F＝extract(Y_T(f)′)＝{f₁′,f₂′,...,f_l′}

Wherein f is_l' is the l-th line spectrum in the feature set.

The invention has the beneficial effects that: compared with the conventional method, the method has the advantages that the extracted line spectrum characteristic position is more accurate, the influence of Doppler frequency offset is avoided, meanwhile, the signal to noise ratio is higher, the line spectrum detection capability is greatly improved, the sonar target identification performance can be improved, and the method can be applied to most active and passive sonar equipment in China at present.

Drawings

FIG. 1 is a functional block diagram of the present invention;

fig. 2 is a schematic diagram of the relative situation of the ship and the target.The absolute speeds of the ship and the target are respectively, and theta and phi are included angles between the ship and the target in the radial direction.

Fig. 3 to 8 are schematic diagrams of the processing results of a sea test.

Detailed Description

The invention will be described in detail below with reference to the following drawings:

as shown in FIG. 1, the process of the present invention is as follows: firstly, target noise data are obtained, noise data in T time are processed in a segmented mode, N snapshot data are obtained, and snapshot power spectrums are calculated respectively. And processing the echo data of the active target to obtain the estimation of the radial relative motion state of the ship and the target. Thirdly, the estimated radial relative motion state of the ship and the target is used for calculating the full-frequency-band frequency offset of the target, and correcting the snapshot power spectrum of the target. And fourthly, accumulating the corrected snapshot power spectrums, and extracting the characteristics of the accumulated power spectrums to obtain the line spectrum characteristics with high signal-to-noise ratio and high precision.

Referring to the relative situation of the ship and the target in FIG. 2, let the ship have an absolute speed ofTarget absolute velocity ofThe ship/target course has an included angle theta/phi relative to the radial directions of the ship/target course, transmits a PCW active pulse detection signal and has the frequency f_aSpectral frequency f of target radiation noise_bThen, then

The ship has a relative radial direction speed of

Target relative radial direction velocity of

According to the Doppler effect, the target echo frequency can be obtained when the active detection is carried out

The relation between the Doppler frequency offset and the relative speed of the ship target can be calculated by the formula (1), namely

Then

Based on the Doppler effect, when passive detection is obtained, the received target frequency is

The target doppler shift is then

It can be seen that the target Doppler frequency shift is related to the relative speed of the ship target, and the result can be obtained by substituting equation (3)

The result of the formula (6) can be used for correcting the received target frequency to obtain a target power spectrum which is not influenced by Doppler frequency offset. A power spectrum of the target at a certain time t is considered,

Y_t(f_n)＝|FFT(X_t)|² l≤n≤h (7)

true frequency position due to Doppler frequency offset

f_n′＝f_n+Δf_n (8)

The formula (6) and (8) are taken into the formula (7), and the corrected power spectrum can be obtained

The correction processing of the formula (9) is carried out on the N time snapshot power spectrums within the integral time T to obtain an aligned power spectrum Y_t(f_n'), t 1,2 … N. For Y_t(f_n') at the point of accumulationPrinciple, then

Extracting the characteristic of the accumulated power spectrum obtained by the formula (10), and obtaining the line spectrum characteristic by matching a maximum value searching mode with a signal-to-noise ratio threshold screening

F＝extract(Y_T(f)′)＝{f₁′,f₂′,...,f_l′}

Wherein f is_l' is the l-th line spectrum in the feature set.

The characteristic is corrected by Doppler frequency offset, so that the characteristic has the characteristic of high precision, meanwhile, the corrected power spectrum has better coherence, a focusing effect can be generated after accumulation, and the signal-to-noise ratio is higher.

Fig. 3 is a comparison result of power spectrums obtained by selecting 3 snapshots at different times, and it can be seen that frequency offset is generated by inherent line spectrum characteristics of a target due to different relative motion states of the ship target at different times.

Fig. 4 shows the doppler shift estimated by the echo of an active target, with which the relative motion state of the target can be calculated.

Fig. 5 is a graph of the target power spectrum before and after the correction, and it can be seen that the target power spectrum before the correction is in a curve form and the target power spectrum after the correction is in a straight form, corresponding to the 3 line spectrums in fig. 3.

Fig. 6 shows a long-time integration processing result before and after the target power spectrum is corrected, which shows that the target frequency spectrum before the correction has a shorter and thicker spectral peak due to the poorer coherence, and the target frequency spectrum after the correction has a sharp peak with a higher signal-to-noise ratio (which is respectively increased by 1-2 dB compared with that before the correction), and has better extractability. At the marked positions 1,2, 3, etc., clear line spectrum features cannot be seen before power spectrum correction, and the line spectrum features are enhanced after correction and are highlighted from the background.

Fig. 7 is a comparison between the target power spectrum history at the selected position 3 before and after correction, which proves that the target weak line spectrum feature exists at the position, and the originally dispersed snapshot line spectrum features are subjected to energy "gathering" through frequency offset correction, so that the line spectrum is highlighted.

Fig. 8 is a comparison of the results after the target power spectrum feature is extracted, and it can be seen that the extracted line spectrum feature has higher signal-to-noise ratio and accuracy after the frequency offset correction, and the line spectrum feature that cannot be extracted before the correction is extracted.

It should be understood that equivalent substitutions and changes to the technical solution and the inventive concept of the present invention should be made by those skilled in the art to the protection scope of the appended claims.