RU2546851C1 - Method of classification of hydroacoustic signals of sea object noise emission - Google Patents

Abstract

FIELD: measurement equipment.

SUBSTANCE: method of classification of hydroacoustic signals of sea object noise emission is offered which comprises the reception by the antenna of signals of sea object noise emission in additive mix with a noise by a hydroacoustic antenna, signal conversion into a digital form, spectral processing of the received signals, accumulation of the received ranges, spectrum smoothing by frequency, determination of a detection threshold proceeding from probability of false alarms and at exceeding of threshold of detection of the current spectrum at this frequency the making a decision on existence of a discrete component using which the sea object is classified in which signals of sea object noise emission in additive mix with a noise are received by two semi-antennas of hydroacoustic antenna, spectral processing of accepted signals is performed at the outputs of semi-antennas, power spectra are summarised at the outputs of two semi-antennas, with determining of total power spectrum $$S_{\sum }^2({\omega }_k),$$

a difference of power spectra $$S_{\Delta }^2({\omega }_k)$$

is found at the outputs of two semi-antennas, a differential spectrum $$\overline{S^2{({\omega }_k)}_{\sum -\Delta }}=\overline{S_{\Sigma }^2({\omega }_k)}-\overline{S_{\Delta }^2({\omega }_k)}$$

- power spectrum of sea object noise emission is determined, and the existence of discrete components is evident at excess of detection threshold by frequencies of power spectrum of sea object noise emission.

EFFECT: elimination of influence of noise spectrum received along lateral field of the characteristic of orientation of the hydroacoustic antenna and correct determination of classification spectral signs.

1 dwg

Description

The invention relates to the field of hydroacoustics and can be used in the tasks of determining the class of an object in the development of hydroacoustic systems.

In systems using classification methods for the analysis of noise emissions of targets, signs based on the characteristics of the spectral composition of the signal, the so-called portrait, are used. V.S. Burdik “Analysis of hydroacoustic systems”. L .: Shipbuilding, 1988, p. 322.

A known classification method described in the work of V.V. Deeva et al. “Information analysis by sonar operator”. L .: Shipbuilding. 1990, pp. 110-111).

The method contains the following operations:

- extraction of the parameters of the noise signal of the object from the additive mixture of signal and interference S (t) = A (t) + Y (t), where A (t) is the signal power of the object, and Y (t) is the noise power (interfering signal);

- dividing the initial implementation of the signal S (t) by r segments of duration T;

- calculation of the spectrum Y (ω _k ) for each such segment, i.e. discrete Fourier transform (FFT) of implementation on a segment of finite duration T;

- accumulation (averaging) of spectra over r realizations in - determination of the averaged estimate Y ′ (ω _k );

- averaging the spectrogram Y ′ (ω _k ) obtained at the previous stage over frequencies using a rectangular window — obtaining an average estimate of Y ”(ω _k );

- determination of the detection threshold α according to the Neumann-Pearson rule with a given probability of false detection R _l ;

- finding the ratio of the averaged estimates Y (′ ω _k ) and Y ”(ω) _k ) and comparing it with the threshold value α. Exceeding the detection threshold indicates the presence of a discrete component at a given frequency.

Information on discrete components is used in solving recognition problems (classification) as one of the main features of noise signals of various objects.

The disadvantage of this method is that the antenna has side lobes directivity characteristics. Therefore, when observing the target simultaneously with the lateral field of the directivity pattern, a noise signal from interference is received, which includes components of noise interference, sea noise and local navigation interference. The interference level will be added to the level of the detected target, which will distort the real level relationships and introduce discrete components that will distort the classification results.

The objective of the invention is to increase the likelihood of the correct classification of noise emissions of a marine object.

The technical result of the invention is to provide a reliable determination of the classification features of noise signals.

, находят разность $$S_{\Delta }^2({\omega }_k)$$

спектров мощности двух полуантенн, определяют разностный спектр $$S{({\omega }_k)}_{\sum }^2{}_{-\Delta ,}=S{({\omega }_k)}_{\sum }^2-S{({\omega }_k)}_{\Delta }^2$$

- спектр мощности шумоизлучения морского объекта, а о наличии дискретных составляющих судят при превышении порога обнаружения частотами спектра мощности шумоизлучения морского объекта.To ensure the technical result indicated, a method for classifying hydro-acoustic noise signals of a marine object, including receiving an antenna of noise signals from a marine object in an additive mixture with hydro-acoustic antenna noise, converting the signal into digital form, spectral processing of received signals, accumulating the obtained spectra, smoothing the frequency spectrum, determining detection threshold based on the probability of false alarms and when the detection threshold of the current spectrum is exceeded at a given time the frequency of deciding on the presence of the discrete component by which the marine object is classified, new features are introduced, namely, the noise signals of the marine object in the additive mixture are received with interference by two half-antennas of the hydroacoustic antenna, and the received signals are spectrally processed at the outputs of the half-antennas, determining the total power spectrum $$S_{\sum }^2({\omega }_k)$$

find the difference $$S_{\Delta }^2({\omega }_k)$$

power spectra of two semi-antennas, determine the difference spectrum $$S{({\omega }_k)}_{\sum }^2{}_{-\Delta ,}=S{({\omega }_k)}_{\sum }^2-S{({\omega }_k)}_{\Delta }^2$$

- the noise spectrum of the marine object, and the presence of discrete components is judged when the detection threshold is exceeded by the frequencies of the noise spectrum of the marine object.

The invention consists in the following.

Work in the conditions of interfering navigation for a low noise target is fraught with difficulties associated with the influence of interfering targets accepted, as a rule, along the lateral field of directional characteristics.

This process occurs simultaneously with the process of working along the main lobe of the directivity characteristic. In this case, the signal at the output of the adder contains both a signal from the target and a signal from the interfering target. The proposed method allows to separate these two signals at the output of the processing system. The difference directivity characteristic, which is formed by subtracting the signal of one half-antenna from the signal of the second half-antenna of one antenna, has a minimum in the center, and a signal in the difference characteristics will be received along the side field (V.N. Tyulin. “Theory of acoustic direction finding”. 1954, p. 35). Thus, in the difference channel there will be a signal from the interfering target received along the side lobes. It is impossible to remove the interfering signal from the sum channel, since it is additively added to the signal from the target. However, after spectral analysis, the spectrum at the output of the total channel will contain the sum of the spectra of the two targets, and the spectrum of the difference channel will contain the spectrum of the interfering target, taken along the side field when the principal zero of the difference characteristic is oriented in the direction of the first maximum of the total directivity. Therefore, if we subtract from the spectrum of the total directivity characteristics the spectrum of the differential directivity characteristics, where only the spectrum of the interfering target is located, then we can obtain the spectrum of the target that interests us.

The invention is illustrated in Fig 1, which shows a block diagram of a device that implements the method.

A device that implements the method includes a sonar antenna 1, divided into two identical semi-antennas A1 and A2, having independent outputs. The device (Fig. 1) has two serial circuits, one of which includes the A1 semi-antenna, block 2 of ADC1, block 4 of BFT1, block 5 of summation of spectra and block 8 of accumulation of total spectra.

An example of the implementation of the claimed method is described by the example of a device that implements it (Figure 1), and the second includes a series antenna A2, block 3 of the ADC2, block 5 of the BPF2, block 7 of the difference of spectra and block 9 of the accumulation of difference spectra. The outputs of blocks 8 and 9 are connected to the inputs of the block 10 determining the spectrum of the signal, the output of which is connected to the input of the block 11 for detecting discrete components (DS). The second output of block 4 is connected to the second input of block 7, and the second output of block 5 is connected to the second input of block 6. The output of block 11 is connected to the input of the classification side 12.

Blocks 2 and 3 can be performed as described in the Handbook "Digital Signal Processing" ed. Radio and Communications 1985, p. 91, blocks 4 and 5 — for example, as described in the Digital Signal Processing Handbook. Ed. Radio and Communications 1985, p. 14. Blocks 8 and 9 are described, for example, in A.A. Kharkevich "Fighting obstruction." Moscow: Nauka, 1965, pp. 70-71.

The implementation of the method, it is advisable to describe the example of the operation of the device (figure 1). Block 11 can be made as described in A.M. Tyurina, “Introduction to the Theory of Statistical Methods in Hydroacoustics”, L. 1963, pp. 127-128.

, где $$\mathrm{Re}S{(k)}_{\sum }={\mathrm{Re}}_1+{\mathrm{Re}}_2$$

, $$JmS{(k)}_{\sum }=J{m}_1+J{m}_2$$

, который поступает в блок 8 накопления суммарных спектров.The signals S ₁ (t) and S ₂ (t) from the outputs of block 1 of the semi-antenna A1 and A2 are received respectively at the input of block 2 of ADC1 and block 3 of ADC2, where ADC1 and ADC2 are analog-to-digital converters. Signals S ₁ (k) and S ₂ (k) from ADC2 and ADC3 in the form of discrete samples are respectively received in block 4 of FFT1 and block 5 of FFT2 to obtain complex spectra of half-antennas A1 and A2. Block 6 receives real (Re ₁ ) and imaginary (Jm ₁ ) samples of the implementation of the complex spectrum of the signal of the half-antenna A1 from block 4 and real (Re ₂ ) and imaginary (Jm ₂ ) samples of the implementation of the complex spectrum of the signal of the half-antenna A2 from block 5. In block 6 determines the total power spectrum of two semi-antennas: $$S_{\sum }^2({\omega }_k)={\mathrm{Re}}^{2}S{(k)}_{\sum }+J{m}^{2}S{(k)}_{\sum }$$

where $$\mathrm{Re}S{(k)}_{\sum }={\mathrm{Re}}_{\mathrm{one}}+{\mathrm{Re}}_2$$

, $$JmS{(k)}_{\sum }=J{m}_{\mathrm{one}}+J{m}_2$$

, which enters the block 8 accumulation of total spectra.

, где $$\mathrm{Re}S{(k)}_{\Delta }={\mathrm{Re}}_1-{\mathrm{Re}}_2$$

, $$JmS{(k)}_{\Delta }=J{m}_1-J{m}_2$$

, который поступает в блок 9 накопления разности спектров.Block 7 of the spectral difference receives real (Re ₁ ) and imaginary (Jm ₁ ) samples of the rth implementation of the complex spectrum of the signal of the half-antenna A1 from block 4 and real (Re2) and imaginary (Jm ₂ ) samples of the rth implementation of the complex spectrum of the signal of the half-antenna A2 from block 5. In block 7, the spectrum of the power difference of two semi-antennas is determined: $$S_{\Delta }^2({\omega }_k)={\mathrm{Re}}^{2}S{(k)}_{\Delta }+J{m}^{2}S{(k)}_{\Delta }$$

where $$\mathrm{Re}S{(k)}_{\Delta }={\mathrm{Re}}_{\mathrm{one}}-{\mathrm{Re}}_2$$

, $$JmS{(k)}_{\Delta }=J{m}_{\mathrm{one}}-J{m}_2$$

, which enters the block 9 accumulation of the difference of the spectra.

суммарных спектров мощности (А.А. Харкевич «Борьба с помехой». Москва: Наука, 1965 г, стр.70).In block 8 of the accumulation of total spectra is determined by the average (accumulated) spectrum $$\overline{S_{\Sigma }^2({\omega }_k)}$$

total power spectra (A. A. Kharkevich “Fighting interference”. Moscow: Nauka, 1965, p. 70).

разности спектров мощности.In block 9 accumulation of the difference of the spectrum is determined by the accumulated spectrum $$\overline{S_{\Sigma }^2({\omega }_k)}$$

difference of power spectra.

In block 10 determining the spectrum of the signal from block 8, the accumulated power spectrum of the total signal is received, and from block 9, the accumulated power spectrum of the difference channel is received.

The difference power spectrum (target signal spectrum) is calculated:

$$\overline{S^2{({\omega }_k)}_{\sum -\Delta }}=\overline{S_{\Sigma }^2({\omega }_k)}-\overline{S_{\Delta }^2({\omega }_k)}$$

The difference power spectrum is transmitted to the DS detection unit 11 for smoothing by a rectangular window, to generate a detection threshold based on a given probability of false alarms (A.M. Tyurin. Introduction to the theory of statistical methods in hydroacoustics. L., 1963, pp. 127-128).

All discrete components that have exceeded the threshold are transmitted to the classification unit 12 to generate classification characteristics from the signal spectrum.

Thus, the technical result, which consists in eliminating the influence of the interference spectrum received along the lateral field of the directivity characteristics, and ensuring the correct determination of the classification spectral features adopted on the main lobe of the directivity characteristics, is achieved.

Claims (1)

A method for classifying hydro-acoustic noise signals of a marine object, including receiving an antenna of noise signals from a marine object in an additive mixture with hydro-acoustic antenna interference, converting the signal into digital form, spectral processing of received signals, accumulating the obtained spectra, smoothing the spectrum in frequency, determining the detection threshold based on the probability of false alarms and when exceeding the detection threshold of the current spectrum at a given frequency, deciding on the presence of a discrete according to which the marine object is classified, characterized in that the noise signals of the marine object in the additive mixture with interference are received by two semi-antennas of the hydroacoustic antenna, spectral processing of the received signals at the outputs of the semi-antennas is carried out, the power spectra from the outputs of the two semi-antennas are summed, determining $$S_{\sum }^2({\omega }_k)$$

find the difference $$S_{\Delta }^2({\omega }_k)$$

power spectra from the outputs of two semi-antennas, accumulate and smooth in frequency the total power spectrum, accumulate and smooth the difference power spectrum, determine the noise power spectrum of the marine object as a difference spectrum $$\overline{S^2{({\omega }_k)}_{\sum -\Delta }}=\overline{S_{\Sigma }^2({\omega }_k)}-\overline{S_{\Delta }^2({\omega }_k)}$$

, determine the detection threshold, and the presence of discrete components is judged by exceeding the detection threshold by the frequencies of the noise spectrum of the marine object.