RU2339050C1 - Method of sea noisy objects detection - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: method includes as follows. Horizontal and vertical orientation characteristic static fan receives noise signals in combination with frequency-time processing within each spatial observation channel, quadrating, time averaging, alignment and signal normalising to interference, observation of current view cycle for received normalised signals and detection decision-making comparing to limit value of signal-interference relation. Thus within each view cycle for each frequency sample the adaptive spatial observation channels are formed, at least by three adjacent spatial channels in horizontal or vertical plane.

EFFECT: improved reliability of detection and long contact with noisy moving target in the sea.

2 dwg

Description

The invention relates to the field of hydroacoustics and can be used in noise direction finding systems.

A known method for detecting spatio-temporal noise signals based on a fully multi-channel spatial adaptive reception of noise signals and adaptive noise suppression using the estimation of acoustic noise correlation coefficients, see V.A. Lazutkin. Statistical methods for processing sonar signals. Kiev: Naukova Dumka. 1987. Ch. 1, p. 2, 3; V.N. Fomin. Recurrent assessment and adaptive filtering. M .: Nauka, 1984. p. 87-88; “Underwater Acoustics and Signal Processing”, ed. L. Bjorne, M .: Mir, 1985, p. 300-301 and p. 321-323; A.M. Vural, Effects of perturbations on the performance of optimum / adaptive array // IEEE Transactions, 1979, vol. AES-15, # 1, p. 76-87. One of the main disadvantages of this method is that in some cases, the high cost can impede the use of a fully adaptive system or there is simply no need for such characteristics that it can provide.

A simplification of this method for detecting and reducing cost is known (due to the decrease and degree of optimality). The method is simplified by using the analysis of the position of the wave front (intense directed noise interference) and processing either in the space of elements of a discrete hydroacoustic antenna or in the space of rays (spatial channels) in the so-called systems with a partially specified structure, see the cited book, ed. L. Bjorne, M .: Mir, 1985, p. 284-286, as well as A.B.Beggeroyer. Signal Processing in Sonar // Application of Digital Signal Processing, ed. E. Oppenheim. M .: World. 1980, p. 6.4; patent 3763490 US, 1973, Adaptive beamformer and signal processor for sonar.

This analysis is based on the representation of the signal field using plane waves, see Yu.G. Sosulin, Yu.N. Parshin. Estimation-correlation-compensation algorithms for detecting multidimensional signals. Radio engineering and electronics. 1981. Issue 26. No. 8. S.1635-1643, and also: R.R. Ramseyer, S.D. Mororge. A distributed microprocessor architecture far fixed and mobile acoustic array adaptive beamforming. IEEE Journal of oceanic engineering. 1979. Vol.OE-4, # 2. p. 46-51; D.J. Chapman. Partial adapting for the large array // IEEE transaction on AP, 1976, vol.24, # 5, p.685-696; D.A. Gray. Formulation of the maximum signal-to-noise ratio array processor in beam space // JASA, 1982, v.72, # 4, p.1195-1201.

All these methods have disadvantages associated with the operating conditions of the detection system with a partially specified structure. In adverse conditions, which are determined by the characteristics of the interference and the presence of intense directional sources of interference, environmental acoustics (profile of the speed of sound, depth and inclination of the bottom, etc.), their effectiveness can deteriorate sharply. The disadvantage of these methods is the need for either the presence of several space-oriented receiving channels for tracking along the direction of the wave fronts of noise sources, or the presence of multi-channel correlators of noise signals of the entire discrete aperture of a hydroacoustic antenna. The consequence of these factors is the high cost of implementing the detection method and the relatively low stability and reliability of detection.

The closest in technical essence to the claimed method is the detection method set forth in RF application No. 2005113369/09 (015399) dated 05/03/2005 (decision of 09/21/2006 on the grant of a patent). A noise signal is received (in fact, it means receiving a mixture of a noise signal and an interference) by an antenna, which is supposed to be sharply directed in the vertical and horizontal plane.

The prototype method implements the operation of receiving noise signals in the horizontal and vertical plane by a multi-element array of sonar and carry out the initial processing, for which:

the voltage of the noise signals of the antenna array is converted into digital form, the Fourier transform of the samples of the voltages of the noise signals of the antenna array is performed, the amplitude and phase coefficients of the in-phase summation of the voltages of the signals of the antenna array are calculated for each of the received frequency samples, the output voltages of the signals of the antenna array with constant weights equal to the product amplitude and phase coefficients, which form the spatial observation channels in the horizontal and vertical plane yy, carry out the optimized time-frequency processing of the received noise signals for each spatial observation channel in the horizontal plane;

after that, they square and carry out secondary processing on each review cycle, for which:

summarize over all frequency samples the output voltages of the formed spatial channels in a fixed frequency range, averaged over time, center and normalize the noise signals to noise, observe at each review cycle the received marks of the received noise signals, and decide whether to detect the signal ratio by comparison with the threshold value interference.

The optimized time-frequency processing is that:

receive a noise signal with a static vertical fan simultaneously in several directions of the vertical plane of each spatial observation channel as part of a static fan in the horizontal plane, optimize the reception of each horizontal spatial channel by selecting the most probable vertical angles of reception for the existing sonar conditions for underwater observation for which:

they process the received noise signals with weights proportional to the calculated signal-to-noise ratio in the vertical spatial channels before accumulating them in successive review cycles and summarize the accepted noise-normalized vertical spatial channel noise signals with the calculated weights.

The method works well with isotropic noise interference and with anisotropic noise of the sea, under conditions where the contribution of intense interference from directed sources coming from other directions in the horizontal plane can be neglected. Moreover, the highest detection efficiency of noise signals is achieved by maximizing the concentration coefficient and noise immunity coefficient in the direction of receiving the noise signal. Hereinafter, the term “noise immunity coefficient” used is generalizing with respect to the classical term “concentration coefficient” (see the Reference book on hydroacoustics, L .: Shipbuilding, 1988, p.305, 308-309). The noise immunity of the antenna in the far anisotropic interference field can be determined in this case through the spatial spectrum of the distributed noise field.

The disadvantage of this method is that this method does not allow for the simultaneous maximization of these coefficients. The disadvantages of the prototype include low selectivity to interference from directional sources due to limitations in the dimensions of the antenna on the media and the presence of fluctuations in the wave fronts of the sources in a large spatial interval. In addition, there is no provision for selection of directional noise interference, focused on obtaining high resolution. The weighted summation used in the prototype leads to success in conditions of relatively weak directed noise interference. The use of processing with constant weighting factors is problematic when the signal approaches the angle with directed noise interference or when exposed to intense noise interference, which leads to the loss of the signal for a long time.

The objective of the invention is: to increase the selectivity to interference of directional sources in the presence of these restrictions, to ensure the selection of directional noise interference, focused on obtaining high resolution, i.e. the creation of a method for detecting noisy objects, which would at the same time make it possible to more accurately determine the presence of a target noise signal and maintain acoustic contact with the target with greater reliability than in the prototype method, reducing the time for masking noise and signal loss with loss of acoustic contact.

The technical result of the proposed method is to increase the reliability of detection and long-term maintenance of the target’s contact by taking into account the hydroacoustic conditions for observing noisy objects and more complete selection of noise signals in an additive mixture of directed noise interference along an angle in a vertical or horizontal plane.

To ensure the specified technical result, a method for detecting objects that are noisy at sea in a fixed frequency range, in which noise signals are received in the horizontal and vertical planes by a multi-element antenna array of a sonar and carry out initial processing, for which:

convert into digital form the voltage of the noise signals of the antenna array, perform the Fourier transform of the samples of the voltage of the noise signals of the antenna array,

calculate for each of the obtained frequency samples the amplitude and phase coefficients of the in-phase addition of the voltage of the signals of the antenna array,

summarize the output voltages of the signals of the antenna array with constant weights equal to the product of the amplitude and phase coefficients, which form the spatial observation channels in the horizontal and vertical planes,

they quadrate and carry out secondary processing on each review cycle, for which:

summarize over all frequency samples the output voltages of the formed spatial channels in a fixed frequency range,

average over time, center and normalize noise signals to interference,

monitor on each review cycle the received marks of the received noise signals and decide on detection by comparing with the threshold value of the signal-to-noise ratio,

new features have been introduced, namely, on each review cycle before squaring, adaptive spatial observation channels are formed for each frequency reference, each of which is formed by at least three adjacent spatial channels in the horizontal or vertical planes, for which:

form mutual power spectra between the noise signals of the spatial channels involved in the formation of the adaptive spatial observation channel,

accumulate mutual power spectra of noise signals for a given accumulation time,

compose a matrix of accumulated mutual power spectra of noise signals and perform orthogonal transformation of the matrix,

calculate the phase coefficient vectors of the in-phase addition of the signals of the spatial channels involved in the formation of the adaptive spatial observation channel,

calculate the vector of the output voltages of the spatial channels involved in the formation of the adaptive spatial channel by solving the vector-matrix equation written for the orthogonally transformed matrix of accumulated mutual power spectra and the phase coefficient vector of the in-phase addition,

the squaring is performed by calculating the response of the resulting adaptive spatial observation channel, which is equal to the reciprocal equal to the sum of the squared elements of the output voltage vector,

and secondary processing is carried out for the output voltages of the adaptive spatial observation channels.

It is known that the introduction of weight processing of noise signals using an adaptive algorithm allows to increase the selectivity to interference of directional sources in the presence of the above limitations, to ensure the selection of directional noise interference, focused on obtaining high resolution. However, the adaptation made according to the full scheme, that is, with a large number of spatial channels of a multi-element sonar antenna array, greatly complicates the procedure and increases the computational time of spatial - frequency - time processing with adaptation.

The proposed method for detecting noise signals, due to the fact that the correlation matrix of noise signals can be reduced to the size of the signal space, allows the processing of information received from adjacent spatial channels in a reduced amount using adaptive devices of a simplified structure.

The present invention is illustrated by graphical images, which show: in Fig.1 is a block diagram of a device that implements the inventive method, Fig.2 is a block diagram of a detection method as a sequence of operations.

The detection method is implemented by a device - a noise-finding station with a spatial signal processing system - UFN (see "Hydroacoustic means ...", Karyakin Yu.A., Smirnov S.A., Yakovlev G.N., 2005, p. 173 , fig. 2.5).

The device for detecting noisy objects in the sea of figure 1 consists of a multi-element, for example, a cylindrical sonar antenna 1. The antenna elements are connected to the preliminary processing device, then to the spatial processing systems of block 2, then the primary information processing 5.1 ... 5.A and the system secondary processing of information 11.

The proposed method is carried out using a receiving system as follows.

до

Noise signals are received by a multi-element sonar antenna 1 in the horizontal and vertical plane. In blocks 2-4 (operations 16-18, figure 2) form a static horizontal and vertical fan, respectively, M and N directional characteristics. Noise signals are received by each of the A = M * N spatial channels that are obtained for the frequency f _k (where f _k = kΔf, Δf is the preselected frequency step as a result of performing, in block 2, operation 17 of the Fourier transform of the noise signal voltage samples). The integers k are in the range from k _n to k _in , with a fixed frequency range located in the range from

before

Next, in blocks 6 (6.1 ... 6.A), A is formed (operation 19, Fig. 2) of adaptive spatial observation channels, each of which is formed by at least three adjacent spatial channels in horizontal Q = 3 or in vertical plane R = 3. In this case, the outputs of adjacent spatial channels are used, spaced at an angle in the horizontal or vertical viewing plane, at least by the width of the directivity characteristic for the k-th frequency component. According to the results of the simulation of the proposed method, it is advisable to select channels with this diversity in angle, since with greater diversity, the noise immunity of the detector decreases, and with a smaller diversity, the stability of the adaptive algorithm decreases, false signals arise.

шумовых сигналов пространственных каналов, участвующих в формировании адаптивного пространственного канала наблюдения. При этом составляют матрицу Ф^ПК размера соответственно Q×Q или R×R накопленных взаимных спектров мощности

шумовых сигналов (операция 20, фиг.2).Then, in blocks 7 (7.1 ... 7.A), in each of A adaptive spatial channels, mutual power spectra are formed and accumulated for a given time

noise signals of spatial channels involved in the formation of an adaptive spatial observation channel. In this case, a matrix F of ^{PCs of} size Q × Q or R × R of the accumulated mutual power spectra is respectively

noise signals (operation 20, figure 2).

и

в матричном видеIn blocks 8 (8.1 ... 8.A), an orthogonal transformation of the matrix Φ ^{PK is} performed (operation 21) using the triangular factorization procedure

and

in matrix form

- нижняя и верхняя треугольные матрицы с элементами, вычисленными по алгоритму квадратного корня (см. например, в книге Б.П.Демидовича и И.А.Марона "Основы вычислительной математики", М., Гос. изд. физ. - мат. л-ры, 1963, стр.287-288).Where

- lower and upper triangular matrices with elements calculated by the square root algorithm (see, for example, in the book by B. P. Demidovich and I. A. Maron, "Fundamentals of Computational Mathematics", M., State Publishing House of Physics and Mathematics. L-ry, 1963, pp. 287-288).

In blocks 15 (15.1 ... 15.A), the phase coefficients P _q (φ ₀ , θ ₀ , k) or P _r (φ ₀ , θ ₀ , k) are calculated and a vector with elements P _q (φ ₀ , θ ₀ , k) or P _r (φ ₀ , θ ₀ , k) in-phase addition of signals in a plane wave for phase centers Q or R of spatial channels involved in the formation of an adaptive spatial observation channel. An obvious implementation option is also possible in which the element P _q (φ ₀ , θ ₀ , k) and P _r (φ ₀ , θ ₀ , k) are calculated, for example, as the directivity of the antenna array for a plane wave (operation 22). The mentioned calculations can be carried out according to the algorithms given, for example, in the book by Matvienko V.N., Tarasyuk Yu.F. "Range of action of hydroacoustic means", L., Shipbuilding, 1981, pp. 212-214. The set of operations 22 in blocks 15 (15.1 ... 15.A) is implemented by preliminary calculation and storing phase coefficients. The calculated values of the phase coefficients P _q (φ ₀ , θ ₀ , k) and P _r (φ ₀ , θ ₀ , k) are recorded in the long-term (permanent) memory of the storage device.

Operations 22 in blocks 15 (15.1 ... 15.A) are carried out independently of other operations and provide data for computational operations 23.

или

образующих соответственно вектор

или

выходных напряжений пространственных каналов, участвующих в формировании адаптивного пространственного канала (операция 23). Вычисление осуществляют по формуле в векторно-матричном виде

где соответственно Р={Р_q(φ₀,θ₀,k)} или Р={P_r(φ₀,θ₀,k} - вектор-столбец фазовых коэффициентов,

- нижняя треугольная матрица ортогонального разложения матрицы взаимных спектров мощности Ф^ПК. Упомянутые расчеты могут быть проведены по алгоритмам, приведенным, например, в упомянутой книге Б.П.Демидовича и И.А.Марона, глава VIII, §2.In blocks 9 (9.1 ... 9.A), the set of Q or R spectral responses is calculated at the k-th frequency

or

forming respectively a vector

or

the output voltages of the spatial channels involved in the formation of the adaptive spatial channel (operation 23). The calculation is carried out according to the formula in a vector-matrix form

where, respectively, P = {P _q (φ ₀ , θ ₀ , k)} or P = {P _r (φ ₀ , θ ₀ , k} is the column vector of phase coefficients,

- the lower triangular matrix of the orthogonal decomposition of the matrix of mutual power spectra Φ ^PC . The above calculations can be carried out according to the algorithms given, for example, in the mentioned book of B.P. Demidovich and I.A. Maron, chapter VIII, §2.

в горизонтальной или вертикальной плоскости (операция 24), равный обратной величине суммы квадратов элементов вектора выходных напряжений, для горизонтальной плоскости по формулеIn blocks 10 (10.1 ... 10.A), for the k-th frequency component, the response of each of A adaptive spatial observation channels is calculated

in the horizontal or vertical plane (operation 24), equal to the reciprocal of the sum of the squares of the elements of the output voltage vector, for the horizontal plane according to the formula

or for the vertical plane according to the formula

усредняют по времени, центрируют и нормируют, производят наблюдение сигналов на каждом цикле обзора (операции 25 и 26).In block 12, the responses of each of A adaptive spatial observation channels in a fixed frequency range are summed over all frequency samples

averaged over time, center and normalize, observe the signals on each review cycle (operations 25 and 26).

Make a decision about detecting the signal in block 13 when the threshold value is exceeded by the signal-to-noise ratio in the spatial channel (operation 27); registration, deployment on the panoramic indicator of signal marks at each review cycle is implemented in block 14 (operation 28).

The simulation results of the proposed method showed that the use of amplitude-phase distribution control by the adaptive algorithm of adjacent spatial channels, which already have high spatial selectivity to distributed interference, provide greater noise immunity than the use of amplitude-phase distribution control of weakly directed receivers of the antenna array.

This made it possible to detect noisy objects with greater reliability than in the prototype method, earlier to determine the presence of the target signal and to maintain acoustic contact with the target for a long time, reducing the time for masking the noise and the time the signal disappears with the loss of acoustic contact. At the same time, the concentration coefficient of the hydroacoustic antenna remains high.

Along with the aforementioned modeling of the inventive object with simulation of signals and interference, processing of real signals and interference records made in the field was carried out, which confirmed the obtained simulation results.

Claims (1)

A method for detecting objects that are noisy at sea in a fixed frequency range, in which noise signals are received in horizontal and vertical planes by a multi-element sonar array and carry out initial processing, which is converted into digital form by the voltage of the noise signals of the antenna array, and the Fourier transform of the samples of voltage of the noise signals by the antenna gratings, the amplitude and phase coefficients of the common-mode voltage addition are calculated for each of the obtained frequency samples of the antenna array signals, summarize the output voltages of the antenna array signals with constant weights equal to the product of the amplitude and phase coefficients, which form spatial observation channels in the horizontal and vertical planes, square and perform secondary processing on each review cycle, for which they are summed over all frequency samples the output voltages of the formed spatial channels in a fixed frequency range are averaged over time, the noise signals are centered and normalized to Exe, they observe on each review cycle the received marks of the received noise signals and make a decision on detection by comparing with the threshold value the signal-to-noise ratio, characterized in that adaptive spatial observation channels are formed on each review cycle for each frequency reference, each which are formed by at least three adjacent spatial channels in the horizontal or vertical plane, for which they form mutual power spectra between the noise the signals of the spatial channels involved in the formation of the adaptive spatial observation channel, accumulate the mutual power spectra of the noise signals for a given accumulation time, compose a matrix of the accumulated mutual power spectra of the noise signals and perform orthogonal transformation of the matrix, calculate the phase coefficient vectors of the common-mode addition of the signals of the spatial channels involved in forming an adaptive spatial observation channel, calculate the output vector voltages of the spatial channels involved in the formation of the adaptive spatial channel by solving the vector-matrix equation for the orthogonally transformed matrix of the accumulated mutual power spectra and the vector of phase coefficients of the in-phase signal addition, the squaring is performed by calculating the response of the resulting adaptive spatial observation channel equal to the inverse of the sum of the squares of the vector elements output voltages, and secondary processing is carried out for output voltages adaptive spatial observation channels.

Images