RU2778168C1 - Method for matching signals detected by spaced hydroacoustic systems - Google Patents

Abstract

FIELD: hydroacoustics.

SUBSTANCE: method is based on receiving, detecting, and measuring the direction of hydroacoustic pulse signal separately by two spaced systems located on a common carrier. In the implementation of the method, a measured value of delay in the detection of two pulse signals and a predicted value of delay are formed for the hypothesis that the signal belongs to a single source. A mathematical measure of similarity between the measured value of delay and the predicted value of delay is formed, and the decision on matching the pulse signals is made if the similarity measure meets the resemblance criterion.

EFFECT: possibility of matching pulse signals if the hydroacoustic observation systems are spaced, and the direction to the signal does not provide information about a possible match, is provided; the observations systems therein may have non-overlapping working frequency ranges; additionally, the method provides a possibility of estimating the distance to the source of pulse signals if signals are detected in two systems, and the matching is successful.

1 cl, 2 dwg

Description

The invention relates to the field of hydroacoustics and is intended to identify impulse signals detected by different hydroacoustic systems located on a common carrier, the antennas of which are separated in space.

The task of integrating hydroacoustic systems, including the identification of signals detected by several systems, has always been posed by equipment developers [Andreev M.Ya., Okhrimenko S.N., Klyushin V.V., Rubanov I.L., Yakovlev V.A. Integrated system of underwater surveillance for a surface ship // Marine collection. 2006. No. 8. S. 50-51]. Systems operating in different frequency bands, having different receiving antennas and (or) information processing methods, are optimized for detecting different parts of a single hydroacoustic signal. Therefore, the identification of the signals detected by these systems will allow obtaining the most complete information about the signal, and, as a result, more fully recognize the source of the signal.

Known methods of identification [Neroslavsky B.L., Shchegoleva N.L. On the identification of track detectors in multichannel direction finding // Hydroacoustics. 2000. No. 2. P. 65-69, Zhandarov A.M. Identification and filtering of measurements of the state of stochastic systems. M.: Science. 1979], according to which, on the basis of a set of general estimates of parameters, the identification of signal forms found in different systems is made. These methods have been developed for cases of identifying signals found in similar systems and having a large set of common parameters (three or more parameters). Theoretical analysis shows the high efficiency of these methods of identification. However, in practice, when integrating highly heterogeneous systems, the list of common parameters in the forms of different systems is limited, and often consists of one parameter - the direction to the signal. This leads to a general disadvantage of identification methods, which is their low efficiency in practice.

There is a method [Volkova A.A., Nikulin M.N. RF patent No. 2684440. A way to identify objects detected by multiple systems. Published on 09.04.2019. IPC G01S 3/80], in which one parameter is used for identification - the direction to the signal. In accordance with this method, signals are received separately by two systems located on a common carrier, the directions to the signals are measured separately in each system, the values of the difference between the directions to the signals are formed, and a decision is made to identify the signals if the direction difference is less than the threshold. A limitation on the application of the method is the condition for measuring directions to signals from a single point in space. Therefore, the method cannot be applied to spatially separated systems, when directions to the signal are measured from different points in space, and even theoretically cannot coincide.

The closest analogue to the used operations to the present invention is the method [Volkova A.A., Nikulin M.N. Patent of the Russian Federation No. 2730103. Method for identifying objects detected by spatially separated systems. Published on 17.08.2020. IPC G01S 3/80], optimized for the case when directions to the signal are measured from different points in space.

This method contains the following main operations: signals are received separately by two systems located on a common carrier, identical arrays of measurements of each signal are formed in both systems, performing spatial selection and time-frequency processing using the same frequency range, a single sampling rate and the same sample length, form pairs of signals as each signal of one system with each signal of another system, calculate the cross-correlation function between the arrays of measurements for each pair of signals, find the maximum value of the cross-correlation function for each pair, make a decision on identification for pairs of signals for which the condition is met : the maximum value of the cross-correlation function is greater than the threshold.

The method is based on the calculation and analysis of the cross-correlation function between the measurement arrays of a pair of signals, which are previously reduced to a single frequency range. However, if the systems operate in fundamentally different, non-overlapping frequency ranges, then reduction to a single range is impossible, and, therefore, the correlation between the signals cannot be detected. Then, if it is necessary to detect broadband signals, including pulsed ones, in systems operating in non-overlapping frequency ranges, then this method cannot be applied. Impulse signals include, for example, the sonar signals of fishing trawlers. Another example of impulse signals is the impact of technical devices operating under water. Thus, the disadvantage of the method according to the patent of the Russian Federation No. 2730103 lies in the impossibility of using when detecting pulsed signals in systems operating in non-overlapping frequency bands.

The objective of the proposed method is to provide the possibility of identifying pulse signals detected in spatially separated systems, the frequency ranges of which do not intersect.

To solve the problem in the method of identifying signals detected by hydroacoustic space-diversified systems, in which

receive signals separately by two systems located on a common carrier,

carry out spatial selection in each system with the measurement of the direction to the signals separately in each system θ _1,i and 0 _2,j , where the first index determines the number of the system, and the second index - the number of the signal received by the system,

carry out frequency-time signal processing in each system, including band-pass filtering, detection and integration procedures,

pairs of signals ij are formed as each signal of one system with each signal of the other system,

make an identification decision for a pair of signals corresponding to the identification criterion,

new signs have been introduced, namely:

preliminarily measure the horizontal distance L between the phase centers of the antennas of the two systems and the speed of sound in water C,

configure both systems to receive pulse signals, for which, when integrating in each system, choose the accumulation time that is optimal for detecting signals of a given duration,

fix the detection time of pulse signals separately in each system T _1,i and T _2j ,

form the measured value of the detection delay for each pair of signals as Δ _ij ·=T _1,i - T _2,j ,

прогнозное значение дальности до источника от антенны второй систем

и прогнозное значение задержки

for the hypothesis of a single source of each pair of signals, a predictive value of the distance to the source from the antenna of the first system is formed

predicted value of the distance to the source from the antenna of the second systems

and predicted delay value

для каждой пары ij,form a mathematical measure of similarity between the measured value of the delay Δ _ij; and predicted delay value

for each pair ij,

and as an identification criterion, a similarity measure that satisfies the similarity criterion is taken, and for pairs ij of pulse signals that satisfy the similarity criterion, the distances R _1,ij and R _2,ij are taken as measured.

The technical result of the invention is the possibility of identifying pulse signals in the case when hydroacoustic surveillance systems are separated in space, and the direction to the signal does not provide information about a possible identification. At the same time, surveillance systems can have non-overlapping operating frequency ranges. Additionally, the method makes it possible to estimate the distance to the source of impulse signals, if the signals are detected in two systems, and the identification is successful.

We will show the possibility of achieving the specified technical result by the proposed method.

Let us determine the conditions under which a signal from one source will be detected by two hydroacoustic systems operating in non-overlapping frequency ranges. The signal from such a source must be broadband enough to appear in both systems. Broadband signals include, among other things, pulsed signals, for which it is known [Kharkevich A.A. Spectra and analysis. M.: State. Publishing House of Physics and Mathematics Literature. 1962] that the wider the frequency range, the shorter the pulse duration. Detection, recognition and, in particular, identification of impulse signals may be of interest for hydroacoustic systems for various purposes. At the same time, the pulse signal is characterized by a pronounced leading edge, which makes it possible to unambiguously determine the moment of its reception. This property of impulse signals can be used to identify them.

The traditional hydroacoustic noise direction finding system is applicable for detecting, among other things, impulse signals. To configure any noise direction finding system for detecting pulsed signals, it is sufficient, when integrating, which is part of a typical time-frequency processing, to select the accumulation time that is optimal for detecting signals of a given duration, which is provided for in the procedures of the claimed method. In addition, in the proposed method, the operations of fixing the time of detection of signals in each of the systems are introduced. This makes it possible to measure an additional parameter, namely the value of the detection delay of two pulse signals. The measured delay value is used further for identification as follows.

Known triangulation method for determining the distance to the signal source [Gamper L.E. On the accuracy of passive sonar methods with spaced onboard antennas // Hydroacoustics. Issue. 9. St. Petersburg: Science. 2009. S. 34-42], based on the reception of a signal by a system of two antennas spaced apart in space. To implement this method, which is based on the geometric solution of the triangle formed by the points Antenna 1-Antenna 2-Source (Fig. 1), it is necessary and sufficient to know any of the three parameters of the triangle. In one version of the method, the known parameters are the distance between the antennas, the direction to the source, relative to the first antenna, and the direction to the source, relative to the second antenna.

In our case, the problem of identification was not initially solved, that is, it is not known whether the Antenna 1-Antenna 2-Source triangle is formed. Therefore, the triangulation method [Gamper L.E. About the accuracy of the methods…] cannot be applied. However, for pulsed signals, as shown earlier, in addition to the three parameters required for the triangulation problem, a fourth parameter can be measured - the time delay between signal reception at one antenna relative to another antenna. This allows us to solve the inverse problem, namely, to check whether the Antenna 1-Antenna 2-Source triangle is formed. If a triangle is formed, this will mean that the signals received on both antennas belong to the same source.

We hypothesize that a triangle is formed. Using the parameters the distance between the antennas, the direction to the source, relative to the first antenna, and the direction to the source, relative to the second antenna, we hypothetically solve the triangulation problem, that is, we determine the predicted values of the distance to the source from both antennas. Further, knowing the speed of sound in water, we determine the predicted values of the signal travel times from the source to both antennas, and the difference between the travel times, that is, the predicted delay of the signal received by one antenna relative to the other. We compare the predicted value of the delay, obtained under the hypothesis that a triangle is formed, with the fourth parameter - the measured delay. If the results of the comparison match, then the hypothesis is correct, a triangle is formed, and the signals received by the two antennas belong to the same source. In this case, the triangulation problem becomes solved not hypothetically, and we get the value of the distance to the source from both antennas.

New operations of the method make it possible to test the indicated hypothesis for impulse signals, and, consequently, to solve the problem of identification in the case when hydroacoustic observation systems are separated in space.

The essence of the invention is illustrated by figures 1 and 2.

In FIG. 1 shows a diagram of the mutual arrangement of the antennas of the two systems and the radiation source in space, explaining the algorithm of the method. In FIG. 1 is marked: A ₁ - antenna of system 1, A ₂ - antenna of system 2, S - estimated position of the signal source, L - distance between the phase centers of the antennas of two systems, θ _1,i - direction to the signal measured in system 1, θ _{2 ,j} is the direction to the signal measured in system 2, R _1,ij is the distance to the radiation source from the system 1 antenna, R _2,ij is the distance to the radiation source from the system 2 antenna.

). Первый выход блока 2 (система 2) соединен со вторым входом блока 3 (измерение Δ), второй выход блока 2 соединен со вторым входом блока 4 (гипотеза

).·Выходы блоков 3 и 4 соединены с первым и вторым входами блока 5 (решение). Выход блока 6 (память) соединен с третьим входом блока 4.In FIG. Figure 2 shows a block diagram of the algorithm for implementing the proposed method for identifying signals detected by space-diversified systems. In FIG. 2 the first output of block 1 (system 1) is connected to the first input of block 3 (measurement Δ), the second output of block 1 is connected to the first input of block 4 (hypothesis

). The first output of block 2 (system 2) is connected to the second input of block 3 (measurement Δ), the second output of block 2 is connected to the second input of block 4 (hypothesis

). · The outputs of blocks 3 and 4 are connected to the first and second inputs of block 5 (solution). The output of block 6 (memory) is connected to the third input of block 4.

In an integrated hydroacoustic complex, consisting of two systems, the antennas of which are separated in space, the proposed method is implemented as follows. If there are more than two spaced systems in the complex, the method can be implemented for each pair of systems. Before starting work, the horizontal distance between the phase centers of the antennas of the two systems L and the speed of sound in water C are measured. The distance L can be measured directly at the antenna installation site or obtained from the dimensional drawings of the hydroacoustic complex carrier. To measure the speed of sound in water C, known indirect or direct methods are used [Komlyakov V.A. Shipborne instruments for measuring the speed of sound and modeling acoustic fields in the ocean. SPb.: Nauka. 2003]. The measured values of L and C are stored in block 6 (memory). Adjust the time-frequency signal processing in system 1 and system 2 so that it is optimized for the detection of pulsed signals. To do this, choose the integration time, consistent with the expected duration of the pulses. For example, for sonar fishing trawlers, the pulse duration is known from the data sheet. To be able to detect pulses of different durations, parallel processing of signals with several accumulation times can be provided.

In the dynamics of operation, system 1 and system 2 independently detect pulse signals, fix the detection time of signals T _1,i , T _2,j and measure directions to them θ _1,i , θ _2,j separately in each system. The operations of detection, processing, measurement of direction and detection time are standard for noise direction finding systems, and can be borrowed from [Handbook of hydroacoustics / A.P. Evtyutov, A.E. Kolesnikov, E.A. Korepin and others. L.: Shipbuilding. 1988]. Pairs of signals are formed as each signal of one system with each signal of another system. Further description is given for the pair ij.

The signal detection times measured in system 1 and system 2 are fed to block 3 (measurement Δ), in which the measured value of the detection delay of a pair of pulse signals ij is formed as Δ _ij ·=T _1,i - T _2,j .

). Одновременно в блок 4 из блока 6 поступают значения L и С. В блоке 4 формируют прогнозное значение задержки

для пары ij. Для этого, выдвигают гипотезу, что сигнал i, обнаруженный в системе 1, и сигнал j обнаруженный в системе 2, принадлежат одному источнику. Тогда, взаимное расположение антенн двух систем и источника излучения в пространстве может быть представлено схемой на фиг.1, На схеме фиг.1 видно, что точки А₁, А₂ и S являются вершинами треугольника, в котором известными (измеренными) параметрами являются углы θ_1,i, θ_2,j и расстояние L. Пользуясь формулами для решения треугольников [Бронштейн И.Н., Семендяев К.А. Справочник по математике. М: гос. изд-во технико-теор. лит-ры. 1956], можно получить выражения для предполагаемой (гипотетической) дальности до источника S от антенн обеих систем А₁ и А₂:In parallel with this, the signal directions measured in system 1 and system 2 are sent to block 4 (hypothesis

). At the same time, block 4 receives the values L and C from block 6. In block 4, the predicted delay value is formed

for the pair ij. For this, it is hypothesized that the signal i detected in system 1 and the signal j detected in system 2 belong to the same source. Then, the relative position of the antennas of the two systems and the radiation source in space can be represented by the diagram in figure 1. The diagram of figure 1 shows that the points A ₁ , A ₂ and S are the vertices of a triangle in which the known (measured) parameters are the angles θ _1,i , θ _2,j and distance L. Using the formulas for solving triangles [Bronshtein I.N., Semendyaev K.A. Handbook of mathematics. M: Mrs. Publishing House of Techno-Theor. liters. 1956], it is possible to obtain expressions for the estimated (hypothetical) range to the source S from the antennas of both systems A ₁ and A ₂ :

Then, knowing the speed of sound in the medium, it is possible to determine the predicted value of the delay between the time of arrival of the signal from the source to antenna A ₁ and antenna A ₂ :

обозначены символом ƒ (функция).Calculations according to formulas (1) and (2) are implemented in block 4. In the diagram of Fig. 2 formulas for R _1,ij , R _2,ij and

are marked with the symbol ƒ (function).

из блока 4 поступают в блок 5, где осуществляется их сравнение. В качестве меры сходства может использоваться любая из известных мер сходства, например, евклидово расстояние [Ту Дж., Гонсалес Р. Принципы распознавания образов / Пер. с англ. М.: Мир. 1978]. Порог для принятия решения выбирается, например, на основании требуемой вероятности ложного отождествления согласно [Тюрин A.M. Введение в теорию статистических методов в гидроакустике. Л. 1963]. На схеме фиг. 2 вычисление меры сходства обозначено символом М. Если для пары сигналов ij, обнаруженных в двух системах выполняется условие: мера сходства между измеренным и прогнозным значением задержки менее порога

то гипотеза о том, что эти сигналы принадлежат одному источнику, считается верной (треугольник образуется), и принимается решение об их отождествлении. Дополнительно, считаются измеренными дальности от источника сигнала до антенн обеих систем, полученные по формулам (1). В противном случае, гипотеза отклоняется, решение об отождествлении не принимается, дальности до источника сигнала считаются не измеренными.Measured delay value Δ _ij from block 3 and predicted delay value

from block 4 go to block 5, where they are compared. As a measure of similarity, any of the known measures of similarity can be used, for example, the Euclidean distance [Tu J., Gonzalez R. Principles of pattern recognition / Per. from English. M.: Mir. 1978]. The decision threshold is chosen, for example, based on the required probability of false identification according to [Tyurin AM Introduction to the theory of statistical methods in hydroacoustics. L. 1963]. In the diagram of Fig. 2, the calculation of the similarity measure is indicated by the symbol M. If for a pair of signals ij detected in two systems, the following condition is met: the similarity measure between the measured and predicted delay values is less than the threshold

then the hypothesis that these signals belong to the same source is considered correct (a triangle is formed), and a decision is made to identify them. Additionally, the distances from the signal source to the antennas of both systems, obtained by formulas (1), are considered to be measured. Otherwise, the hypothesis is rejected, the identification decision is not made, the distances to the signal source are considered not measured.

The method works for every pair of signals. The number of such pairs will be n×m, where n is the number of signals in one system, m is the number of signals in another system. In hydroacoustics, usually no more than ten signals are detected simultaneously in each system. Then the maximum number of pairs will not exceed 100. For each pair of signals, the operations of blocks 3 and 4 are independently repeated. For each pair of signals, its own similarity measure is formed in block 5, which is compared with the threshold, and a decision is made or not made to identify this particular pair of signals.

All of the above allows us to consider the problem of the invention solved. A method for identifying signals detected by spatially separated hydroacoustic systems is proposed, which can be applied in the case when the detected signals are pulsed.

Claims (1)

A method for identifying signals detected by hydroacoustic spatially separated systems, in which signals are received separately by two systems located on a common carrier, spatial selection is carried out in each system with direction measurement to signals separately in each system θ _1,i and θ _2,j , where the first index determines the number of the system, and the second index - the number of the signal received by the system, in each system, frequency-time processing of signals is carried out, including the procedures of band-pass filtering, detection and integration, pairs of signals ij are formed as each signal of one system with each signal of another system, make a decision on identification for a pair of signals corresponding to the identification criterion, characterized in that they first measure the horizontal distance L between the phase centers of the antennas of two systems and the speed of sound in water C, tune both systems to receive pulse signals, for which, when integrating in each system choose the accumulation time that is optimal for detecting signals of a given duration, fix the detection time of pulse signals separately in each system T _1,i and T _2,j , form the measured value of the detection delay for each pair of signals as Δ _ij =T _1,i -T _{2 ,j} , for the hypothesis of a single source of each pair of signals, a predictive value of the distance to the source from the antenna of the first system is formed

predicted value of the distance to the source from the antenna of the second system

and predicted delay value

form a mathematical measure of similarity

as the Euclidean distance between the measured delay value Δ _ij and the predicted delay value

for each pair ij, and as an identification criterion, they take a similarity measure that satisfies the similarity criterion

and for pairs ij of pulse signals that satisfy the similarity criterion, the distances R _1,ij and R _2,ij are taken as measured.

Images