RU2797779C1 - Method for determining the range to a moving radiation source using passive sonar - Google Patents

Abstract

FIELD: hydroacoustics.

SUBSTANCE: method for determining the range to a moving radiation source using a passive sonar contains M spaced sonar antennas (M≥3), preliminary processing, including discretization, digital transformation and Fourier transform (FT), means of review in a given sector of observation and range interval, identification of the coordinates of the detected signal mark on the view indicator α_n, D_n, and to determine and track the current direction to a moving radiation source α_s ^*(t) each m-th antenna is phased in the direction from its center to the OT, at each i-th DFT cycle, the complex paired mutual power density spectra (PMPDS) of the signals of the antenna pairs of the receiving system are measured and the direction deviation discriminator is used. To determine the current range value on each i-th cycle of the DFT, the discriminator of the deviation of its reference range value D_refi from the desired true D_Si is used, while applying a nonlinear range scale to the virtual focus points of the receiving system R_fi with coordinates <α_refi, r_f>, for which the lower limit of the range interval D_fn is predetermined on the linear scale D, in which the radiation source is supposed to move in the process of observation and tracking: D≥D _fn, and on each next i-th cycle of the DFT the reference virtual focusing point of the receiving system R_fi with coordinates <α_refi, r_refi> is designated in it so that the direction to R_fi coincides withα_refi, and the range r_refi is related to D_refi by a rigid non-linear relationship: r_refi = D_refi ^-1 * γ_f, where γ_f is the scale factor, and the magnitude and sign of the deviation δ_Ri of the current reference value r_refi is determined based on the desired range value to the virtual source, and the current estimate of the actual range on the i-th DFT cycle is determined as D_refi ^* = D_refi -δ_{R i} and subjected to dynamic averaging by known methods for determining the parameters of successive measurements and predicting their values for subsequent measurement cycles.

EFFECT: improving the accuracy of determining the range, especially at large values of D/B, where B is the base size of the PS.

2 cl, 5 dwg

Description

The invention relates to the field of hydroacoustics, namely to passive sonar using airborne, towed and stationary extended and spaced antennas, and can be used in the near zone of radiation from a source with a circular (spherical, cylindrical) wave front, as well as in atmospheric acoustics and passive radar .

With passive sonar (PGL) sources of hydroacoustic radiation in the near field of the source field (Fresnel zone), extended (wide-aperture) receiving systems (PS) from M (M≥3) hydroacoustic receivers or spaced directional local antennas (hereinafter referred to as "PS antennas") are used. Means of preliminary and inter-antenna spatiotemporal information processing (MA air defense) provide a two-coordinate view in a given sector of directions α and range interval D on an indicator with a coordinate grid <α, D> [Gamper L.E. Optimal spatio-temporal processing in passive sonar systems with diversity antennas // Nauchno-tekhn. Sat. Hydroacoustics. Issue. 10 St. Petersburg, 2009. S. 49-60]. These tools ensure the observation of the “useful” signal of the radiation source S in the form of a “signal mark” due to the “response” of the receiving system to its signal against the background of interference emissions, with a “main (global)” maximum that provides reading of the desired coordinates of the radiation source α _S , D _S on the scales α, D [Patent of the Russian Federation No. 2507531, registration in the State Register on February 20, 2014.]. MA air defense includes "focusing" the receiving system to the point of the supposed location of the source by compensating for the difference in the time of arrival of the signal from the source S to the centers of the antennas of the PS ("phasing" of the antennas) and determining the square of the modulus of the sum of these signals. Another MA air defense method used for accurate reading and autotracking of the coordinates of the radiation source is the use of discrimination of the deviation of the current values of the estimates of the direction α ^* (t) and the distance D ^* (t) from their reference values α _op (t) and the range D _on (t ), specified or tracked by previous measurements, by the system [Loskutova G.V., Polkanov K.I. Spatial-frequency and frequency-wave methods for describing and processing hydroacoustic fields. St. Petersburg: Science. 2007].

Determining the range in the near field is associated with specific complexity. It lies in the fact that the main and side maxima of the response of the focused receiving system z _max (α, D) in the Fresnel zone have the property of asymmetric expansion in the range section z _max (D, α _s =const) with an increase in the measured parameter D [Le Cadre JP Performance Analysis of Wavefront Curvature Methods for Range Estimation. IEEE Trans, on Aerospace and Electronics. Vol. 31, no. 3, July 1995. P. 1082-1103]. This asymmetric expansion is small at short ranges, but rapidly increases at large distances, worsening the accuracy of determining the coordinate and the range resolution of sources at D _s >>B _PS , where D _s is the distance to the source, B _PS is the basic size of the PS (distance between centers end antennas of the PS). With the expansion of the main maximum of the response, the steepness of the discrimination characteristic also decreases (dependence of the output value of the discriminator on the magnitude of the desired deviation, i.e. sensitivity), since it is proportional to its derivative, therefore, the error σ _D of the range estimate and its dependence on the measured parameter D _s increase. This property of an asymmetric expansion of the range section of the response of the receiving system, a decrease in the steepness of the discrimination characteristic of the range and an increase in the error from the range estimate is an objective [Article Le Cadre JP], a "geometric" feature of the spatial selectivity of the receiving system in the near zone [Schultheiss PM Passive Sonar Detection in the presence of Interference // Jorn. Acoust. soc. Am. 1968. V. 43. R. 418-425]. In this case, the cross section in the direction does not depend on the range, similar to the directivity characteristic of the antenna in the far zone. The purpose of the proposed technical solution is to improve the accuracy of determining the range to the radiation source at any distance in the observation interval in the near zone by reducing the "geometric" dependence of the slope of the discrimination characteristic on the measured parameter - range.

пропорциональна производной характеристики направленности (ХН) антенны z(α,α_оп). Главный максимум ХН - приблизительно симметричная функция относительно α_оп, с максимумом в точке (α=α_оп), следовательно ДХ - антисимметричная функция с нулевым значением в точке (α=α_оп). Линейный (приблизительно, «квазилинейный») участок в окрестностях нулевого значения используется для измерения отклонения δ_α с использованием крутизны ДХ:

There is a method for determining the angle of arrival of the plane-wave front of the signal of a moving source in the far zone of a two-channel receiving system [Loskutova G.V. and etc.]. For this, measuring devices are used that contain discriminators of the measured parameter and dynamic smoothing devices. This method uses a receiving system of two antennas or an antenna with two (or more) channels formed on it with phase centers separated by a distance of d _f (for example, the centers of the two halves of the antenna). The channel antennas compensate (phase) independently, in one assigned reference direction α _op and, to determine the desired unknown direction α _s , measure the magnitude and sign of the deviation of the reference direction α _op from the desired unknown α _s using the signals of both channels, and then correct the reference direction by the value of the measured deviation δ _α . A feature of this discrimination method is that its characteristic (DH)

is proportional to the derivative of the directivity characteristic (ХН) of the antenna z(α,α _op ). The main maximum XH is an approximately symmetric function with respect to α _op , with a maximum at the point (α=α _op ), therefore, DH is an antisymmetric function with zero value at the point (α=α _op ). A linear (approximately "quasi-linear") section in the vicinity of the zero value is used to measure the deviation δ _α using the slope of the DH:

The disadvantages of this method are that, firstly, it uses the information of only two receiving channels, while passive range determination by the curvature of the wave front in a circular field is possible only at M≥3, and secondly, it can only be used to determine one coordinate - the angle of arrival of a signal with a flat wave front, and the purpose of passive sonar is to determine both the direction and range, in the general case also the elevation angle. To determine the coordinates of the radiation source, including the range when receiving a signal with a circular (spherical, cylindrical) wavefront, this method cannot be used, but the deviation discrimination method itself, as shown below, is applied both in the prototype and in the proposed method, with necessary changes and new features to determine both direction and range.

To reduce the dependence of the steepness of the discriminator characteristic on the measured parameter - the range, it is necessary to select some parameters of the model of the system <source / receivers of the PS> and the method of their deformation in the desired direction to transform the shape of the DC, which ensures the stabilization of its steepness in the required range interval D.

, при этом значения распределяются на шкале нелинейно и не эквидистантно, что в некоторых случаях неудобно наблюдателю. Другим примером этого метода является замена графика z(x) с линейной шкалой х, графиком z'(x^-1), в котором помечены деления шкалы х^-1 соответствующими значениями х с неравномерным шагом (что также не всегда удобно на практике). Этот метод частично решает одну из поставленных выше задач, позволяя получить дискриминационную характеристику приблизительно одинаковой крутизны в окрестностях нулевого отклонения х - х₀=0, однако, это не улучшает точности определения параметра х, т.к. сама зависимость (уже измеренная) при этом не меняется, а лишь графически деформируется.Methods are known for changing the shape of a graphic image of an object on an indicator picture for any purpose according to the principle of a “distorted mirror”: by changing the linear scale of the image along one or another axis of the graph according to a given non-linear rule. A well-known example of this method is the logarithmic scale, where the plot of z(x) with the linear scale x is replaced by the plot

, while the values are distributed on the scale non-linearly and not equidistantly, which in some cases is inconvenient for the observer. Another example of this method is to replace the z(x) graph with a linear x scale with a z'(x ^-1 ) graph that labels the x ^-1 scale divisions with the corresponding x values in non-uniform increments (which is also not always convenient in practice). This method partially solves one of the above problems, making it possible to obtain a discriminatory characteristic of approximately the same steepness in the vicinity of the zero deviation x - x ₀ =0, however, this does not improve the accuracy of determining the parameter x, because the dependence itself (already measured) does not change, but only graphically deforms.

There are also known methods for determining the coordinates of a moving radiation source by dynamic processing of the sequence of measuring the direction to it [Kuzmin S.Z. Digital processing of radar information. M.: Sov. radio. 1967]. They often use the technique of estimating not the sought-for parameter itself - the distance D, but the parameter R=D ^-1 , which makes it possible to simplify the solution of the problem with less computational resources. The disadvantages of these methods are: the use of only direction measurements, a long time for solving the problem, the need for a hypothesis about the nature of the source trajectory, etc.

The closest in terms of the set of common features to the proposed technical solution is the “Method for passively determining the coordinates of a moving radiation source” [RF Patent No. 2623831, published on June 29, 2017], containing the reception of signals with a circular wave front by a receiving system (PS) of M antennas ( M≥3) located in space in a known way, preliminary and inter-antenna space-time processing, an indicator with a two-coordinate grid and direction scales α and range D, displaying the location of radiation sources in the field of observation in the form of signal marks. To accurately determine the coordinates and support, a reference point (OT) is assigned with coordinates α _op and D _op , which at the time of pointing are taken equal to: α _op (t ₀ )=α _max (t ₀ ) and D _op (t ₀ )=D _max (t ₀ ) and then, in the process of their maintenance, they correct and predict by known methods of dynamic smoothing of estimates [Kuzmin S.Z., monograph].

. Определяют значение «измерительной» части ДХН - с₁, суммируя по частоте значения ПВСПМ каждой пары с умножением на частотную характеристику приемного тракта по выходу предварительной обработки h_k, на коэффициент exp(-j2πƒτ_mni) компенсации разности времен прихода сигнала mn-й пары антенн и на частоту ƒ_k, затем суммируя результаты всех пар антенн с весовыми коэффициентами w_amni. Аналогично определяют значение «нормирующей» части ДХН - c₂, только в первой сумме слагаемые умножают на квадрат частоты, а во второй - на квадрат w_amni. Таким образом, получают значение отклонения направления δ_ai=c₁/c₂, в единицах измеряемого параметра, а оценку направления на i-м цикле ДПФ определяют как:

.At each i-th cycle of the DFT, according to the signals received by the antennas, estimates of the deviation δ _αi , δ _{Di of} the OT coordinates from the desired current coordinates α _s (t _i ), D _s (t _i ) of the source, respectively, are determined, for which each m-th antenna is phased in direction from its center to OT and form mn pairs of antennas, with m, n=1, ..., M, with m<n, in total (M2-M)/2=M n _pairs with non-repeating indices mn. For each mn-th pair of antennas, the pairwise cross power density spectra (PRSPs) of the signals are measured. The subsequent stages of determining the estimates of the deviations of the direction δ _αi and range α _Di in the prototype are different. The angular sections of the PS response, like the directivity characteristics of antennas in the far field, are symmetrical with respect to the phasing direction and, in the first approximation, depend little on the direction and range, and the discriminatory directional deviation characteristic (DKhN) is approximately antisymmetric about zero at α _opi = α _Si . Therefore, for discrimination δ _αi use the known method [Loskutova G.V. et al., monograph], including: calculation of the differences τ _mni of the times of signal arrival from point S to the centers of each mn-th pair of antennas and weight coefficients

. The value of the "measuring" part of the DHN is determined - from ₁ , summing up the frequency of the values of the PVSP of each pair multiplied by the frequency response of the receiving path according to the preprocessing output h _k , by the coefficient exp(-j2πƒτ _mni ) compensation for the difference in the arrival times of the signal of the mn-th pair of antennas and frequency ƒ _k , then summing the results of all pairs of antennas with weights w _amni . Similarly, the value of the "normalizing" part of the DCN - c ₂ is determined, only in the first sum the terms are multiplied by the square of the frequency, and in the second - by the square w _amni . Thus, the direction deviation value δ _ai =c ₁ /c ₂ is obtained, in units of the measured parameter, and the direction estimate on the i-th DFT cycle is determined as:

.

To simplify the description (without losing the essence of the features under consideration), we agree to assume that the steepness of the normalized DR does not depend on the signal level (it is supported by independent adjustment of the power of the radiation source), and the standard deviation (RMS) of fluctuations in the output value of the discriminator is determined by the signal-to-noise ratios at the outputs antennas, which, for simplicity, we will accept constant and the same on all antennas of the PS.

оси дальностей D, иначе говоря искать нулевую точку ДХ на оси D в окрестностях D_опi. Для этого сканируют величину D в окрестностях D_опi, используя М_п нечетных попарных взаимно-корреляционных функций (НВКФ) сигналов антенн, комплексное накопление и обратное преобразование Фурье.In contrast to the direction, the long-range sections of the PS response are not symmetrical, and their shape and the width of the main maximum strongly depend on the range, more precisely, on the ratio D _S /B, where B is the size of the PS base. Accordingly, the range discrimination characteristic (DCD) is not an antisymmetric curve, and its steepness in the linear section depends on the range, rapidly decreasing with increasing D _S /B, and therefore, use the same algorithm for determining the deviation in range as in the direction not advisable. Therefore, in the prototype, it is proposed to measure not the deviation δ _Di , but to look for the intersection point of the DC

distance axis D, in other words, to look for the zero point of the DH on the D axis in the vicinity of D _opi . To do this, scan the value of D in the vicinity of D _opi using M _p odd pairwise cross-correlation functions (NVKF) antenna signals, complex accumulation and inverse Fourier transform.

The advantage of this method is to increase the accuracy of determining the coordinates of a moving radiation source due to the use of a discriminatory method for estimating coordinates (zero count) and to reduce errors associated with the "inertia" of static averaging, which depend on the speed of the relative movement of the source and receivers. The disadvantages of the method include the "geometric" dependence of the error in determining the range σ _D on the value of the measured range due to the decrease in the steepness of the characteristics of the range discriminator Δ _D with an increase in its value D _S .

The objective of the invention is to improve the accuracy of determining the distance to the radiation source by a passive sonar in the near field zone of a moving radiation source in the observation interval, by reducing the "geometric" dependence of the steepness of the discrimination characteristic of the range on the measured parameter.

The technical result of the invention is to provide discrimination of the deviation of the range estimate with a characteristic that is quasi-invariant to the range value in a given observation interval of a passive sonar, which leads to an increase in the accuracy of determining the range, especially at large values of the range ratio D to the base size of the PS: D/B _PS .

, которые на момент начала работы средств точного определения координат и сопровождения используют как опорную точку (ОТ) с координатами

, а для определения и сопровождения текущего направления на движущийся источник излучения

фазируют каждую m-ю антенну в направлении от ее центра на опорную точку (ОТ) и на каждом i-ом цикле ДПФ измеряют комплексные попарные взаимные спектры плотности мощности (ПВСПМ) сигналов всех mn-х пар антенн приемной системы (m,n=1,…, М), используют дискриминацию текущих отклонений направления от опорной точки α_опi и известные методы динамического сглаживания последовательности оценок

введены следующие новые признаки:To provide the specified technical result in a method for passively determining the coordinates of a moving radiation source, containing the reception of signals M by passive sonar antennas, M≥3, pre-processing, including amplification, filtering with a frequency response h _k in the reception band ƒ _n ≤ƒ _k ≤ƒ _in , synchronous sampling, digital transformation, cyclic discrete Fourier transform (DFT), means of viewing in a given angular sector of observation and range interval, determining the coordinates of a detected signal mark on the view indicator

, which at the time of the start of work of the means of precise determination of coordinates and tracking are used as a reference point (GT) with coordinates

, and to determine and track the current direction to a moving radiation source

each m-th antenna is phased in the direction from its center to the reference point (RT) and on each i-th DFT cycle the complex pairwise mutual power density spectra (CPRMS) of the signals of all mn-th pairs of antennas of the receiving system (m, n = 1) are measured ,…, M), use the discrimination of the current deviations of the direction from the reference point α _opi and the known methods of dynamic smoothing of the sequence of estimates

The following new features have been introduced:

где γ_ф - масштабный коэффициент, равный квадрату нижней границы D_фн:

, и определяют величину и знак отклонения δ_Ri текущего опорного значения r_опi от искомого значения дальности до виртуального источника, а текущую оценку фактической дальности на i-м цикле ДПФ определяют как

и подвергают последующему динамическому осреднению известными методами определения параметров последовательных измерений и прогноза их значений на следующие циклы измерений.to determine the current range value by a passive sonar on each i-th cycle of the DFT, discrimination of the deviation of its reference range value D _opi from the desired true D _Si is used, while applying a non-linear range scale r _f to the virtual focus points of the receiving system with coordinates <α _opi , r _f >, for this, on a linear range scale D, the lower limit D _fn of the range interval is predetermined, in which the radiation source will supposedly move in the process of observation and tracking: D≥D _fn . The upper limit is determined only by the validity of the application of the wavefront curvature method in given specific conditions - source power, signal-to-noise ratio, hydrology, the technical level of sensitivity to the phase relationships of the PS antenna signals, etc.). At each next i-th DFT cycle, a virtual reference focusing point of the receiving system R _opi is assigned with coordinates <α _opi , r _opi >, so that the direction to R _opi coincides with α _opi , and the distance r _opi is related to D _opi by a nonlinear dependence :

where γ _f - scale factor equal to the square of the lower boundary D _fn :

, and determine the magnitude and sign of the deviation δ _Ri of the current reference value r _opi from the desired range value to the virtual source, and the current estimate of the actual range on the i-th DFT cycle is determined as

and subjected to subsequent dynamic averaging by known methods for determining the parameters of successive measurements and predicting their values for the next measurement cycles.

, используют дискриминацию отклонения дальности. Для этого рассчитывают разность времен распространения сигнала t_mRi из виртуальной точки R_опi(α_опi, r_опi) к центрам Om, O_n каждой mn-й пары антенн: τ_mnRi=t_mRi-t_nRi (m, n=1,…, М) по формулам тригонометрии, используя скорость звука в среде с, а также весовые коэффициенты w_rmni, равные производной разности времен распространения τ_mnRt no r_i, т.е.

, суммируют по частоте ƒ_k мнимые части ПВСПМ каждой mn-й пары антенн в полосе приема [ƒ_н, ƒ_в] с умножением каждого слагаемого на коэффициент частотной характеристики h_k, на частоту ƒ_k и на коэффициент e_kmni=exp(-j2πƒ_kτ_mnRi) компенсации разности времен распространения сигнала от опорной виртуальной точки R_фi к центрам каждой mn-й пары антенн, в результате получают для каждой mn-й пары суммарные величины C_mni i-го цикла ДПФ, затем суммируют по частоте ƒ_k вещественные части ВСПМ с умножением каждого слагаемого на h_k, на коэффициент компенсации e_kmni и на квадрат частоты

получая для каждой пары суммарную величину G_mni, затем суммируют по индексу mn величины C_mni с весовыми коэффициентами w_rmni, получая величину c_i, которую делят на сумму по индексу mn величин G_mnt с тем же весовым коэффициентом в квадрате

, получая величину g_i и определяют отклонение δ_ri=c_i/g_i, выполняют осреднение оценок отклонения δ_ri по i=1, 2, …, I циклам ДПФ, получая малоосредненную оценку

, а текущую оценку фактической дальности на i-м цикле ДПФ определяют как

.To determine the magnitude and sign of the deviation δ _gi of the current reference value r _opi from the desired value of the distance to the virtual source

, use range deviation discrimination. To do this, calculate the difference in the propagation time of the signal t _mRi from the virtual point R _opi (α _opi , r _opi ) to the centers Om, O _n of each mn pair of antennas: τ _mnRi =t _mRi -t _nRi (m, n=1,… , M) according to the trigonometry formulas, using the speed of sound in the medium c, as well as the weighting coefficients w _rmni equal to the derivative of the propagation time difference τ _mnRt no r _i , i.e.

, sum over the frequency ƒ _k the imaginary parts of the FSPDM of each mn-th pair of antennas in the reception band [ƒ _n , ƒ _in ] with the multiplication of each term by the coefficient of the frequency response h _k , by the frequency ƒ _k and by the coefficient e _kmni =exp(-j2πƒ _k τ _mnRi ) compensation for the difference in signal propagation time from the reference virtual point R _fi to the centers of each mn pair of antennas, as a result, for each mn pair, the total values C _mni of the i-th DFT cycle are obtained, then _the real parts of the SPDT with each term multiplied by h _k , by the compensation factor e _kmni and by the square of the frequency

obtaining for each pair the total value G _mni , then sum over the index mn the values C _mni with weight coefficients w _rmni , obtaining the value c _i , which is divided by the sum over the index mn values G _mnt with the same weight coefficient squared

, receiving the value g _i and determine the deviation δ _ri =c _i /g _i , perform the averaging of estimates of the deviation δ _ri over i=1, 2, ..., I DFT cycles, obtaining a slightly averaged estimate

, and the current estimate of the actual range on the i-th DFT cycle is defined as

.

. Этот масштаб обладает следующими свойствами: фактически, при любой дальности D в интервале D=D_фн÷∞ дискриминационная характеристика дальности формируется с использованием значений расстояния в интервале r_ф=0÷D_фн, где велика крутизна ДХД, и при этом сохраняется тождество*):

, т.е. нулевое значение виртуальной ДХД на основной, линейной шкале дальностей D всегда находится в точке фактического расположения источника излучения D_S, а ее крутизна близка к крутизне на дальности r_фн, даже несколько увеличиваясь с увеличением фактической дальности D_S. Предложенный выбор нелинейности использует особенность геометрии фрагмента дуги радиуса R_д, стянутого хордой L_д: высота его сегмента h_д равна:

, где α_ц - центральный угол дуги. Применяя метод приближенных вычислений при R_д>>L_д/2, можно получить:

, где η_д - коэффициент, определяемый геометрией схемы <дуга/радиус>:

. Применяя эту нелинейную зависимость в расчетах коэффициента компенсации разности времен распространения сигнала от опорной виртуальной точки R_фi к центрам каждой mn-й пары антенн e_kmni=exp(-j2πƒ_kτ_mnRi) получают стабильную форму главного максимума отклика ПС и ДХД.The introduction of new features makes it possible to increase the accuracy of determining the range to a moving source of radiation by a passive sonar at long ranges by providing a high steepness of the discriminatory range characteristic (DRD) - at the level of its value in a given lower boundary of the observation interval D _fn , and its stabilization at any distance in the interval of near zones. The scale of the angular sections does not change in this case. The novelty of the proposed solution lies in the fact that when forming a discriminatory characteristic, a nonlinear scale is used in long-range sections, of the form:

. This scale has the following properties: in fact, at any distance D in the interval D=D _fn ÷∞, the discriminatory characteristic of the range is formed using the distance values in the interval r _f =0÷D _fn , where the steepness of the DCD is large, and at the same time the identity is preserved *) :

, i.e. the zero value of the virtual DCD on the main, linear range scale D is always located at the point of the actual location of the radiation source D _S , and its steepness is close to the steepness at the distance r _fn , even slightly increasing with increasing actual range D _S . The proposed choice of nonlinearity uses a feature of the geometry of a fragment of an arc of radius R _d , contracted by a chord L _d : the height of its segment h _d is equal to:

, where α _c - the central angle of the arc. Applying the method of approximate calculations for R _d >>L _d /2, you can get:

, where η _d is a coefficient determined by the geometry of the scheme <arc/radius>:

. Applying this non-linear dependence in calculating the compensation coefficient for the difference in signal propagation times from the reference virtual point R _phi to the centers of each mn-th pair of antennas e _kmni =exp(-j2πƒ _k τ _mnRi ) we obtain a stable shape of the main maximum of the PS and DCD response.

*) Note: Please note that when moving from direct focusing to point S to focusing to a virtual point _Rf , the sign of DCD changes to the opposite (see Fig.5).

The essence of the alleged invention is illustrated by figures 1-5:

Fig. 1 is a geometric diagram of a source-antenna system for a passive sonar with three antennas.

Fig. 2 graphs explaining the ratio of distances to the focal points of the receiving system for real (prototype) and virtual (proposed) positions of the radiation source.

Fig. 3 is a diagram of a passive sonar that solves the problem of determining the range by the proposed method.

Fig. 4 is a diagram of block 4 for determining the range to the radiation source by the proposed method.

Fig. 5 graphs of discrimination characteristics of range deviations from the reference point when using the traditional and proposed methods of discrimination in the range interval, FIG. a) according to the prototype method, b) according to the proposed method.

The principle of operation of the proposed invention can be illustrated by an example that implements the solution of a two-coordinate problem with three linearly located antennas A ₁ , A ₂ , A ₃ , as shown in the diagram of Fig. 1. where O ₁ , O ₂ , O ₃ are the centers of the PS antennas, S is the radiation source with coordinates (α _s , D _s ), D _fn /R _fn is the common point of the linear scale D and non-linear R with coordinates (α _s , r _fn ).

On the graphs of Fig. 2 shows the ratio of the distances to the focusing points of the receiving system in real D _f (prototype) and virtual r _f (proposed) positions of the radiation source, D _s - the position of some source S on the linear scale D _f , r _fS - its position on the non-linear scale r _f .

- блок 3 (например, по Патенту РФ №2623831) и дальности

- блок 4 (по предлагаемому способу), блок динамического сглаживания 5 (программируемый процессор), вычислитель 6 (программируемый процессор), блок управления 7 (программируемый контроллер), индикатор 8 и блок обзора 9 (например, по Патенту РФ №2507531).The proposed method can be implemented using a device (passive sonar), the schemes of which are shown in Fig. 3, 4. The device (Fig. 3), contains known blocks and modules according to the proposed method: block of the receiving system 1 of M antennas 1.1 + 1.M, in block 2: phasing module 2.1 (by known methods of pre-processing and shaping the directivity of antennas) and a module for measuring cross-spectrum power density (CPDS) from the outputs of all pairs of antennas - module 2.2, blocks for determining low-average estimates: directions

- block 3 (for example, according to RF Patent No. 2623831) and range

- unit 4 (according to the proposed method), dynamic smoothing unit 5 (programmable processor), calculator 6 (programmable processor), control unit 7 (programmable controller), indicator 8 and review unit 9 (for example, according to RF Patent No. 2507531).

, цифровое преобразование, циклическое дискретное преобразование Фурье (ДПФ) выборок объема N_ДПФ сигналов с выходов антенн.Block 1 includes M antennas, as well as pre-processing elements controlled from block 7: synchronous sampling with a frequency

, digital transformation, cyclic discrete Fourier transform (DFT) of samples of volume N _DFT of signals from antenna outputs.

и их производных

в блоке 5.Block 2 includes a module 2.1 for phasing antennas in directions from the center of the antenna to the reference point (OT) α _mopi , m=1÷M, and a module 2.2 for forming M _p pairs of mn antennas and determining the complex cross power density spectra (CPDS) of their signals . OT coordinates come from block 6 (computer) based on the results of dynamic smoothing of coordinate estimates

and their derivatives

in block 5.

(известным способом, например, по Патенту РФ №2623831) и дальности

(предлагаемым способом), соответственно, путем измерения отклонений координат источника излучения δ_αi, δ_ri от координат виртуальной опорной точки R_фопi. Для этого в эти блоки поступают измерения ВСПМ М_п пар антенн (Re и Im) из блока 2, коэффициенты компенсации разностей задержек сигнала в парах антенн e_kmni=exp(-j2πƒ_kτ_mnRi) и весовые коэффициенты

и

.Blocks 3 and 4 implement estimates of the current direction

(in a known way, for example, according to the Patent of the Russian Federation No. 2623831) and the range

(by the proposed method), respectively, by measuring the deviations of the coordinates of the radiation source δ _αi , δ _ri from the coordinates of the virtual reference point R _fopi . To do this, these blocks receive measurements of the HSPD M _p pairs of antennas (Re and Im) from block 2, the compensation coefficients for the differences in signal delays in the pairs of antennas e _kmni =exp(-j2πƒ _k τ _mnRi ) and weight coefficients

And

.

показан на фиг. 4, где: 4.1 - модуль суммирования по частоте мнимых элементов ВСПМ для каждой пары антенн с умножением на h_k, ƒ_k и e_kmnij (здесь и далее в скобках указан блок или модуль из которого поступают данные или куда отправляются), результаты - величины C_mni для всех Мп пар антенн поступают в модуль 4.2 - модуль суммирования по индексу mn с весовым коэффициентом w_αmni, получая величине c_i, пропорциональне. оцениваемому отклонению δ_ri. 4.3 - модуль суммирования по частоте вещественных элементов ВСПМ для каждой пары антенн с умножением на h_k,

и e_kmnij. 4.4 - модуль суммирования по индексу mn с весовым коэффициентом w² _αmni - получение нормирующей величины g_i. 4.5 - модуль оценки отклонения δ_αij путем нормирования ci на g_i. 4.6 - малое осреднение оценки отклонения по i. 4.7 - оценка направления путем корректировки опорной точки на величину отклонения.An example of the implementation of block 4 for measuring the deviation δ _ri of the distance from the desired one and obtaining an estimate of the coordinate

shown in FIG. 4, where: 4.1 is the summation modulus for the frequency of the imaginary elements of the SSPD for each pair of antennas multiplied by h _k , ƒ _k and e _kmnij (hereinafter, the unit or module from which the data is received or where it is sent is indicated in brackets), the results are the values C _mni for all Mn pairs of antennas enter module 4.2 - summation module by index mn with weight coefficient w _αmni , getting the value of c _i proportionally. estimated deviation δ _ri . 4.3 - the summation modulus for the frequency of the real elements of the VSPM for each pair of antennas multiplied by h _k ,

and e _kmnij . 4.4 - summation module over the index mn with a weighting factor w ² _αmni - obtaining the normalizing value g _i . 4.5 - deviation estimation module δ _αij by normalizing ci to g _i . 4.6 - small averaging of the deviation estimate for i. 4.7 - estimation of the direction by correcting the reference point by the amount of deviation.

, объема I каждая, с основным осреднением оценок за заданное время Т_н (J циклов малого осреднения) и определением их производных для прогноза ОТ на следующий шаг.Block 5 (programmable computer) performs dynamic smoothing of low-average coordinate estimates

, volume I each, with the main averaging of estimates for a given time T _n (J cycles of small averaging) and determining their derivatives for predicting OT for the next step.

Block 6 (programmable computer, controller) provides computational operations for all blocks of inter-antenna space-time processing,

Using the considered device, the proposed method is performed as follows. The observer from the control panel assigns a sector of angles and an interval of observation ranges and monitors the appearance of signal marks (SO) of radiation sources on the review indicator. Adjusts the pre-processing parameters, if necessary: gain, frequency range boundaries, sampling rate, FFT sampling parameters - _Ndft , parameters of small and dynamic averaging J, I, etc. Having chosen the CO for determining the exact coordinates and tracking, assigns the lower bound of the range interval D _fn , bearing in mind that the smaller D _fn , the sharper the main maximum of the response and, therefore, the steeper the DH and the higher the accuracy of determining the range, in addition, in the process of observation, the distance to the source should not become less than D _fn , because in the area D _S <D _fn some calculations are incorrect (negative number under root).

и назначаются первой опорной точкой (ОТ):

, после чего включаются в действие все блоки обработки согласно схемам рисунков фиг. 3 и фиг. 4.On a signal from the remote control, the determination of the exact coordinates and tracking begins. On the scales of the indicator, the sighting device reads the coordinates of CO

and are assigned to the first reference point (FROM):

, after which all processing units are activated according to the diagrams of the drawings of FIG. 3 and FIG. 4.

, для удобства и наглядности отнесенные к СКО флуктуаций этих значений σ_z(α=α_оп,D=D_оп) при некотором заданном отношении сигнал/помеха на антеннах ПС. Конкретное значение уровня помехи в нашем рассмотрении не существенно. Методика таких расчетов описана в работе [Гампер Л.Е. Критерии оценки дальности действия пассивной гидролокации с разнесенными антеннами // НТС «Гидроакустика», вып. 11. СПб, 2010. С. 50-60.]. Она удобна тем, что значение графика z является по существу величиной выходного отношения сигнал/помеха дискриминации при рассогласовании (D-D_оп). Крутизна квазилинейного, рабочего участка ДХ

определяет точность измерения дальности пассивным гидролокатором. На графиках фиг. 5а видно, что при всех дальностях ДХ не антисимметрична и ее крутизна в точке D=D_оп значительно уменьшается с увеличением фактической дальности до источника излучения. На графиках фиг. 5б крутизна ДХ в точке D=D_оп практически постоянна, а ее разброс (менее 0.25%) определяется незначительным расхождением на краях рабочего квазилинейного участка. Величины отношений крутизны ДХ при предлагаемом, нелинейном η_нс и известном способе η_трс, на дальностях до источника излучения от 1.0 до 20.0 км, приведены ниже:It is most convenient to show the performance and advantages of the proposed method using a numerical example, using graphs of the discriminatory characteristics of a given receiving system, calculated in the range of ranges using the prototype method and the proposed method under equal conditions. For calculations, it is assumed: PS with three antennas, base size B _PS =191.5⋅λ _sr m, where λ _sr is the wavelength at the middle frequency of the reception range, the speed of sound is c=1500 m/s. The range interval of interest to the observer is D=(1.0-20.0) km. The distance to the lower limit of the virtual focusing interval D _fn =r _fn =1.0 km, as shown in Fig. 2. In the graphs of FIG. Figure 5 shows the discriminatory characteristics of the deviation of the range from its reference value D _op , (DCD), measured by the current i-th sample of the FFT, calculated by a known method (prototype, Fig. 5a) and proposed (Fig. 5b) for the actual location of the radiation source at ranges D _S =(1, 2, 3, 5, 10 and 20) km. On the horizontal axis are the values of deviations (DD _op ) from the reference point (D=D _op ) in meters. On the vertical axis of the axis - the average (in time) output values of the discriminators

, for convenience and clarity, referred to the RMS fluctuations of these values σ _z (α=α _op ,D=D _op ) for some given signal-to-noise ratio at the antennas of the PS. The specific value of the noise level in our consideration is not essential. The methodology for such calculations is described in [Gamper L.E. Criteria for assessing the range of passive sonar with diversity antennas // NTS "Hydroacoustics", vol. 11. St. Petersburg, 2010. S. 50-60.]. It is convenient in that the value of the plot z is essentially the value of the output signal-to-noise discrimination discrimination at mismatch (DD _op ). The steepness of the quasi-linear, working section of the DC

determines the accuracy of the passive sonar range measurement. On the graphs of Fig. It can be seen from Fig. 5a that, at all ranges, the DH is not antisymmetric and its slope at the point D=D _op decreases significantly with increasing actual distance to the radiation source. On the graphs of Fig. 5b, the slope of the DC at the point D=D _op is practically constant, and its spread (less than 0.25%) is determined by a slight divergence at the edges of the working quasi-linear section. The values of the ratios of the steepness of the DC with the proposed, non-linear η _ns and the known method η _trs , at distances to the radiation source from 1.0 to 20.0 km, are given below:

The values of the table in the second row are equal to the decrease in the root-mean-square error (RMS) of determining the range by the proposed method σ _Dns compared to the known method σ _Dtrs at distances to the radiation source from 1.0 to 20.0 km. It should be noted that the use of the proposed method is not associated with any deterioration in the energy parameters of signal processing in comparison with the prototype, because both in the first and in the second case, the output discrimination level is equally determined by the sum of the phased pairwise correlations of the PS antenna signals. The parameters and spatial characteristics of the response in directions when using the proposed method do not change.

Claims (2)

1. A method for determining the range to a moving source of radiation by a passive sonar, containing the reception of signals by M antennas, M≥3, pre-processing, including amplification, filtering with a frequency response h _k , synchronous sampling, digital transformation, cyclic discrete transformation (DFT) in the reception band ƒ _n ≤ƒ _k ≤ƒ _v (Fourier transform), survey means in a given observation sector and range interval, determination of the coordinates of the detected signal mark on the survey indicator α _n , D _n , which at the time of the start of work of the means for accurately determining coordinates and tracking are used as reference point (OT) with coordinates D _op (t ₀ )=D _n , α _op (t ₀ )=α _n , and to determine and track the current direction to the moving radiation source

each m-th antenna is phased in the direction from its center to the OT and at each i-th DFT cycle, the complex pairwise mutual power density spectra (CPRMS) of the signals of the antenna pairs of the receiving system are measured, discrimination of the current direction deviations from the reference point α _opi and known methods are used dynamic smoothing of the evaluation sequence

, characterized in that to determine the current range value by a passive sonar on each i-th cycle of the DFT, discrimination of the deviation of its reference value D _opi from the desired true D _Si is used, while applying a non-linear range scale r _f to the virtual focus points of the receiving system R _fi c coordinates <α _opi , r _f >, for this, on a linear scale D, the lower limit of the range interval D _fn is predetermined, in which the radiation source will supposedly move in the process of observation and tracking: D≥D _fn , at each next i-th cycle of the DFT designate in it the reference virtual focusing point of the receiving system R _phi with coordinates <α _opi , r _opi >, so that the direction to R _phi coincides with α _opi , and the distance r _opi is related to D _opi by a non-linear relationship:

, where γ _f - scale factor equal to the square of the lower boundary

, and determine the magnitude and sign of the deviation

the current reference value r _opi from the desired range value to the virtual source r _Si , and the current estimate of the actual range on the i-th DFT cycle is determined as

and subjected to subsequent dynamic averaging by known methods for determining the parameters of successive measurements and predicting their values for the next measurement cycles. 2. The method according to p. 1, characterized in that to determine the magnitude and sign of the deviation

the current reference value r _opi from the desired value of the distance to the virtual source

distance deviation discrimination is used, for which the difference in signal propagation times t _mRi (m, n=1, ..., M) is calculated from the virtual point R _opi to the centers O _m , O _n of each mn-th pair of antennas τ _mnRi =t _mRi -t _nRi , according to the trigonometry formulas, using the speed of sound in the medium c, and also determine the weight coefficients equal to the derivative with respect to r of the propagation time difference τ _mRi (α _opi , r _opi ), i.e.

, sum over the frequency ƒ _k the imaginary parts of the FSPDM of each mn-th pair of antennas in the reception band [ƒ _n , ƒ _in ] with the multiplication of each term by the coefficient of the frequency response h _k , by the frequency ƒ _k and by the coefficient e _kmni =exp(-j2πƒ _{k mn R i _ k are} the real parts of the HSTM with each term multiplied by h _k , by the compensation factor e _kmni and by the square of the frequency

obtaining for each pair the total value G _mni , then sum over the index mn the values C _mnt with weight coefficients w _mnRj , obtaining the value c _i , which is divided by the sum over the index mn values G _mnt with the same weight factor squared

obtaining the value g _i , and determine the deviation

perform averaging of variance estimates

i=1, 2, …, I cycles of the DFT, obtaining a low-average estimate of the deviation

, and the current estimate of the actual range on the i-th DFT cycle is defined as

.

Images