RU2300118C1 - Mode of detection noisy objects in the sea - Google Patents

Abstract

FIELD: the invention refers to hydro acoustics and may be used in passive sonar systems.

SUBSTANCE: the technical result of the invention is increasing of noise immunity of hydro acoustics system of detection? Realization of prescribed time of accumulation and increasing time of maintenance of acoustics contact with a moving target in the sea. The mode includes the following operations. It is necessary to receive the primary field of noisy radiation of objects with static sheaf of characteristics of direction in the horizontal plane, execute frequency-time processing in each space observation channel, frame, average on time, center and standardize signals to the disturbance, execute observation on the current cycle of survey of received standardized signals and take decision about detection by way of comparison with the threshold value of ratio signal-disturbance. In the process of receiving tracking of energetic parameters of the noisy signal on level, dispersion of the level and selection of local false maximums of noisy signals in the static sheaf of space channels is executed.

EFFECT: increases noise immunity of the passive sonar system of detection.

2 dwg

Description

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

A known method of detecting a point source using a correlator of antenna array signals with a multipath directivity. With this treatment, the transducers of the antenna array are divided into two groups, each of which introduces a phase shift and linear weighting of the output signals of the transducers. The output signals of the two groups correlate, and the results are summarized in the taps of the delay line, squared and recorded in the coordinates of the bearing - time. Using the target position lines observed on the output image in a multi-beam diagram system, one can follow the change in the target situation; to obtain estimates of the ranges and bearings of various targets, a joint analysis of the movement of targets and data from the navigation system of one of its ships can be carried out. In addition, one can observe the appearance and disappearance of certain goals, as well as situations corresponding to false alarm (Application of digital signal processing, trans. From English under the editorship of E. Openheim, M., Mir, 1980, pp. 477, 478 )

One of the main disadvantages of this method is that for small signal-to-noise ratios at the outputs of both gratings, the multiplication operation in calculating the correlation changes some threshold characteristics of the system. The operation of multiplication can also cause the appearance of false side lobes, which, in essence, are spatial analogues of the results of modulation of signals.

Closest to the technical nature of the proposed method is the detection method described in L. Kamp's monograph "Underwater Acoustics", trans. from English Moscow, Mir, 1972, pp. 262 ... 263, according to which, using an antenna and an energy receiver of a signal, it is possible to detect a target in the passive mode by comparing the signal-to-noise ratio in the acoustic illumination zone with a threshold value.

An audio signal is received (in fact, it means receiving a mixture of a noise signal and an interference) by an antenna in a system with a multipath diagram. The broadband signal from the fan output of the directivity characteristics of the horizontal spatial channels of the receiving antenna is fed to the power receiver with a given averaging time and a panoramic indicator with a recording device or with a cathode ray tube, and beam brightness modulation is used. Hereinafter, the term “fan of spatial channels” used is generalizing with respect to the classical term “multi-beam radiation pattern” (see Application of Digital Signal Processing, trans. From English under the editorship of E. Openheim, M., Mir, 1980. , p. 467, 468), when the set of directivity characteristics of adjacent spatial channels is fan-shaped with a step-discrete in angle.

Each scan of the fan outputs on one review cycle is reproduced on the indicator as a line modulated in brightness. Subsequent sweeps in the review cycles shift downward relative to the previous ones. The resolution cell that registers the target in the spatial channel of the fan is associated with a mark that appears at the same distance from the edge of the indicator for each sweep.

With this type of scan, if the hydroacoustic is waiting for the appearance of, for example, 20 bursts of brightness in each review cycle, it can observe the presence of target signals in the form of vertical lines on the indicator. Observation in this case is carried out at a threshold when, for example, half of the resolution time resolving cells give noise marks on the indicator, and the signal plus noise mark appears in the cell of the spatial channel to which it belongs to approximately three quarters of the total time of all review cycles. (L. Kamp, “Underwater Acoustics”, transl. From English. M., Mir, 1972, pp. 265 ... 266).

This detection method contains the following operations:

- receiving a hydroacoustic noise signal using a receiving antenna with a developed aperture in the horizontal plane,

- time-frequency processing of received noise signals for each spatial observation channel in the horizontal plane,

- measurement of the level at the output of the spatial channel of the fan, including the accumulation in time, centering and normalization in units of the signal-to-noise ratio,

- deployment on successive review cycles of the received signals of the spatial channels of the fan in the horizontal plane on a panoramic indicator in the coordinates of the angle - time.

At each given instant, these operations provide observation only in a certain sector of angles in the vertical plane, determined by the width of the directivity. The method works well in a homogeneous medium and in conditions where signal fading as a result of the addition of signal beams after refraction in a layered-inhomogeneous medium and reflections from the boundaries can be neglected. In a real sea, in an acoustic field, inhomogeneous and anisotropic, the algorithm showed unsatisfactory operation, gave a large number of false signals - noise marks and unstable detected weak useful signals. Detection is carried out by comparing noise levels in neighboring directions. The signal accumulation time is determined only by the value set for one sweep cycle. In the dynamics of the input signals with the relative motion of the receiving system, the signal accumulation time is limited to one review cycle.

Thus, it is desirable to have a method for detecting noisy objects, which would simultaneously increase the signal accumulation in time and with greater reliability than in the prototype method to determine the presence of the target signal and maintain acoustic contact with the target for a long time, reducing the time of the signal loss in the resolution cell by angle acoustic contact loss time.

The objective of the proposed method is to increase the noise immunity of the hydroacoustic system by implementing a given accumulation time of the moving target signal relative to the receiving antenna for several viewing cycles and increasing the time of maintaining acoustic contact with the moving target.

To solve this problem, in a known method for detecting objects that are noisy at sea in a fixed frequency range, including receiving the primary noise field of objects in the horizontal plane, in which the time-frequency processing of the received noise signals for each observation channel in the horizontal plane is carried out, they are squared, averaged over time , center and normalize the signals to interference, observe on each cycle of the survey marks of the received noise signals of the entire set of spatial channels x in the horizontal plane and make a decision about its detection by comparing with a threshold signal-to-interference introduced new operations, namely:

at the next review cycle, observation is carried out by two independent sequences of operations:

in the first sequence, the energy parameters of the noise signal are maintained by level, level dispersion, and rejection of local false maxima of noise signals, for which the following operations are performed:

separating noise signals from background noise above a level that is several times lower than the detection threshold,

determine the level of all local signal maxima,

calculate, according to a given approximation law, the refined value of the noise signal level according to several responses of spatial channels in the vicinity of a given local maximum of the signal forming the signal fragment,

determining a signal level shift for the time between the previous and current review cycles and calculating the probability density of the measured signal level shift and the probability density of false alarms for a given accumulation time,

in the second sequence, information parameters of the noise signal are monitored by bearing, bearing dispersion, by speed and acceleration of bearing change, and bearings of local false maxima of noise signals are rejected, for which the following operations are performed:

fix the bearing of the spatial channel, in which each local maximum of the signal is observed,

calculate the specified value of the bearing of the noise signal according to several responses of the spatial channels in the vicinity of this local maximum of the signal forming the signal fragment according to a given approximation law,

make up from the totality of the bearing estimation and the magnitude of the bearing change (VIP) the vector of motion parameters of the local signal maxima,

calculate the matrix of cross-correlation functions for the vector of motion parameters of the local maximums of the signal,

calculate the rate of change of VIP and bearing for the time between the previous and current review cycles,

as a result, predictive estimates of the bearing and VIP of the local maximums of the signal for the time between the previous and current review cycles for a given accumulation time are determined,

determine the variance of the predictive bearing estimate for a given accumulation time and calculate the strobe width from the bearing, within which each signal is monitored,

calculate the probability density of the displacement of the measured bearing for a given accumulation time,

the results of two sequences of operations calculate the total weight of the local maximums of the signal,

comparing the generalized weight of the local signal maxima with the signal detection threshold, which corresponds to the threshold signal-to-noise ratio.

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

Figure 1 shows a block diagram of the proposed method as a sequence of operations. Operation 1 provides for the formation of a horizontal static fan of directivity characteristics and the simultaneous reception of the primary field of the mixture of the noise signal and interference with each spatial channel. Operation 2 provides frequency processing of the received signal of each spatial channel.

The meaning of operations 3 and 4 is determined by their names.

The set of operations 5 ... 8 provides tracking of the energy parameters of the noise signal by level, level dispersion and rejection of local false maxima of noise signals.

Operation 5 provides for the separation of the primary marks of noise signals from background noise using a special threshold for a given signal-to-noise ratio (SIR), for example, within 0.5 ... 3.

The set of operations 6 ... 8 provides the determination of the level of local maximums of a signal, the calculation of the specified level of a noise signal according to several responses of spatial channels in the vicinity of the local maximum of the signal forming the signal fragment, and the determination of the shift of the signal level during the review cycle. The selection of primary marks is based on the identification of local maxima that exceed the specified special threshold and the refinement of the signal level, defined as the position and magnitude of the maximum for a parabola drawn through three points in the vicinity of the local maximum.

Operation 6 is carried out by smoothing the level (in OSB fractions) of the label U _{M of the} local signal maxima according to the formula U _et = U _M / α ₀ + (1-1 / α ₀ ) U _et-1 , where U _et , U _et-1 - smoothed estimates of the current noise signal level at time t, t _1, respectively; α ₀ - the value of the smoothing window; U _et = α ₁ U _et-1 + (1-α ₁ ) U _M , α ₁ - constant smoothing the level of the SIR. The smoothing constants can be selected and depend on a priori or experimental data on the stationarity of the behavior of the signal level in a layered inhomogeneous marine environment.

Operation 7 is carried out by parabolic approximation of the level of the SIR by several, for example, three, points in the neighborhood of the label of the local maximum of the signal u ₁ , u ₂ , u ₃ , according to the formula U _a (φ) = u ₁ + φu ₂ + φ ² u ₃ , U ₀ (φ ₀ ) = maxU _a (φ), where U = U ₀ is the adjusted level value, φ is the bearing corresponding to the number of the spatial channel in which the local maximum of the signal is observed.

при выбранном нормировании.Operation 8 involves determining the signal level offset U _et for the review cycle and calculating the probability density p (U ₀ -U _M ) of the measured signal level offset in fractions of the probability density p (U _M ) of the false alarm level. Thus, an additional selection of labels is carried out according to the probability density of false alarms for a given accumulation time. When recording explicitly the distribution densities, it is assumed: the signal fluctuations are completely determined by the interference fluctuations. The distribution law of the maximum values is in good agreement with the experimentally obtained one during modeling and is approximated by the normal law with zero mean and unit dispersion

with selected rationing.

The set of operations 9 ... 12 provides support for the information parameters of the noise signal from the bearing, the dispersion of the bearing, the speed and acceleration of changes in the bearing, and the rejection of bearings of the local false maxima of noise signals.

In operation 9, the bearing φ is fixed corresponding to the number of the spatial channel in which the local maximum of the signal is observed.

Operation 10 involves calculating, according to a given law, the approximation of the updated value of the bearing of the noise signal according to several responses of spatial channels in the vicinity of a given local maximum of the signal forming the signal fragment. The calculation is carried out by parabolic approximation of the level of the SIR by several, for example, three, points in the vicinity of the label of the local maximum of the signal u ₁ , u ₂ , u ₃ , according to the formula U _a (φ) = u ₁ + φu ₂ + φ ² u ₃ , U ₀ = (φ ₀ ) = maxU _a (φ), φ = φ ₀ is the adjusted bearing value.

Operation 11 provides for the determination of forecast estimates of the bearing and the angular velocity of the bearing (VIP) of the local signal maximums during the time between the previous and current review cycles for a given accumulation time, for which:

вектор параметров движения локальных максимумов сигнала

,make up from the combination of the assessment of the bearing φ and VIP

vector of motion parameters of local signal maxima

,

для вектора параметров движения локальных максимумов сигнала

,calculate the matrix of cross-correlation functions

for the vector of motion parameters of local signal maxima

,

и пеленга

за время между последовательными циклами обзора,calculate the rate of change of VIP

and bearing

during the time between successive review cycles,

для заданного времени накопления и вычисляют ширину строба по пеленгу ±δ, в пределах которого осуществляется наблюдение каждого сигнала.determine the variance of the forecast estimate of the bearing

for a given accumulation time and calculate the strobe width from the bearing ± δ, within which each signal is observed.

,

,

, на текущем цикле обзора осуществляется по формулам Калмановской фильтрации либо по упрощенным формуламForecast information parameters φ _i ,

,

,

, in the current review cycle is carried out according to the Kalman filtering formulas or according to simplified formulas

where dt is the time between adjacent cycles.

на заданный коэффициент запаса и проверка величины дисперсии пеленга

, которая не должна быть ниже заданной, в противном случае матрица ковариаций заменяется на начальную.Multiplication of all elements of the covariance matrix is performed.

for a given safety factor and checking the bearing dispersion

, which should not be lower than the given one, otherwise the covariance matrix is replaced by the initial one.

Operation 12 - calculating the probability density of displacement for the review cycle of the measured bearing p (φ ₀ -φ _M ) for a given accumulation time according to the formula

Operation 13

The generalized weight of the local maximums of the signal is calculated by the formulas:

where N _b is the number of responses of spatial channels in the vicinity of local signal maxima, POR _MW .

The detection flag is established if the total weight of the local maximums of the signal Wr _mp exceeds the corresponding threshold for detecting POR _MW , and the probability of false alarms p (U _M ) is lower than the corresponding threshold for POR _VLT .

The detection method is implemented by a device - a noise-finding station with a spatial signal processing system - UFKH (see Ship hydroacoustic equipment, Koryakin Yu.A., Smirnov S.A., Yakovlev G.N., St. Petersburg, Nauka, 2004, p. 173, Fig. 2.5).

The block diagram of the device is shown in figure 2. The device for detecting noisy objects in the sea of figure 2 consists of a multi-element, for example, a cylindrical antenna 1, the elements of which are connected to a pre-processing system, then to a spatial processing system 2, a primary information processing system 5 and a secondary information processing system 9.

The proposed method is carried out using a receiving system as follows. Noise signals are received by antenna 1. From the output of antenna 1, the signals are transmitted to block 2 for forming spatial observation channels — static fans of directivity in the horizontal plane. Further, through the blocks 6 of the range filters, detection and accumulation (averaging), the signals enter the blocks 7 of the signal processing.

On the indicator device 10, a panorama of the set of received signals at the output of the receiving system, for which the above procedures are performed, is recorded.

Claims (1)

A method for detecting objects that are noisy at sea in a fixed frequency range, including receiving the primary noise field of objects in the horizontal plane, in which the time-frequency processing of the received noise signals for each observation channel in the horizontal plane is carried out, they are squared, averaged over time, centered and normalized to interference, carry out observation on each review cycle of the marks of the received noise signals of the entire set of spatial channels in the horizontal plane and mayutsya decision about its detection by comparing with a threshold signal-to-noise ratio, characterized in that at the next review cycle, observation is carried out by two independent sequences of operations: the first sequence is the operation of tracking the energy parameters of the noise signal by level, level dispersion, and rejecting local false maxima of noise signals, for which separating noise signals from background noise above a level that is several times lower than the detection threshold, determine the level of all local signal maxima, calculate, according to a given approximation law, the refined value of the noise signal level according to several responses of spatial channels in the vicinity of a given local maximum of the signal forming the signal fragment, determining a signal level shift for the time between the previous and current review cycles and calculating the probability density of the measured signal level shift and the probability density of false alarms for a given accumulation time, the second sequence is the operation of tracking the information parameters of the noise signal in the bearing, bearing dispersion, in speed and acceleration of the bearing change and rejection of bearings of local false maxima of noise signals, for which fix the bearing of the spatial channel, in which each local maximum of the signal is observed, calculate the specified value of the bearing of the noise signal according to several responses of the spatial channels in the vicinity of this local maximum of the signal forming the signal fragment according to a given approximation law, make up from the totality of the bearing estimation and the magnitude of the bearing change (VIP) the vector of motion parameters of the local signal maxima, calculate the matrix of cross-correlation functions for the vector of motion parameters of the local maximums of the signal, calculate the rate of change of VIP and bearing for the time between the previous and current review cycles, determine the forecast estimates of the bearing and VIP local maximums of the signal for the time between the previous and current review cycles for a given accumulation time, determine the variance of the predictive bearing estimate for a given accumulation time and calculate the strobe width from the bearing, within which each signal is monitored, calculate the probability density of the displacement of the measured bearing for a given accumulation time, the results of two sequences of operations calculate the total weight of the local maximums of the signal, comparing the generalized weight of the local signal maxima with the signal detection threshold, which corresponds to the threshold signal-to-noise ratio.

Images