RU2287840C1 - Method of detection and classification of surface targets - Google Patents

Abstract

FIELD: radar technique; methods of detection of signals from surface targets and selection of signal from false targets - sources of concentrated passive jamming; coherent-pulse radar stations.

SUBSTANCE: proposed method is based on radiation of coherent radio pulses at constant carrier frequency, reception and processing of reflected signals; additionally targets are selected from targets received at second selection stage and considered to be true; frequency of outgoing pulses is changed from burst to burst; carrier frequency of received signals is reduced; power of frequency components is accumulated and decision is taken on class of target observed.

EFFECT: enhanced efficiency in classification of true targets and their identity to class of targets.

2 cl, 9 dwg

Description

The invention relates to radar technology, mainly to methods for detecting signals from surface targets with the selection of signals from false targets - sources of concentrated passive interference - and can be used in coherent-pulse radar stations installed on high-speed carriers, in particular aircraft of various destination, including for aerospace reconnaissance of ships.

Currently, in coherent-pulse radars with filtering the received signals by Doppler frequency to detect the signal from the true target in the presence of reflections from false targets - sources of passive interference, a difference in the Doppler frequencies of signals from the true target and interference sources is used, which is a consequence of the difference in their radial velocities [one].

The informative feature used to classify targets in these technical solutions is the average Doppler frequency | F _d | ≥0.

The disadvantage of this method is the erroneous classification (selection) in cases where the true target is stationary or moving in a direction perpendicular to the direction of radiation propagation - in these cases, the frequency F _d = 0, and also when the source of passive interference (false target) moves, so the frequency F _{d of the} echo from the source of interference is non-zero.

There is another method, considered in [2], which is based on the emission of coherent radio pulses with a constant carrier frequency, the reception of reflected signals, filtering the received at each resolution element by the range of the signals at the Doppler frequency on M equally spaced frequency components, measuring their powers, comparing the measurement results with a threshold level, deciding on the presence of a signal from a true target, if the powers of at least one but not more than k adjacent frequency components exceed the threshold level. The disadvantage of this method is the inability to select true targets from single corner reflectors (UO). The reason for this phenomenon, as is known, is that, in contrast to dipole clouds (DOs), as well as from UO bundles connected by flexible files (cables), the inter-period fluctuation spectrum of signals from single UOs, as well as signals from surface ships (NK) ), is narrowband because the signals reflected from those and other fluctuate slowly [3], so their selection in this way is impossible.

Closest to the technical nature of the analogue adopted as a prototype of the proposed method, is a method for selecting surface targets [4].

The prototype method is based on two-stage selection, namely, after the first stage, which consists in emitting coherent radio pulses with a constant carrier frequency, receiving reflected signals, filtering the received signals: by the Doppler frequency on M equally spaced frequency components in each resolution element in range, measuring the frequency power components, comparing the measurement results with a threshold level, making a preliminary decision about the presence of a signal from a true target, if the power of at least one about no more than k adjacent frequency components exceeds the threshold level, where k is the ratio of the maximum expected width of the spectrum of inter-period fluctuations of the signals from the true target to the bandwidth of one frequency component, go to the second stage, which consists in the emission of coherent radio pulses during n repetition periods varying from pulse to pulse carrier frequency f _i ,

i = 1, 2, ..., m, and

S_ir, r∈1, 2,..., М, вычислении нормированного значения S_H по формулеwhere c is the speed of light, l is the minimum in view of the expected length of the true target in the radial direction, receiving reflected signals from targets accepted as true at the first stage of selection, filtering the signals at each of m carrier frequencies by Doppler frequency on M equally spaced frequency components, measuring their powers S _ir (r = 1, 2, ..., M), finding the maximum power values of the frequency components S _i =

S _ir , r∈1, 2, ..., M, calculating the normalized value of S _H by the formula

where S _{m + 1} = S ₁ ,

comparing the result with the second threshold level and deciding on the presence of a signal from the true target when the threshold level is exceeded.

The disadvantage of the prototype method is the inability to classify true goals, i.e. making decisions on the belonging of a surface ship to a certain class (in particular, the class of "main" targets).

The objective of the invention is to enable the classification of true goals and determine their belonging to a particular class of goals.

The solution to this problem is based on the assessment of informative features (longitudinal and transverse dimensions, radial speed and architecture) of a surface ship by its radar image obtained by synthesizing an aperture during anterolateral viewing with coherent processing, and deciding on the class of the surface ship by comparing the value of the statistical decisive function informative signs with a threshold level.

где с - скорость света, l - минимальная по ракурсу ожидаемая протяженность истинной цели в радиальном направлении, приеме отраженных сигналов от целей, принятых за истинные на первом этапе селекции, фильтрации сигналов на каждой из m несущих частот по допплеровской частоте на М равноотстоящих частотных составляющих, измерении их мощностей S_ir (r=1, 2,..., М), нахождении максимальных значений мощностей частотных составляющих

, вычислении нормированного значения S_H по формулеThe essence of the proposed method for detecting and classifying surface targets is that in the known method of selection based on the emission of coherent radio pulses with a constant carrier frequency, receiving reflected signals, filtering the received signals by Doppler frequency on M equally spaced frequency components in each range resolution element, measuring the power of the frequency components, comparing the measurement results with a threshold level, deciding on the presence of a signal from a true target, if at least one, but no more than k adjacent frequency components exceed the first threshold level, where k is the ratio of the maximum expected width of the spectrum of inter-period fluctuations of signals from the true target to the bandwidth of one frequency component, radiation at the second stage of coherent radio pulses for n repetition periods with a carrier frequency f _i , i = 1, 2, ..., m, varying from pulse to pulse, and

where c is the speed of light, l is the minimum expected angle of the true target in the radial direction, receiving reflected signals from targets taken as true at the first stage of selection, filtering the signals at each of m carrier frequencies by Doppler frequency on M equally spaced frequency components, measuring their powers S _ir (r = 1, 2, ..., M), finding the maximum powers of the frequency components

calculating the normalized value of S _H according to the formula

where S _{m + 1} = S ₁ ,

comparing the result with the second threshold level and deciding on the presence of a signal from the true target when the second threshold level is exceeded, in addition from the goals accepted at the second stage of selection as true, choose targets located in the directions β _j relative to the direction of the carrier velocity vector V , and satisfying the condition

where ψ _0.5 is the width of the antenna radiation pattern at the level of -3 dB in the plane passing through the vector V and the direction to the jth target, and for their classification the parameters M, m and n are selected from the relations:

n = m

where R _j is the slant range to the j-th target,

F _p - the repetition frequency of the probe pulses,

V is the speed of the radar carrier,

change the carrier frequency f _i (i = 1, 2, ..., m) of the probe pulses from a packet to a packet of M pulses, keeping it constant inside the packets, reduce the carrier frequency of the received signals by F _{to the} Doppler frequency corresponding to the projection of the velocity V radar carrier in the direction β _j to the j-th target, i.e.:

carry out the accumulation of power frequency components S _irq (r = 1, 2, ..., M) in each resolution element in range q _min <q <q _{max of} received m ₁ <m bursts of pulses taking into account the predicted change in the coordinates of the j-th target for accumulation time due to the movement of the radar carrier, form signals:

where q _min , q _{max are} determined by the boundaries of the range of the range when observing the j-th target, find the coordinates r ₁ (q), r ₂ (q) of the contour of the radar image of the j-target from the conditions:

and boundary - the minimum q ₁ and maximum q ₂ q values under which these conditions are met,

where C ₀ ≥0 is the threshold level determined by the admissible probability of false response due to noise and interference,

determine the angle view α _{j of the} j-th target, the maximum longitudinal L and transverse In its size from the relations:

according to the formula:

Where

r (q) = aq + d, parameters a. d is selected from the condition

Z _cf = 0.5 [Z ₁ (q) + Z ₂ (q)],

ΔR, ΔX are the dimensions of the resolution element in the radial (in range) and transverse directions, respectively, verify that the condition q ₂ -q ₁ ≥r (q ₂ ) -r (q ₁ ) is satisfied, and if it is not satisfied, repeat the operation from finding the coordinates contour to target size estimates after permutations:

make a decision about the class of the observed goal "main" if:

where L ₁ , L ₂ , B ₁ , B ₂ - a priori known values of the length and width of the radar images of the "main" and "not main" targets, respectively,

σ _L , σ _B are the corresponding a priori known average standard deviations,

A ₀ is a threshold value selected depending on the desired relationship between the probabilities (and costs) of classification errors.

где F₀ - пороговое значение, определяют координаты носа (или кормы) наблюдаемой j-ой цели по правилам вида:In addition, for classification, the coordinates of the most reflecting points of the target in the coordinate system associated with the target can be additionally used as informative signs, for which the Doppler frequency F _{dj of the} reflected signals arriving in the direction B _{j of} the antenna pattern axis relative to the direction of the carrier velocity vector V is measured Radar at

where F ₀ is the threshold value, determine the coordinates of the bow (or stern) of the observed j-th target according to the rules of the form:

for F _dj > 0: q _n = q ₁ , r _H = a · q ₁ + d (nose),

for F _dj <0: q _H = q ₂ , r _H = a · q ₂ + d (feed),

compare the accumulated values of the signals z (r, q) and determine the coordinates r * _i , q * _j ., the maximum values i = 1, 2, ..., u, u> 0, find the distances of these points from the nose according to the rules :

- в продольном и поперечном направлениях соответственно, и принимают решение о классе наблюдаемой цели по правилу:

- in the longitudinal and transverse directions, respectively, and decide on the class of the observed target according to the rule:

if

then the goal is "main",

where l _01i , l _02i , b _01i , b _02i are the a priori known distances averaged inside the classes of the most reflecting points from the bow of the ship in the longitudinal and transverse directions for the classes of “main” and not “main” targets, respectively,

σ _li , σ _bi are the corresponding averaged standard deviations,

And ₁ is a threshold value selected depending on the desired ratio between the probabilities (and costs) of classification errors.

We give the necessary explanations.

According to the proposed method, radar observation of targets accepted as true after selection and located in the front hemisphere in the directions β _j relative to the direction of the carrier velocity vector V satisfying the condition

with coherent processing of the reflected signals over a time t _k in the anterolateral view (TSP) mode [5]. Defining the probing signal in the form

where a (t), φ (t) are the amplitude and phase modulation functions, respectively, we obtain for the signal reflected from the target element at the beginning of the observation at an angle β where β _j - 0.5ψ _0.5 ≤β≤β _j + 0.5ψ _0.5 relative to the direction of the vector V, the expression (see Fig.8):

where R (t} is the range to the target element in question,

t is the time counted from the start of irradiation.

Next, we have

относительно 1:Using the expansion in Newton's series, we obtain from (2) up to terms of order

regarding 1:

Substituting (3) in (1) we obtain the expressions for the phase incursion and the instantaneous circular frequency of the received signals in the form:

In expressions (2) - (4), the target’s own speed is not taken into account, i.e. the target is assumed to be stationary.

As can be seen from (4), the motion of the radar carrier with speed V leads, firstly, to a linear phase incursion Δψ ₁ (t), which corresponds to a frequency shift in the spectrum of the reflected signal by the Doppler frequency:

- длина волны излучаемых сигналов,Where

- wavelength of the emitted signals,

secondly, to the phase incursion with a quadratic time dependence:

depending also on the range to the target.

Restricting ourselves to the case of the so-called unfocused survey [5], we accept the condition of permissible phase distortions:

then for the limiting value of the coherent processing time we get:

whence, in turn, we have an expression for the number of coherently processed pulses when observing the jth target:

further, the bandwidth of one processing channel by Doppler frequency is expressed by the formula

whence it follows that the number M of processing channels equal to the number of spectral components of the signal

coincides with the number of coherently processed pulses and is determined by expression (8).

We obtain further the expression for the angular and linear resolutions in the transverse direction. Comparing (9) with the relation

following from (5), we obtain for equivalent angular resolution

so the corresponding linear resolution is

further since the time of radar contact with the jth is

получим верхнее предельное значение для m.then taking into account the relations n≤F _p T _j and

we get the upper limit value for m.

, которое, как правило, выполняется для надводных целей, сводится к замене Vcosβ в выражениях (4) на Vcosβ+ V_jcos(β-α_j), где α_j - ракурс j-ой цели. Действительно, при α_j=0, когда цель движется навстречу РЛС, доплеровская частота сигналов увеличивается, а при α_j=π - наоборот, уменьшается.Accounting for the target’s own speed V _j under the condition

, which, as a rule, is performed for surface purposes, reduces to replacing Vcosβ in expressions (4) with Vcosβ + V _j cos (β-α _j ), where α _j is the angle of the j-th target. Indeed, when α _j = 0, when the target moves towards the radar, the Doppler frequency of the signals increases, and when α _j = π, on the contrary, it decreases.

повышают точность оценок по приводимым формулам в ситуации, когда радиальный размер цели оказывается меньше поперечного (фиг.9, цель 2).The rules for finding the coordinates of the contour of the radar image of the target and size estimates follow from the geometric relationships (Fig.9) and do not require explanation. Permutations

increase the accuracy of estimates by the given formulas in a situation where the radial size of the target is less than the transverse (Fig. 9, target 2).

The classification rule is obtained as follows.

Consider the case of binary classification of objects according to the set of informative features γ ₁ , γ ₂ , ..., γ _n , the values of which will be assumed to be statistically independent, normally distributed with parameters γ ₁₁ , γ ₂₁ , ... γ _n1 , σ ₁ , σ ₂ , ..., σ _n and γ _l2 , γ ₂₂ ..., γ _n2 σ ₁ , σ ₂ , ... σ _n - for the first and second classes, respectively. Then composing the likelihood ratio with respect to the totality of informative features

after its logarithm and simple transformations, we obtain the following decision rule:

if

then the object belongs to class 1, otherwise - to class 2.

The threshold value A of the statistical decision function is selected depending on the desired relationship between the probabilities of classification errors (and their costs). In the special case important for practice, when these errors (erroneous classification of an object of class 1 as an object of class 2 and vice versa) are equivalent, the expression

их максимальных значений, определяют их координаты сначала в исходной системе, связанной с носителем, а затем, с помощью известных правил преобразования координат, расстояния соответствующих элементов цели от ее носа в продольном и поперечном направлениях.To assess the direction of movement of the target, it is proposed to measure the Doppler frequency of the reflected signals arriving in the direction β _{j of the} axis of the antenna pattern, which, in turn, is determined with possible accuracy by the maximum antenna pattern of the surveillance radar, or by zero of the total-difference diagram in the case of single-pulse The radar, however, as already mentioned, pre-take into account the Doppler frequency corresponding to the own radial velocity of the radar carrier in the same direction. Further, taking into account that surface ships usually move with their bow in the direction of movement, they determine the coordinates of the bow (or stern), and find a given amount by comparing the signals of the accumulated values of the power of the frequency components

their maximum values, determine their coordinates first in the original system associated with the carrier, and then, using the known rules for the transformation of coordinates, the distance of the corresponding elements of the target from its nose in the longitudinal and transverse directions.

The invention is also illustrated by a further description and drawings of a device that implements the proposed method for the detection and classification of surface targets, which are:

figure 1 - structural diagram of a radar that implements the inventive method,

figure 2 is a structural diagram of a frequency adjustment unit,

figure 3 is a structural diagram of the Converter "code-frequency",

4 is a structural diagram of a double balanced modulator,

5 is a diagram of a program of an information processing device,

6 - waveforms of the signals at the outputs of the synchronizer,

Fig.7 is a structural diagram of a synchronizer,

Fig is a diagram of the anterolateral view,

Fig.9 is a diagram of a radar image of the ship.

In figure 1, the following notation:

1 - antenna

2 - antenna switch,

3 - transmitter,

4 - frequency adjustment unit,

5 - synchronizer,

6 - receiver

7 - phase detector

8 - block video amplifiers,

9 - range gating unit,

10 - Doppler frequency calculator,

11 - Converter "code-frequency"

12 - double balanced modulator,

13 - switch

14 ₁ ,, 14 _m - shapers of the complex envelope,

15 ₁ , ..., 15 _m - Doppler frequency filters,

16 ₁ , .... 16 _m - detectors of frequency components,

17 - information processing device.

In the diagram of FIG. 1, a synchronizer 5, a frequency tuner 4, a transmitter 3, an antenna switch 2 and an antenna 1 are connected in series. The signal input of a receiver 6 is connected to the third arm of the antenna switch 2, the heterodyne input of which is connected to the output of the oscillator oscillations of the transmitter, and the output the receiver 6 is connected to the signal input of the phase detector 7, at the outputs of which the quadrature components of the received signals are formed. The outputs of the phase detector 7 are connected in series to a block 8 of video amplifiers, a block 9 of range gating, and a switch 13.

The output of the Doppler frequency calculator 10 is connected to the first input of the code-frequency converter 11 connected to the first input of the double balanced modulator 12, the second input of which is connected to the oscillation output of the reference frequency of the transmitter 3, and the output is connected to the reference frequency input of the phase detector 7.

The outputs from the first to 2 m-th switch 13 are connected in pairs to the corresponding inputs of identical frequency channels from the first to m- _th , in which the processing of pulse sequences with a carrier frequency f _i , i = 1, 2, ..., m. Each of the frequency channels contains a shaper 14 ₁ (14 ₂ ,, 14 _m ) of a complex envelope and a multi-channel filter 15 ₁ (15 ₂ ,, 15 _m ) of Doppler frequency, in which the received signals are filtered by the Doppler frequency on M equally spaced frequency components . The outputs of the multi-channel filters 15 ₁ , ..., 15 _{m are} channel-connected to the inputs of the multi-channel detectors 16 ₁ , ..., 16 _{m of} frequency components, the outputs of which are channel-connected to the corresponding signal inputs (fifth input) of the information processing device 17.

The first output of the information processing device 17 is connected to the control input of the antenna 1, its second output is connected to the control input of the frequency tuning unit 4, and the third output is connected to the input of the synchronizer 5.

The first and second information outputs of the antenna 1, according to the azimuth and elevation angle signals, are connected respectively to the first and second inputs of the information processing device 17 and the Doppler frequency calculator 10, to the third input of which the carrier speed values from the on-board navigation system are received.

To the second output of the synchronizer 5 is connected to the control input of the transmitter 3 and the third input of the information processing device 17. The third inputs of the synchronizer 5 are connected to the control inputs of the range gating unit 9 and the information processing device 17, and the fourth is the input of clock pulses of the (second) code-to-frequency converter 11.

The first output of the frequency adjustment unit 4 is also connected to the control input of the switch 13, and its second output is connected to the fourth inputs of the Doppler frequency calculator 10 and the information processing device 17.

Figure 2 presents the structural diagram of the block 4 frequency adjustment, where the following notation:

18 is a noise voltage generator,

19 - amplifier limiter

20 is a counter

21 - the first element And

22 - register

23 - decoder,

24 - pulse frequency divider,

25 - the second element And

26 is a key block,

27 is a delay element.

In the diagram of FIG. 2, a noise voltage generator 18, a limiter amplifier 19, a counter 20, a first AND element 21, a register 22 and a decoder 23 are connected in series.

The second input of the first element And 21 is connected to the output of the key unit 26, the second input of which forms the second (control) input of the frequency adjustment unit 4, the first (synchronizing) input of which is connected via the pulse frequency divider 24 to the first input of the second element And 25 and through the element 27 delays - to his second entrance. The output of the second element And 25 is connected to the input of the key block 26.

The output of the decoder 23 forms the first output of the frequency tuning unit 4, and the output of the register 22 forms its second output.

Figure 3 presents the structural diagram of the Converter 11 code-frequency, where the following notation:

23 - decoder,

29 - controlled divider

30 - low pass filter,

In the diagram of FIG. 3, a decoder 28, a controlled divider 29, and a low-pass filter 30 are connected in series. The first and second inputs of the code-frequency converter 11 are formed respectively by the input of the decoder 28 and the second input of the controlled divider 29, and the output of the low-pass filter 30 forms the output of the converter 11.

Figure 4 presents the structural diagram of a double balanced modulator (DBM) 12, where the following notation:

31, 33 - phase shifters at an angle

32, 34 - multipliers,

35 - adder.

In the diagram of Fig. 4, the first input of the double balanced modulator 12 is connected directly to the first input of the multiplier 32, and through the first phase shifter 31 to the first input of the multiplier 34, the second input of the DBM 12 is connected directly to the second input of the second multiplier 34, and to the second input of the first the multiplier 32 through the phase shifter 33. The outputs of the multipliers 32 and 34 are connected to the corresponding inputs of the adder 35, the output of which is the output of the DBM 12.

Figure 5 presents the program diagram of the device 17 information processing, where indicated:

36 - command to block 4 frequency adjustment - work at a constant frequency f _i = const,

37 - selection of dipole clouds according to the rule:

то цель - дипольное облако, где

граничные частоты спектра в q-м канале дальности, в котором обнаружена цель,if

then the goal is a dipole cloud, where

the boundary frequencies of the spectrum in the q-th range channel in which the target is detected,

38 - command to block 4 frequency adjustment - work with frequency adjustment,

39 - selection of corner reflectors according to the rule:

, то j-я цель - истинная, здесьif

, then the j-th goal is true, here

where S _{m + 1} = S ₁ ,

40 - verification of the condition

41 - accumulation of power frequency components m _i <m packets of reflected signals with the formation of signals

42 - finding the coordinates of the contour of the j-th target on the plane (r, q),

43 - determination of the parameters a, d of the line r _j (q) = aq + d - the longitudinal axis of the radar image of the NK,

44 - finding the values of informative features of the j-th target, in particular the length L _j and the width B _{j of the} radar image,

45 - verification of the fulfillment of the condition q ₁ -q ₂ ≥r (q ₂ ) -r (q ₁ ), if "no" - 46 - circular permutation and repetition of operations 39-41;

47 - making a decision about the class of the goal - the main or not the main ("yes", "no");

48 - verification of the fulfillment of condition j <N, if “yes”, transition to the next goal, if “no” - “end of work”, where N is the number of goals accepted as true and satisfying the condition of clause 40.

Figure 6 presents the waveforms of the signals at the outputs of the synchronizer 5. Here, the following notation:

49 is an oscillogram of control pulses of frequency tuning unit 4 with a period T _p , with lead time t ₁ ≪ T _p , relative to the beginning of the repetition period and duration τ ₀ ≪ t ₁ - at the first output of synchronizer 5;

50 is an oscillogram of the control pulse of the transmitter 3 and the information processing device 17 with a period T _p starting at the beginning of the repetition period and duration T _and ≪T _p at the second output of the synchronizer 5;

относительно начала периода повторения, длительностью импульсов

и числом импульсов в последовательности51 - waveforms of the control pulses of the range gating unit 9 and the information processing device 17, which are periodic sequences of adjacent pulses with a period T _p , the beginning of the sequence at the time

relative to the beginning of the repetition period, pulse duration

and the number of pulses in the sequence

where R _max , R _min - the maximum and minimum ranges of the field of view, respectively, at the third output of the synchronizer 5;

52 - oscillograms of clock pulses for controlling the converter 11 code-frequency with a period of T = τ _c ≪T _p and a duration of τ _t ≪T at the fourth output of the synchronizer 5.

Figure 7 presents the structural diagram of the synchronizer 5, where the following notation:

53 - clock generator,

54 - counter

55 - decoder,

56 - block triggers.

In the diagram of Fig. 7, a clock generator 53, a counter 54, a decoder 55, and a trigger block 56 are connected in series, the first, second, and third outputs of which form the outputs of the same name 5. The output of the clock generator 53 is also connected to the fourth output of the synchronizer 5, and the input of the synchronizer is connected to the control input of the decoder 53.

As an antenna 1, a mirror antenna can be used [6, p. 46], controlled by a servo drive in horizontal and vertical planes according to commands received from the information processing device 17.

Antenna switch 2 can be made on the basis of a ferrite Y-circulator [6, p. 46].

The transmitter 3 is an amplification chain consisting of a pathogen and a power amplifier connected in series with a pulse modulator, and the pathogen can be made according to the scheme shown in RF patent No. 2083995 of 12/27/94, figure 2, consisting of m = n _f amplification-multiplying chains of quartz frequencies commuted by code signals from the frequency tuning unit 4, the reference frequency generator and mixers, and the power amplifier, depending on the required power and the band of amplified frequencies p It is realized on the basis of an electrovacuum device (an amplitron, a traveling wave lamp, etc.) or a semiconductor device [7, p. 19-52] and a pulse modulator constructed according to well-known schemes [7, p. 103-107].

Receiver 6 is a serial connection of a high-frequency amplifier, the input of which is the first input of the receiver, a mixer, the local oscillator input of which is the second input of the receiver, and an intermediate-frequency amplifier, the output of which is the output of the receiver.

The complex envelope shapers 14 ₁ , 14 ₂ , ..., 14 _m can be constructed on the basis of a series connection of peak detectors that expand the signal pulses and low-pass filters (in both quadrature channels).

Filters 15 ₁ , 15 ₂ , ..., 15 _m Doppler frequencies are multichannel filters with the number of channels M equal to the number of coherently processed pulses.

A radar station that implements the claimed method for the detection and classification of surface targets, operates as follows.

The transmitter 3 generates coherent pulsed sounding signals with a carrier frequency controlled by the frequency tuning unit 4, these signals pass through the antenna switch 2 to the antenna 1, are emitted into space. The coordinates of the targets (range R _j and angle β _j ) detected during the search are received from the third and first outputs of the information processing device 17 to the inputs of the synchronizer 5 and the drive of the antenna 1, respectively, and determine the location of the viewing zones when classifying the targets.

At the same time, from the second output of the information processing device 17, the command “work on a constant carrier frequency” is received at the second input of the frequency tuning unit 4, and at the same time, dipole clouds are selected.

Block 4 frequency adjustment works as follows (figure 2).

The noise voltage generator 18, constructed, for example, on the basis of a noise diode, generates noise oscillations with a spectral width significantly exceeding the repetition frequency (Δf _w ≫F _p ). Further, these oscillations are amplified and limited in the amplifier-limiter 19 and fed to the counter 20, which counts, for example, positive edges modulo m and, thus, has m equiprobable states. The clock pulses coming from the first output of the synchronizer 5 to the first input of the frequency tuning unit 4 through the repetition period T _p with a lead time of t ₁ relative to the beginning of the next period come to the pulse frequency divider 24, which can be constructed, for example, based on the counter, and divides the repetition rate by M times, and also through the delay element 27 by the edge time, to the input of the And 25 element, to the other input of which pulses with a period of MT _p come from the output of the divider 24, as a result of which pulses are formed at the output of the And 25 element with a period of MT _p , and with a lead of t ₁ relative to the beginning of the next repetition period of the probe pulses.

These pulses pass through the key block 26, open when there is a “frequency switcher operation” control command at its control input that goes to the second input of the frequency switcher 4, and open the And 21 element, through which the readings of the counter 20 are recorded in register 22 and converted in the decoder 23 into a parallel code with the number of bits m, where only one of m bits is nonzero, which is stored for at least M repetition periods and determines the frequency frequencies of the signal f _{Гi of the} local oscillator in the next packet of M pulses.

When entering the control input of the key block 26 through the second input of the frequency adjustment unit 4 of the command from the device 17 "operation at the frequency f _i = const", the block 26 closes, the pulses to the second input of the And 21 element no longer arrive, and the new meter 20 readings cease to arrive to the register 22, so that its code and the code at the output of the decoder 23 remain unchanged and, therefore, the frequencies f _i and f _gi also remain constant until the command changes at the second input of block 4, and in all cases the relation

The signals reflected from the target are received from the antenna 1 through the third arm of the antenna switch 2 to the input of the receiver 6, where they are amplified at a high frequency, converted to an intermediate frequency under the action of oscillations of the local oscillation frequency f _gi , received at the local oscillator input of the receiver 6 from the second output of the transmitter 3, and after amplification at an intermediate frequency, they come from the output of the receiver to the signal input of the two-channel phase detector 7, to the input of the oscillations of the reference frequency of which the oscillations of the frequency f _pc + F _d come in New Balanced Modulator 12.

Dual balanced modulator 12 operates as follows (figure 4).

приходят на фазовращатель 31, на выходе которого они сдвигаются по фазе на

приобретают вид U_дsin(2πF_дt-φ_д), после чего попадают на первый вход перемножителя 34, на второй вход которого приходят со второго входа ДБМ 12 из возбудителя передатчика 3 колебания промежуточной частоты, имеющие вид U_пчcos(2πF_пчt-φ_пч). В результате перемножения на выходе перемножителя 34 появляется напряжение kU_пчU_дsin(2πF_дt-φ_д)cos(2πF_пчt-φ_пч), которое поступает на вход сумматора 35. Аналогично колебания промежуточной частоты, поступающие на второй вход ДБМ 12, сдвигаются по фазе фазовращателем 33 и попадают на второй вход перемножителя 32, на первый вход которого поступают колебания частоты Доплера F_д с первого входа ДБМ 12. В результате перемножения на выходе перемножителя 32 образуется напряжение kU_пчU_дcos(2πF_дt-φ_д)sin(2πF_пчt-φ_пч), которое попадает на другой вход сумматора 35.The first input of the DBM 12 receives the oscillation of the Doppler frequency F _d corresponding to the radial speed of the radar carrier in the direction of the observed target. They are formed in the converter 11 code-frequency (see below). These fluctuations, expressed as

come to the phase shifter 31, at the output of which they are phase shifted by

take the form of U _d sin (2πF _d t-φ _d), and then fall to a first input of multiplier 34, to the second input of which comes from the second input DBM 12 of the transmitter exciter 3 fluctuations of the intermediate frequency of the form U _nv cos (2πF _pch t -φ _pch ). As a result of multiplying the output of multiplier 34 receive voltage kU _pch U _d sin (2πF _d t-φ _e) cos (2πF _IF t-φ _nq), which is input to the adder 35. Similarly, the intermediate frequency vibrations input to the second input of the DBM 12 , phase shifted by phase shifter 33 and fall to a second input of multiplier 32, a first input of which the oscillation frequency F _d Doppler received from the first input DBM 12. As a result of multiplying the output of the multiplier 32 forms the voltage U _d kU _nv cos (2πF _d t-φ _d ) sin (2πF _pch t-φ _pch ), which falls on the other input of the adder 35.

As a result of the summation at the output of the adder 35, the voltage of the support U _op with a frequency f _pc + F _{d is formed} , which has the form

where designated U _op = U _P U _d

Fluctuations in the frequency F _d are formed in the converter 11 code-frequency, which operates as follows (figure 3).

, по поступающим на входы вычислителя 10 значениям скорости V носителя, несущей частоты f_i и угла β_j. Прямоугольные импульсы в форме меандра с требуемой частотой поступают с выхода делителя 29 частоты на фильтр 30 нижних частот, выделяющий первую гармонику, синусоидальные колебания требуемой частоты поступают на выход преобразователя 11 код-частота, соединенный с входом ДБМ 12.The second input of the Converter 11 receives the clock pulses of the quartz frequency from the fourth output of the synchronizer 5, which come to the signal input of a controlled frequency divider 29, made on the basis of the counter. The control input of the divider 29 from the first input of the converter 11 through the decoder 28 receives control signals that determine the frequency division coefficient, which corresponds to the frequency value F _d coming from the Doppler frequency calculator 10, where it is calculated in accordance with the expression

, according to the values of the carrier speed V, the carrier frequency f _i and the angle β _j coming to the inputs of the calculator 10. Rectangular pulses in the form of a meander with the required frequency are supplied from the output of the frequency divider 29 to the low-pass filter 30 that selects the first harmonic, sinusoidal oscillations of the required frequency are fed to the output of the code-frequency converter 11 connected to the input of the DBM 12.

один относительно другого вследствие прохождения через специальный фазовращатель в цепи опорного напряжения одного из них. На выходе фазового детектора 7 образуются когерентные видеосигналы в двух квадратурных каналах, в которых проведена компенсация доплеровской частоты, соответствующей проекции скорости V носителя в направлении оси диаграммы направленности антенны при наблюдении j-ой цели.Further, as described above, in the DBM 12, frequency oscillations f _pc + F _d are generated, which are fed to the input of the reference frequency of the phase detector 7, consisting of two identical phase detectors, to which the oscillations of the reference frequency are shifted by

one relative to the other due to passage through a special phase shifter in the voltage reference circuit of one of them. At the output of the phase detector 7, coherent video signals are generated in two quadrature channels, in which the Doppler frequency compensation is performed corresponding to the projection of the carrier velocity V in the direction of the antenna pattern axis when observing the jth target.

каналов дальности благодаря стробам, поступающим с третьего выхода синхронизатора 5 на вход блока 9 стробирования по дальности.These video signals are amplified in block 8 of video amplifiers, consisting of two video amplifiers, the band of which is consistent with the width of the spectrum of the received video pulses, and fall into block 9 of the gating range, in which they are gated in range in both quadrature channels, while in each of them is formed

range channels due to the gates coming from the third output of the synchronizer 5 to the input of the range gating unit 9.

The quadrature components of the pulsed video signals from the gated target are fed to the inputs of the switch 13, which connects to the input of one of the m processing channels depending on the number (code) of the carrier frequency that is input to the control input of the switch 13 from the output of the frequency tuning unit 4.

The processing of video signals from the gated target begins in block 14 of the formation of the complex envelope, in which the complex envelope is formed by expanding the pulses (using peak detection) and low-pass filtering (in both quadrature channels).

The quadrature components of the complex envelope thus formed are then filtered by the Doppler frequency in a multichannel filter of 15 Doppler frequencies with the number of channels M equal to the number of coherently accumulated repetition periods.

From the outputs of all M channels of the PDF 15, the filtered signals enter a multichannel detector 16 of the frequency components, in which the amplitude detection of signals in all M channels occurs, so that the blocks 15 and 16 form essentially a spectrum analyzer.

Further, the signals from the outputs of all the detectors 16 of the frequency components in each of the range channels fall into the information processing device 17, the operation of which is described above (see Fig. 5).

After the end of the first stage of selection - selection of dipole clouds, which is performed at a constant carrier frequency, by the command from the second output of the device 17 to the second input of the frequency tuning unit 4, the latter switches to the frequency tuning mode from a burst to - a burst of M pulses. This is achieved by opening the key block 26 (figure 2), as described above.

Further, in the device 17 is the selection of corner reflectors and the classification of targets, as described above.

The operation of the synchronizer 5 (Fig.7) consists in the formation of control signals (Fig.6), while the signal 52 is generated at the output of a clock pulse generator 53 connected to the fourth output of the synchronizer 5, and other signals with a counter 54, a decoder 55 and block RS-flip-flops 56, forming signals of the desired duration and delay anticipation relative to the beginning of the repetition period. Moreover, since for various purposes the delay time t _{2 of the} set of strobe pulses 51 relative to the beginning of the repetition period turns out to be different, the required values of t ₂ are received for the next target from the third output of the device 17 through the synchronizer 5 input to the decoder 55.

The technical advantage of the proposed method for the detection and classification of surface targets in comparison with the prototype method is the expansion of functionality, expressed in the ability to classify surface targets not only into "true" and "false", but also to classify true targets into "main" and "not main "

The presented description and drawings make it possible to implement the inventive method using the existing elemental base and technology for the production of radio devices in accordance with the purpose, which characterizes the industrial applicability of the object of the invention.

Bibliography

1. Bakulev P.A. Radar of moving targets, M .: Sov. radio, 1964.

2. US Patent No. 4119966, cl. G 01 S 13/52, 1973.

3. Krasyuk I.P., Rosenberg V.I. Shipborne radar and meteorology. L .; Shipbuilding, 1970.

4. RF patent No. 2083996, class G 01 S 13/524 of 06/27/95, publ. 07/10/97. bull. No. 19, prototype.

5. Reutov A.P., Mikhailov B.A., Kondratenkov G.S., Boyko B.V. Radar side view. M .: Sov. radio, 1970.

6. Handbook of Radar, ed. M. Skolnik, vol. 2, M .: Sov. radio, 1977.

7. Handbook of Radar, ed. M. Skolnik, vol. 3, M .: Sov. radio, 1979.

Claims (2)

1. A method for detecting and classifying surface targets, including at the first stage emitting coherent radio pulses with a constant carrier frequency, receiving reflected signals, filtering the received signals by Doppler frequency on M equally spaced frequency components in each resolution element by range, measuring the power of frequency components, comparing the results measurements with a threshold level, the decision on the presence of a signal from a true target, if the power of at least one, but not more than k adjacent frequency ulation exceed the first threshold level, where k - the ratio of the maximum expected width range mezhperiodnyh fluctuation of signals from true targets to the bandwidth of one frequency component comprising the second stage emission of coherent radio pulses during n repetition periods varying from pulse to pulse carrier frequency f _i, i = 1, 2, ..., m, receiving reflected signals from targets accepted as true at the first stage of selection, filtering the signals at each of the carrier frequencies by the Doppler frequency at M equally spaced parts components, measuring their powers, determining the normalized power values of the frequency components, comparing the result with a second threshold level and deciding whether there is a signal from a true target when the second threshold level is exceeded, characterized in that of the goals accepted as true at the second stage of selection , choose targets located in the directions β _j relative to the direction of the carrier velocity vector V and satisfying the condition

where Ψ _0.5 is the width of the antenna pattern at the level of -3 dB in the plane passing through the vector V and the direction to the j-th target, for their classification, bursts of M probe pulses are emitted, each with a duration of n repetition periods of the carrier frequency f ₁ , changing the carrier frequency from the burst to the burst of probe pulses and keeping it constant inside the bursts and choosing the parameters M, m and n from the relations:

n = m where R _j is the slant range to the j-th target; F _p - the repetition frequency of the probe pulses in the packet; V is the speed of the radar carrier, when detecting the received signals, the Doppler frequency shift corresponding to the projection of the carrier velocity onto the direction of the directional axis axis is compensated when observing the jth target, the capacities of the frequency components S _irq (r = 1, 2, ..., M) of the received m ₁ < m bursts of pulses in each element q of range resolution from the range q _min <q <q _max , taking into account the predicted change in the coordinates of the j-th target during the accumulation time due to the movement of the radar carrier, the following signals are generated:

where q _min , q _{max are} determined by the boundaries of the field of view in range when observing the j-th target, find the coordinates r ₁ (q), r ₂ (q) of the contour of the radar image of the j-target from the conditions Z (r ₁ -1, q) <C ₀ Z (r ₁ , q) ≥C ₀ Z (r ₂ , q) ≥C ₀ z (r ₂ +1, q) <C ₀ and boundary ones - minimum q ₁ and maximum q ₂ - the values at which these conditions are satisfied, where С ₀ > 0 is the threshold level determined by the admissible probability of false operation due to noise and interference, then assess the angle α _j , j-th target and its maximum longitudinal and transverse dimensions L, B in accordance with the relations

where r (q) = ad + d, and the parameters a, d are determined from the condition

Z _cf = 0.5 [Z ₁ (q) + Z ₂ (q)], ΔR, ΔХ are the dimensions of the resolution element in the radial (in range) and transverse directions, respectively, then verify that the condition q ₂ -q ₁ ≥r (q ₂ ) - r (q ₁ ) is met and, if it is not satisfied, then perform permutations:

and repeat the operation of finding the coordinates of the contour of the radar image and the operation of determining the size of the target until the condition q ₂ -q ₁ ≥r (q ₂ ) -r (q ₁ ) is fulfilled, after which they make a decision about the class of the observed target, considering that it is "main" if the relation holds

where L ₁ , L ₂ , B ₁ , B ₂ - a priori known values of the length and width of the radar images of the "main" and "not main" targets, respectively; σ _L , σ _B are the corresponding a priori known average standard deviations, And ₀ is the threshold value chosen depending on the desired ratio between the probabilities (and costs) of classification errors. 2. The method according to claim 1, characterized in that the coordinates of the most reflective targets are additionally used as informative signs, for which the Doppler frequency F _{dj of the} reflected signals arriving in the direction β _{j of} the antenna pattern axis relative to the direction of the radar carrier velocity vector V is measured, at

where F ₀ is the threshold value, determine the coordinates of the bow or stern of the observed j-th target according to the rules of the form: for F _dj > 0: q _n = q ₁ , r _n = a · q ₁ + d (nose), when F _dj <0: q _n = q ₂ , r _n = a · q ₂ + d (feed) compare the accumulated values of the signals Z (r, q), determine the coordinates r * _i , q * _{j of} their maximum values, where i = 1, 2, ..., u, u≥0, find the distances of these points from the nose, according to the rules:

- along the longitudinal axis,

- in the transverse direction, and decide on the class of the observed target according to the rule: if

then the goal is "main", where l _01i , l _02i , B _01i , B _02i are the a priori known distances averaged inside the classes of the most reflecting points from the bow of the surface ship in the longitudinal and transverse directions for the classes of "main" and not "main" targets, respectively, σ _li , σ _bi are the corresponding averaged standard deviations, A ₁ is a threshold value chosen depending on the desired relationship between the probabilities (and costs) of classification errors.

Images