RU2389039C1 - Method of measuring air target radial velocity in carrier frequency adjustment mode from pulse to pulse in accordance with random law at low signal-to-noise ratio - Google Patents

Abstract

FIELD: physics. ^ SUBSTANCE: method can be used in pulsed Doppler radar with carrier frequency adjustment from pulse to pulse. The random law of variation of probing frequency is algorithmically generated from a linear-stepped law of variation of frequency in a packet of pulsed signals. The centimetre wavelength range with frequency adjustment within the limits of hundreds of MHz is used. The number of signals used in a packet equals 2k, where k=6Ç8. Duration of the packet must not exceed the interval for angular correlation of rotation of the air target. Velocity is measured during one packet using a technique for finding the extremum of the multiplicative coefficient which is a product of the estimated variance and a third central moment for the complex vector of the range portrait, obtained from the frequency characteristic of the target, re-phased with complex factors which take into account selection of all possible radial velocities of the object. ^ EFFECT: increased accuracy of measuring radial velocity and noise-immunity due to adjustment of frequency from pulse to pulse in accordance with a random law and coherent integration of amplitude of reflected pulses during inverse Fourier transformation with a complex vector of the frequency characteristic of the target. ^ 12 dwg, 4 tbl

Description

The invention relates to the field of radar and is intended to measure the radial speed of an object when using the mode of tuning the carrier frequency from pulse to pulse according to a random law that eliminates the negative impact of interference-targeted interference frequency.

попарных произведений комплексно-сопряженных отсчетов отраженных сигналов одной частоты в смежных периодах повторения по формулеA known method of measuring the radial velocity of a target when using multi-frequency signals [1], which includes emitting multi-frequency signals consisting of two frequency components, receiving signals reflected from the target by two frequency channels of the radar, lowering the frequency of the received signals to an intermediate one, amplifying the received signals, converting the received signals to video frequency using two quadrature phase detectors, converting the quadrature components of the signals into digital form using analog-to-digital conversion teli and carrying out with them algorithmic processing in a digital computer. During algorithmic processing, in each k-th frequency channel, the complex sum is found

pairwise products of complex conjugate samples of reflected signals of the same frequency in adjacent repetition periods according to the formula

,

,

where the symbol * means the operation of complex pairing;

k is the number of the frequency channel, with k = 1.2;

j is the number of the repetition period of the multi-frequency signal.

Then, phase arguments are determined in each frequency channel.

, которые являются доплеровскими сдвигами фаз сигнала за период повторения сигналов одинаковой частоты Т, вычисляют разность Δφ полученных фазовых аргументов φ^(k) сумм в двух, отличающихся частотой излучения, каналахamounts found

which are Doppler phase shifts of the signal during the repetition period of signals of the same frequency T, calculate the difference Δφ of the obtained phase arguments φ ^{(k) of the} sums in two channels differing in the frequency of radiation

Δφ = φ ⁽¹⁾ -φ ⁽²⁾ ,

цели по формулеwhere φ ⁽¹⁾ and φ ⁽²⁾ are the Doppler phase shifts of the reflected signal during the repetition period T of single-frequency signals in the first and second channels, respectively. Further, the ambiguity associated with the possibility of obtaining the Δφ value of a different sign is eliminated, and then the radial velocity estimate is calculated from the phase difference Δφ

goals by formula

,

,

where c is the propagation velocity of electromagnetic waves; T is the value of the pulse repetition period of the same frequency (the period is the same in both frequency channels);

f ⁽¹⁾ and f ⁽²⁾ are the values of the sounding frequencies in the first and second channel, respectively.

The disadvantage of this method is the inability to measure the speed of the target in the conditions of application of interference-oriented interference in frequency. The use of a single-frequency or two-frequency signal with a regular frequency change creates favorable conditions for revealing the radiation frequency used and for setting powerful interference at the detected frequency in the following repetition periods.

A known method of measuring the radial speed of an air target in the frequency-from-pulse to pulse-frequency adjustment mode [2], which consists in the fact that the number of pulse signals generated for radiation is chosen equal to the number N of used radiation frequencies, and the number N is chosen equal to 2 ^k , where k - an integer that takes a value from 6 to 8.

The time Δt for radiation of a packet of signals with frequency tuning is chosen no more than the interval of angular correlation T _y rotation of an air target in flight when yawing, amounting to 5 ms. Hence, the pulse repetition period T _and is chosen no more than T _and = T _yк / (N-1).

Using a radar station, a packet of signals with tuning of the carrier frequency is emitted during the time interval Δt. When emitting pulses with frequency tuning, a random law of frequency variation is used, for which a sequence of frequency values used in a packet from f ₀ to f ₀ + F _per with a step Δf = F _per / (N-1) is generated and pre-stored in the random access memory where f ₀ is the main carrier frequency of the probe signal of the centimeter range, F _per = 150 MHz is the range in which the frequency is tuned from pulse to pulse. The radiation frequency numbers are distributed according to a random law, in which the radiation time t _n pulse at the n-th frequency f ₀ + nΔf is determined by the formula

- порядковый номер импульса на n-й частоте, принимающий значение от 1 до N, единожды повторяющееся в пределах пачки сигналов с перестройкой частоты.Where

- the serial number of the pulse at the nth frequency, taking a value from 1 to N, repeating itself once within a packet of signals with frequency tuning.

Receive signals reflected from the target at n-th frequencies, where n is the frequency number of the signal, lower the frequency of the received signals to an intermediate one. When lowering the frequency of the received signals to the intermediate, take into account the value of the additive nΔf used when emitting the signal at the n-th frequency f ₀ + nΔf, so that the frequency difference of the signal received at the n-th frequency is f ₀ + nΔf + F _{d n} , where F _{d n} - additive Doppler frequency of the reflected signal at n-th frequency due to target radial velocity, and the local oscillator signal (f ₀ -f _pr) + nΔf always determined only by the magnitude of the intermediate frequency f and corresponding value _ave Doppler additives F _{d n.}

Using quadrature phase detectors, the quadrature components of the received signals are isolated. Quadrature components are converted to digital form using analog-to-digital converters. Each signal reflected at the nth frequency is converted into a complex form of the form

,

,

- амплитуда отраженного на n-й частоте сигнала;Where

- the amplitude of the signal reflected at the nth frequency;

- фаза отраженного на n-й частоте сигнала;

- phase of the signal reflected at the nth frequency;

и

- значения квадратурных составляющих принятого сигнала на n-й частоте.

and

- the values of the quadrature components of the received signal at the nth frequency.

в порядке увеличения n. Для этого формируют вектор G из N элементов, записывая в n-й элемент вектора G комплексное значение

отраженного на n-й частоте сигнала. Формируют двумерную матрицу D данных из N строк и Z=2V_{p max}/dV+1 столбцов, где V_{p max} - максимально возможная радиальная скорость цели, выбираемая заблаговременно, dV - интервал дискретизации радиальной скорости, определяющий точность измерения радиальной скорости, записывают в элемент n-й строки z-го столбца матрицы D комплексную величину

, рассчитанную по формулеSet Values

in increasing n order. To do this, form a vector G of N elements, writing the complex value to the nth element of the vector G

reflected at the nth frequency of the signal. A two-dimensional data matrix D is formed from N rows and Z = 2V _{p max} / dV + 1 columns, where V _{p max} is the maximum possible radial velocity of the target, chosen in advance, dV is the sampling interval of the radial velocity, which determines the accuracy of measuring the radial velocity, is written into the element nth row of the zth column of matrix D is a complex quantity

calculated by the formula

,

,

- комплексная величина n-го элемента вектора G; z - номер столбца матрицы D. Получают матрицу D1 путем проведения обратного быстрого преобразования Фурье с комплексными векторами данных каждого столбца матрицы D, в результате в каждом столбце получают дальностный портрет (ДлП) воздушной цели. При этом учитывают, что истинный информативный ДлП цели может быть сформирован только при точной фазовой фокусировке принятой реализации сигналов, т.е. при точном устранении фазовых сдвигов, обусловленных радиальном перемещением цели. При игнорировании негативного влияния фазовых сдвигов, обусловленных радиальным перемещением, ДлП получается размытым, т.е. не отражающим взаимного расположения рассеивающих центров поверхности цели в радиальным направлении. В этом случае ДлП имеет шумовую форму. Для поиска условий, при которых формируется информативный ДлП цели находят максимальное значение модуля комплексного сигнала в матрице D1 и делят комплексные величины всех элементов матрицы D1 на это значение, а затем рассчитывают величину энтропии данных Hz для каждого z-го столбца матрицы D1 по формулеWhere

is the complex value of the nth element of the vector G; z is the column number of the matrix D. Get the matrix D1 by performing the inverse fast Fourier transform with complex data vectors of each column of the matrix D, as a result in each column receive a long-range portrait (DL) of an air target. At the same time, it is taken into account that a true informative target DL can be formed only with exact phase focusing of the received signal implementation, i.e. with the exact elimination of phase shifts due to radial movement of the target. When ignoring the negative effect of phase shifts due to radial displacement, the DLP is blurred, i.e. not reflecting the relative position of the scattering centers of the target surface in the radial direction. In this case, the DLP has a noise shape. To find the conditions under which an informative DLP is formed, the targets are found the maximum value of the modulus of the complex signal in the matrix D1 and the complex values of all elements of the matrix D1 are divided by this value, and then the entropy value of the data Hz for each z-th column of the matrix D1 is calculated by the formula

.

.

The column number z _{min H} corresponding to the lowest entropy value H _{z min} is determined, based on which the radial velocity of the target is determined by the formula

.

.

.The found estimate is taken as the measured value of the radial velocity of the target

.

для N=const, определялось по 3000 реализациям. В качестве критерия возможности измерения V_p выбиралось выполнение условия, чтобы ошибка измерения в каждом опыте из 3000 не превосходила 10 м/с. В табл.1 приведены значения q_min, соответствующие различному числу N импульсов в пачке.The performance of this method depends on the intensity of the noise [2]. High noises distort the DL so that the entropy of the elements representing the DL vector becomes a random variable. In [3], the value of the minimum signal-to-noise ratio (SNR) q _min at which measurement is permissible

for N = const, was determined by 3000 implementations. As a criterion for the possibility of measuring V _{p, the} fulfillment of the condition was chosen so that the measurement error in each experiment out of 3000 does not exceed 10 m / s. Table 1 shows the q _min values corresponding to a different number N of pulses in a packet.

Table 1. Dependence of the minimum signal-to-noise ratio on the number of pulses in a packet N 64 128 256 512 1024 g _min , dB 5 1,2 -1.4 -3.4 -5.5

An increase in N leads to a decrease in the minimum allowable signal-to-noise ratio and standard deviation of the velocity measurement error.

Thus, there are limit values of the noise level, limiting the applicability of the above method. Therefore, it is of interest to develop a method that provides a reliable estimate of the radial velocity at a lower signal-to-noise ratio compared to the method [2].

The objective of the invention is the provision of measuring radial velocity followed by the angular coordinates and range of an air target in the mode of tuning the carrier frequency from pulse to pulse according to a random law with a reduced signal-to-noise ratio.

To solve this problem, it is proposed in the known method [2, 3] instead of the entropy of the DL elements used to search for the air target’s speed, use the multiplicative coefficient calculated by the formula

,

,

- оценка дисперсии амплитуд z-го столбца матрицы D1,Where

- assessment of the variance of the amplitudes of the z-th column of the matrix D1,

- оценка математического ожидания амплитуд z-го столбца матрицы D1,

- assessment of the mathematical expectation of the amplitudes of the z-th column of the matrix D1,

- estimate of the third central moment of the amplitudes of the z-th column of the matrix D1.

All other operations that reflect the essence of the method [2], it is advisable to leave unchanged.

может быть описан следующей совокупностью последовательных действий. В оперативном запоминающем устройстве формируют последовательность величин частот, используемых в пачке СПЧ, от f₀ до f₀+F_пер. Затем номера частот излучения распределяют по случайному закону, при котором время излучения t_n от начала пачки для импульса на n-й частоте f₀+nΔf, где n - номер используемой частоты

, определяется по формуле

. Например, если в 15-м периоде излучен импульс на 7-й частоте, то

при n=7. Порядок использования частот запоминается для последующей расстановки принятых сигналов в порядке линейного увеличения частоты. Величина T_и выбирается, исходя из требования обеспечения однозначности отсчетов по доплеровской частоте во всем диапазоне возможных радиальных скоростей цели [8], что в символьном виде выражается неравенствомThus, an improved measurement method

can be described by the following set of sequential actions. In the random access memory, a sequence of frequency values used in the FH bundle is formed, from f ₀ to f ₀ + F _per . Then the radiation frequency numbers are distributed according to a random law, in which the radiation time t _n from the beginning of the burst for a pulse at the n-th frequency f ₀ + nΔf, where n is the number of the used frequency

determined by the formula

. For example, if in the 15th period a pulse is emitted at the 7th frequency, then

for n = 7. The frequency usage order is remembered for the subsequent arrangement of the received signals in the order of a linear increase in frequency. The value of T _and is selected based on the requirement to ensure the uniqueness of the samples at the Doppler frequency in the entire range of possible radial velocities of the target [8], which is symbolically expressed by the inequality

F _and > F _{d max} ,

- частота повторения импульсов внутри пачки;Where

- pulse repetition rate inside the packet;

F _{d max} = 2V _{p max} (f ₀ + F _per ) / c - the maximum possible Doppler frequency of the target.

During the time Δt, limited by the interval of the angular correlation of the rotation of the air target T _yk , which is not more than 5 ms [5, 6], a packet of 2 ^k pulsed FHRs is randomly generated, where k is an integer taking a value from 6 to 8. The number of pulse signals is more convenient to take equal to 2 ^k , since in this case, the processing can use fast Fourier transform algorithms, which can significantly reduce the amount of computation.

It is further proposed to receive signals reflected from the target at different frequencies f ₀ + nΔf + F _{d n} , where F _{d n} is the Doppler frequency addition of the reflected signal at the nth frequency due to the radial speed of the target, then it is proposed to lower the frequency of the received signals to an intermediate f _pr + F _{d n} , where f _CR is the value of the intermediate frequency, while taking into account the value of the additive nΔf used when emitting a signal at the n-th frequency f ₀ + nΔf, so that the frequency difference f _{pr n} of the signal received at the n-th frequency f ₀ + nΔf + F _{d n} and the local oscillator signal (f ₀ -f _pr) + nΔf always determined Referring only to the intermediate frequency value f _etc. and the corresponding value of the Doppler additives F _{d n}

f _{pr n} = (f ₀ + nΔf + F _{d n} ) - [(f ₀ -f _pr ) + nΔf] = f _pr + F _{d n} .

Then, using the quadrature phase detectors [7], it is necessary to extract the quadrature components of the received signals, convert the quadrature components to digital using analog-to-digital converters, and convert each signal reflected at the nth frequency into a complex form of the form

.

.

отраженного сигнала на n-й частоте. При этом принятые сигналы внутри вектора G расставляют в порядке линейно-ступенчатого изменения частоты. После этого формируют двумерную матрицу D данных из N строк и Z=2V_{p max}/dV+1 столбцов. В элемент n-й строки z-го столбца матрицы D записывают комплексную величину

, рассчитываемую по формулеThen it is proposed to form a vector G from N elements, write the complex value to the nth element of the vector G

reflected signal at the nth frequency. In this case, the received signals inside the vector G are arranged in the order of a linear-step change in frequency. After that form a two-dimensional matrix D of data from N rows and Z = 2V _{p max} / dV + 1 columns. In the element of the nth row of the zth column of the matrix D write the complex value

calculated by the formula

,

,

- комплексная величина n-го элемента вектора G.Where

is the complex value of the nth element of G.

It is further proposed by performing the inverse fast Fourier transform with complex data vectors of each column of the matrix D to obtain the matrix D1, and then calculate the coefficient k _z for each z-th column of the matrix D1 by the formula

.

.

по формулеAt the final stage, find the number of the column z _{max k} corresponding to the largest value of the coefficient k _{z max} , with which determine the estimate of the radial velocity of the target

according to the formula

.

.

and take this estimate as the measured value of the radial speed of the air target.

при измерении радиальной скорости воздушной цели в условиях понижения отношения сигнал-шум состоит в следующем.The essence of the method, namely the essence of replacing entropy with a multiplicative coefficient

when measuring the radial speed of an air target in conditions of lowering the signal-to-noise ratio is as follows.

As is known, the variance of a random variable is a characteristic of dispersion, the dispersion of a quantity near its mathematical expectation. In this regard, it will increase in the case of the formation of an informative DLP (the appearance of unclear responses in amplitude exceeding the DLP noise) and will be minimal if the complex amplitudes of the reflected signals are randomly and evenly distributed over the numbers of the DLP elements. This situation is observed if the true value of the radial velocity does not coincide with one of its values used for phasing the vectors in the columns of the matrix D, since this does not compensate for phase incursions associated with moving the target, which means that the conditions for the formation of an informative DL are violated. The long-range portrait in this case is a vector with randomly distributed amplitudes (Fig. 1).

Such a distribution of complex amplitudes in unphased DLP is an analog of a random noise process. It is also known that the third central moment serves to characterize the asymmetry (or “slanting”) of the distribution. If the distribution is symmetric with respect to the mathematical expectation, then the moment tends to zero, i.e. in the case of the formation of informative DLP (figure 2) this moment will have a maximum value.

It should be noted that long-range portraits blurred by phase noise will have the property of stationarity, i.e. constancy of numerical characteristics extracted from vectors of their complex amplitudes.

объясняется, во-первых, удобством использования одного числового коэффициента вместо двух при алгоритмизации предлагаемого способа для реализации на ЭВМ, а во-вторых - повышением входного отношения сигнал-шум, при котором предложенный способ сохраняет работоспособность. При расчетах удобнее использовать мультипликативный коэффициент, так как в случае использования аддитивного коэффициента возникает необходимость в нормировке входящих в него слагаемых. На фиг.3, 4, 5 представлены соответственно зависимости энтропии, оценки дисперсии амплитуд z-го столбца матрицы D1, оценки третьего центрального момента амплитуд z-го столбца матрицы D1 от значения радиальной скорости. Проведенные расчеты показали, что, например, при отношении сигнал-шум 10 дБ для 128 импульсов в пачке коэффициент корреляции зависимостей энтропии и оценки дисперсии от значения радиальной скорости равен -0,99. Это означает, что данные функции имеют отрицательную корреляцию и высокую линейную зависимость.Expediency of using one generalized multiplicative coefficient

firstly, it is explained by the convenience of using one numerical coefficient instead of two when algorithmizing the proposed method for computer implementation, and secondly, by increasing the input signal-to-noise ratio, in which the proposed method remains operational. In calculations, it is more convenient to use the multiplicative coefficient, since in the case of using the additive coefficient, it becomes necessary to normalize the terms included in it. Figures 3, 4, 5 show, respectively, the dependences of entropy, variance estimates of the amplitudes of the z-th column of the matrix D1, estimates of the third central moment of the amplitudes of the z-th column of the matrix D1 on the radial velocity value. The calculations showed that, for example, at a signal-to-noise ratio of 10 dB for 128 pulses in a packet, the correlation coefficient of the dependences of entropy and dispersion estimate on the value of the radial velocity is -0.99. This means that these functions have a negative correlation and a high linear dependence.

If from the values of the functions of the estimates of entropy and dispersion, the graphs of which are presented in Figs. 3 and 4, subtract their average values, and then divide all the values of entropy by the maximum absolute value of entropy, and the values of the estimates of dispersion by the maximum value of dispersion, then the graphs of the newly formed functions will take the form shown in Fig.6 and 7, respectively. On Fig presents the result of the above operation of subtracting the average value and dividing by the maximum value for the evaluation function of the 3rd central moment. The result of term-by-term addition of the functions of normalized values of the estimates of entropy and dispersion is shown in graphical form in Fig. 9. Figure 9 shows that the scatter of the values of the inconsistencies of the functions is in the range of ± 3% with respect to the modulus of the maximum value of the normalized functions of the estimates of entropy and dispersion (equal to 1). This, combined with a high correlation of the functions of the estimates of entropy and dispersion, allows us to conclude that the possibilities of entropy and dispersion for measuring the radial velocity are almost identical. That is, in the known method [2], the measurement of radial velocity by the extreme value of entropy can (without loss of accuracy) be replaced by its measurement by the extreme value of dispersion. The correlation coefficient of the dependences of the entropy and estimation of the third central moment on the radial velocity value at a signal-to-noise ratio of 0 dB for 128 pulses in a packet has a value of -0.45 and decreases with increasing noise intensity.

If we subtract their average values from the values of the entropy functions and the estimates of the third central moment, and then divide all the values of entropy by the maximum value of entropy, and the values of the estimates of the third central moment by the maximum positive value of the 3rd central moment (Figs. 6, 8), and add the resulting functions, then the result of this addition will have the graphical interpretation shown in Fig. 10. It can be seen that the spread in the values of the inconsistencies of the functions is in the range of more than ± 65%. Moreover, on all graphs (Figs. 3, 4, 5) there is a surge (extremum) in the region of the true value of the radial velocity. With the term-by-term multiplication of the values of the variance estimation functions and the third central moment, the product result on all intervals except the points of global extrema (where the function argument is close to the true value of the radial velocity) will always be inferior in magnitude to the product of these functions at the point corresponding to the true value of the radial velocity. This, as shown by the results of mathematical modeling, leads to an increase in the operating range of the signal-to-noise ratio for the proposed method.

вектора ДлП от предполагаемого значения скорости приведен на фиг.11. График получен методом моделирования на ЭВМ. Оценка радиальной скорости в этом примере составила 100,3 м/с при истинной радиальной скорости объекта 100 м/с при отношении сигнал-шум 0 дБ для 128 импульсов в пачке. Моделирование показывает, что экстремальное значение коэффициента k_z в большей степени превышает среднее значение этого коэффициента, чем аналогичное превышение для энтропии в способе [2].Variant of coefficient dependence

vector DLP from the estimated value of the speed shown in Fig.11. The graph is obtained by computer simulation. The estimate of the radial velocity in this example was 100.3 m / s with a true radial velocity of the object of 100 m / s with a signal-to-noise ratio of 0 dB for 128 pulses per burst. Modeling shows that the extreme value of the coefficient k _z exceeds the average value of this coefficient to a greater extent than the similar excess for the entropy in the method [2].

There are limit values of the noise level, limiting the applicability of the proposed method. The value of the minimum signal-to-noise ratio q _min , at which the measurement of V _{p is} permissible for N = const, was determined during mathematical modeling using 3000 realizations. As a criterion for the possibility of measuring V _{p, the} fulfillment of the condition was chosen so that the measurement error in each experiment out of 3000 does not exceed 10 m / s. Table 2 shows the q _min values corresponding to different numbers of pulses in burst N.

Table 2. Dependences of the minimum signal-to-noise ratio on the number of pulses in a packet and the selected method of extracting information N 64 128 256 512 1024 q _min , dB (Entropy) 5 1,2 -1.4 -3.4 -5.5 q _min , dB (Multiplicative factor) 3 -0.5 -2.9 -4.9 -7 Δq _min , dB (difference) 2 1.7 1,5 1,5 1,5

Analysis of table 2 allows us to conclude that the minimum allowable signal-to-noise ratio is reduced when using the multiplicative coefficient k _{z by} an average of Δq = 1.5-2 dB. In addition, when using the multiplicative coefficient k _z increases the accuracy of measuring the radial velocity V _p . Table 3 shows the standard deviation (RMS) of the measurement error of the radial velocity depending on the number of pulses in the packet, pulse repetition period, signal-to-noise ratio for the selected method of extracting radial velocity information using the multiplicative coefficient k _z .

Table 4 shows the standard deviation (RMS) of the error of measuring the radial velocity depending on the number of pulses in the packet, the pulse repetition period, the signal-to-noise ratio for the selected method of extracting information about the radial velocity from the DLP using entropy

H _z .

A comparative analysis of tables 3 and 4 allows us to conclude that the reduction of the standard deviation of the velocity measurement error at N≤256 by 15-20% when using the coefficient k _z instead of entropy (Fig. 12). With an increase in the number of pulses, the decrease in the standard deviation becomes more significant, reaching in some cases 50%.

The proposed method is easily implemented and has the following advantages: increased noise immunity (increase up to 40% compared to the method [2]);

increased accuracy of measuring radial velocity (increase up to 50%) compared with the prototype.

The proposed method, being the development of the prototype [2], is currently the only one implemented for constructing systems for measuring the radial velocity of a target in radars with pulse-frequency tuning of the carrier frequency.

The proposed method can find application:

in promising radar systems with high-speed target selection and the construction of moving target selection systems;

in phase focusing algorithms for synthesized apertures in radio-vision problems;

in algorithms for recognizing aerial targets by their radar images [9,10,11].

Information sources

1. RF patent No. 2166772. IPC ⁷ G01S 13/58. Detector-meter of multi-frequency signals. / Popov D.I., Belokrylov A.G. Application No. 2000105563. Priority 6.03.2000 Publ. 05/10/2001 (analog).

2. RF patent №2326402. IPC ⁷ G01S 13/58. A method of measuring the radial speed of an air target in the frequency tuning mode from pulse to pulse. / Savostyanov V.A., Mayorov D.A., Mitrofanov D.G., Prokhorkin A.G. Application No. 2007101537. Priority 01/17/2007. Publ. 06/10/2008, BIPM 2008. No. 16. Part 3 p.752 (prototype).

3. Mayorov D.A., Savostyanov V.A., Mitrofanov D.G. Measurement of the radial speed of airborne objects in frequency tuning mode. // Measuring equipment, 2008. No. 2. p. 43-47.

4. Grigorin-Ryabov VV Radar devices. - M .: Owls. Radio, 1970 .-- 680 p.

5. Barton D.K., Ward G.R. Handbook of radar measurements. Per. from English / Ed. M.M. Weisbane. - M .: Owls. Radio. 1976.- 392 p.

6. Radar systems of multifunctional aircraft. T. 1. Radar - the information basis of the combat operations of multifunctional aircraft. Systems and algorithms for primary processing of radar signals. / Ed. A.I. Kanaschenkov and V.I. Merkulov - M .: Radio engineering, 2006. - 656 p.

7. Handbook of radar. / Ed. M.I.Skolnika. Per. from English - M .: Owls. Radio, 1967. Vol. 1. Basics of radar. - 456 p.

8. Guide to the basics of radar technology. / Ed. Druzhinina V.V. - M .: Military Publishing House, 1967 .-- 768 p.

9. RF patent No. 2234110. IPC ⁷ G01S 13/89. A method of constructing a two-dimensional radar image of an air target. / Mitrofanov D.G., Bortovik V.V. and other Application No. 2003100255. Priority January 24, 2003 Publ. 08/10/2004, BIPM 2004. No. 22.

10. Mitrofanov D.G. A comprehensive adaptive method for constructing radar images in dual-purpose control systems. // Proceedings of the RAS. Theory and control systems, 2006. No. 1. p. 101-118.

11. Mitrofanov D.G., Silaev N.V. An adaptive multi-frequency method for constructing a radar image of a fluctuating air target. // Radio engineering, 2002. No. 1. p. 53-60.

Claims (1)

A method for measuring the radial speed of an air target in the mode of tuning the carrier frequency from pulse to pulse according to a random law with a reduced signal-to-noise ratio, namely, that the number of pulse signals equal to the number of radiation frequencies used in the signal packet with frequency tuning is chosen equal to 2 ^k , where k is an integer taking a value from 6 to 8, the time Δt for radiation of a signal packet with frequency tuning is chosen no more than the interval of angular correlation T _y rotation of an air target in flight, amounting to its value is 5 ms, a random law of frequency change is used when emitting pulses with frequency tuning, for which a sequence of frequency values is used in the random access memory used in a packet of signals with frequency tuning from f ₀ to f ₀ + F _per with step Δf = F _per / (N-1), where f ₀ is the main carrier frequency of the probe signal of the centimeter range, F _lane = 150 MHz is the range in which the frequency is tuned from pulse to pulse, N is the number of frequencies used, the radiation frequency numbers are allocated as new law, in which the radiation time t _{n of the} pulse at the n-th frequency f ₀ + nΔf is determined by the formula

where T _and is the pulse repetition period inside the packet, selected on the basis of the requirement to ensure the uniqueness of the samples at the Doppler frequency in the entire range of possible radial velocities of the target;

- the serial number of the pulse at the n-th frequency, taking a value from 1 to N, repeating itself once within the burst of signals with frequency tuning, remember the order in which the frequencies are used for radiation, using the radar station for the time interval Δt emit the burst of signals described above with tuning carrier frequency, receive signals reflected from the target at n-th frequencies, where n is the signal frequency number, reduce the frequency of the received signals to an intermediate one, and quadrature phase detectors are extracted using quadrature phase detectors components of the received signals, convert the quadrature components into digital form using analog-to-digital converters, convert each signal reflected at the nth frequency into a complex form of the form

Where

- the amplitude of the signal reflected at the nth frequency;

- phase of the signal reflected at the nth frequency;

and

- the values of the quadrature components of the received signal at the nth frequency, when lowering the frequency of the received signals to the intermediate, take into account the value of the addition nΔf used when emitting the signal at the n-th frequency f ₀ + nΔf so that the frequency difference of the signal received at the nth frequency is f ₀ + nΔf + f _{d n,} where f _{q n} - Doppler frequency of the reflected signal to supplement n-th frequency due to target radial velocity, and the local oscillator signal (f ₀ -f _pr) + nΔf always determined only by the value of the intermediate frequency f and a corresponding _straight the magnitude of the Doppler ext forks F _{d n} , form a vector G of N elements, write the complex value to the nth element of the vector G

reflected at the nth frequency of the signal, a two-dimensional data matrix D is formed of N rows and Z = 2V _{p max} / dV + 1 columns, where V _{p max} is the maximum possible radial speed of the air target selected in advance, dV is the sampling interval of the radial speed, determining the accuracy of measuring the radial velocity, write the complex quantity into the element of the nth row of the zth column of the data matrix D

calculated by the formula

Where

is the complex value of the nth element of the vector G; z is the column number of the data matrix D, the data matrix D1 is obtained by performing the inverse fast Fourier transform with complex data vectors of each column of the data matrix D, characterized in that for each z-th column of the data matrix D1 the coefficient k _{z is} calculated by

Where

- assessment of the variance of the amplitudes of the z-th column of the generated data matrix D1,

- assessment of the mathematical expectation of the amplitudes of the z-th column of the generated data matrix D1,

- estimate of the third central moment of the amplitudes of the z-th column of the generated data matrix D1, find the column number k corresponding to the largest coefficient k _z , using the found value z _{max k} determine the target radial velocity estimate

according to the formula

and take this estimate as the measured value of the radial speed of the air target.

Images