RU2414721C1 - Method for radar measurement of speed of an object - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: invention can be used in radar stations of different types to determine the speed of the radar set carrier relative the underlying surface, as well as measuring radial speed of observed objects. The disclosed method also enables its realisation in a multifunctional radar set as one of the operating modes.

EFFECT: high accuracy of measuring speed compared to the method of measuring radial speed based on Doppler filtration for different values of accumulation time of radar signals reflected from the observed object owing to use of wideband probing signals and correlation processing of the obtained range profiles of objects.

3 dwg

Description

The invention relates to radar and can be used in various types of radar stations (radar) to determine the own speed of the radar carrier relative to the underlying surface, as well as to measure the radial speed of objects of observation.

A known method of radar (RL) to determine the speed of the observed objects relative to the radar carrier by measuring the Doppler frequency shift of the radar signals reflected from the object. The method is used to measure the aircraft’s own speed relative to the underlying surface [1, p. 294-297], as well as to determine the radial speed of objects of observation in radars of various types, for example, onboard multi-function radars [2, p. 69-106] .

This method is based on the emission of coherent radar from the probing radar signals in the direction of the object, which may be a portion of the underlying surface or any other object, receiving the signals reflected from the radar object and measuring the Doppler frequency shift F _D = 2V _r / λ, where λ is the wavelength of the probing signals, V _r is the radial component of the relative speed of the object.

Three or four beams with non-coplanar longitudinal axes are usually used in Doppler self-velocity meters (DIS). For each beam measure its value V _r = F _D λ / 2 using the data from the on-board navigation system find the full speed vector of the aircraft.

In modern digital radars, Doppler shift measurement is performed using the fast Fourier transform (FFT). Moreover, the accuracy of the frequency measurement is approximately equal to the width of the FFT filter - ΔF _D. The accuracy of measuring the velocity ΔV _r is determined by the accuracy of measuring the Doppler frequency shift ΔF _D

ΔV _r = ΔF _D λ / 2.

In turn, the width of the FFT filter is inversely proportional to the accumulation time T of the RL array of signals needed to complete each FFT operation (ΔF _D = 1 / Tн).

Thus, the accuracy of the velocity measurement is proportional to the signal accumulation time, but is limited by various fluctuation errors of the reflected radar signals, which are determined by the characteristics of the radar itself, as well as by the properties of the irradiated object — its backscattering diagram.

The closest prototype method is a correlation method for measuring the aircraft’s own speed relative to the underlying surface. In the correlation speed meter (CIS) uses a spatio-temporal method for processing radar signals received simultaneously by three or four antennas spaced along the baseline with wide radiation patterns, the axes of which are oriented in the vicinity of the normal [1, pp. 297-299]. In this case, the baseline is oriented along the velocity vector of the aircraft. In CIS, the aircraft’s own speed is determined by searching for the maximum correlation function of the signals received by pairs of diversity antennas.

It was experimentally proved [3] that the correlation method can have higher accuracy than the Doppler method of measuring velocity. However, its use is limited by technical difficulties associated with the need to use a baseline orientation system for the aircraft velocity vector. In addition, this method requires its own antenna system, so it is difficult to implement it in multifunctional radars as one of the operating modes.

The objective of the invention is to develop a method of radar measurement of radial velocity, combining the properties of Doppler and correlation speed meters, having a higher measurement accuracy than the Doppler method, and the possibility of implementation as part of multi-function radar as one of the operating modes.

The essence of the invention lies in the fact that to measure the radial velocity of the object V _r in the direction of the object through the transmitting radar antenna emit the first and second sequences of N radio pulses having a duration τ, the same and constant initial phases, the period T _p . In each sequence, the filling frequency of the radio pulses is discretely linearly changed from pulse to pulse in increments of Δf = 1 / τ, but with different signs so that the radio pulses of one of the sequences (either the first or second) have an increasing filling frequency f _a (n) = f _o + Δf (n- (N-1) / 2), and the radio pulses of another sequence have a decreasing filling frequency f _b (n) = f _o -Δf (n- (N-1) / 2), where n - pulse numbers of sequences (n = 0, 1, 2, ..., N-1), f _o - average frequency of filling of radio pulses f _o = (f (0) + f (N-1)) / 2.

The beginning of the second sequence is delayed relative to the beginning of the first sequence for the time T _c = NT _p (emit the second sequence after the end of the first) or T _c = T _p / 2 (every n-th pulse of the second sequence is emitted after the n-th pulse of the first).

After the emission of each radio pulse through the receiving antenna of the radar in the selected time interval receive radar signals reflected from the irradiated object. Moreover, the beginning of this interval is delayed relative to the beginning of the emitted pulse for a time t ₀ > τ, and the end of the interval is delayed for a time t ₁ <T _p if T _c = NT _{p is set} or t ₁ <T _p / 2 if T _c = T _p / 2. The value of t ₀ choose the same for all pulses of both sequences.

The received signals with a high pulse filling frequency f _a (n) and f _b (n) are transferred to the zero frequency with simultaneous conversion to a complex form and the received signals are digitized with a time discrete Δt, which does not exceed the duration of the radio pulse τ. The obtained digital samples are recorded in two-dimensional arrays of complex samples of input signals s _a (m, n) for a sequence with increasing pulse filling frequency f _a (n) and s _b (m, n) for a sequence with decreasing pulse filling frequency f _b (n) , where m are the numbers of samples (m = 0, 1, 2, ..., M-1). The number of such samples is selected from the condition M = (t ₁ -t ₀ ) / Δt, and the first samples (m = 0) correspond to time instants t ₀ .

Then, using the obtained arrays s _a (m, n) and s _b (m, n), the radial velocity V _r is determined by selecting this velocity V _r from the interval V _min ÷ V _{max of} possible values of the desired velocity V _r according to the extremum of the correlation module the functions

,

,

where F _b ^∗ (m, k) is the complex conjugate array F _b (m, k),

moreover, F _a (m, k) = Ф {u _a (m, n)} and F _b (m, k) = Ф {u _b (m, n)}, where k = 0,1,2 ,. .., N-1; Ф {} is the discrete Fourier transform, u _a (m, n) = s _a (m, n) e ^{iφ (n)} and u _b (m, n) = s _b (m, Nn-1) e ^{iφ (Nn -1)} , and the values of the phase coefficients φ (n) are calculated for each selected value of the velocity V according to the formula φ (n) = - 4πТ _p (V / c) (n + 1/2) (f _o + Δf (n- ( N-1) / 2)), where c is the speed of propagation of radio waves in space.

The technical result of the proposed method of radar measurement of the radial velocity of an object is to increase the accuracy of its measurement compared to the method of measuring radial velocity based on Doppler filtering at equal values of the accumulation time of the radar signals reflected from the object of observation, through the use of broadband probing signals and correlation processing of the obtained long-range profiles objects.

The probing radar signals used in the proposed method belong to broadband signals of the type of discrete frequency sequences (PDP) and are described in [4, pp. 289-307]. Signals of the PDP type are sequences (bursts) of N radio pulses of duration τ. Each n-th pulse of the PDP (n = 0, 1, 2 ..., N-1) has its own filling frequency, selected from a frequency series having a step Δf = 1 / τ, and an average frequency value f _o . In the general case, the order of sorting frequencies in the PDP from pulse to pulse can be any, but this order must be taken into account when processing the received radar signals. In this case, the reception of radar signals reflected from the object of observation is carried out in the gaps between the emitted radio pulses of the PDP.

The period of the following radio pulses T _p is selected based on the required unique range of the radar observation. The total duration of one DCH T _dchp = NT _p .

A separate pulse of the PDP allows one to obtain a range resolution ΔR = s τ / 2. Special processing of the entire PDP (range compression) using a matched filter [4] provides a range resolution Δr = cτ / (2N).

High resolution allows you to get long-range profiles (DP) of extended objects, which are areas of the underlying surface that fall into the opening of the antenna beam, and other objects. DP is a sequence of amplitudes and phases of the signals reflected from the elements of the object, spaced from each other by a distance Δr along the line of oblique range.

Signals of the type of PDP and their processing algorithms have the property that if the object of observation in the time interval T of the _PDP moves along the range line relative to the aircraft with a constant speed V _r , and the law of change in the frequency of filling pulses of the PDP is linear, then the observation object is shifted range in the output DP relative to the true range. This property is a consequence of the Doppler frequency shift of the radar signals.

The direction of this displacement is determined by both the sign of the radial velocity of the object and the sign of the linear tuning of the frequency of the pulses in the PDP. Therefore, if an object has a certain velocity V _{r of} one or another sign, then in the output DP for an PDP with an increasing frequency of filling from pulse to pulse, the image of the object will shift in range in one direction, and for a PDP with a decreasing filling frequency in the opposite direction. Moreover, the displacement of the image of the object relative to its true position on the range line, which would be in the absence of movement, is proportional to the value of V _r .

If the value of speed V _{r is} measured in some other way, for example, using a navigation system, then this distance shift of the object in the DP can be compensated during processing by phase correction of the received radar signals. The indicated property of radar signals of the type of PDP is laid in the basis of a new radar method for measuring the radial velocity of objects of observation.

In the proposed method, the probe signal is two practically simultaneously emitted PDPs in which the frequency of filling of the radio pulses is linearly changed from pulse to pulse, with a constant discrete Δf = 1 / τ, but with different signs. Moreover, the radio pulses of one sequence have an increasing frequency of filling (direct PDP) equal to f _a (n) = f _o + Δf (n- (N-1) / 2), where n = 0, 1, 2, ... N-1 , f _o - average frequency of filling of radio pulses f _o = (f (0) + f (N-1)) / 2.

Radio pulses of a different sequence have a decreasing filling frequency (reverse PDP) equal to f _b (n) = f _o -Δf (n- (N-1) / 2). Moreover, to measure the speed, there is enough radiation and processing of one pair of AFP - direct and reverse. The relative order of the forward and reverse PDPs in the pair does not matter.

In the method, you can use two options for the emission of pairs of PDPs (see figure 1). In the first embodiment, the second PPP is emitted immediately after the end of the first one after a time equal to T _p (a pair of successive PPPs). In the second embodiment, every nth pulse of the second PPP is emitted after the nth pulse of the first PPP after a time equal to T _p / 2 (a pair of embedded PPPs). The sign of the linear frequency modulation of the first PDP can be any, and the sign of the second PDP must be opposite to the sign of the first. The accuracy of measuring the speed in both variants of the radiation of PDP at equal values of T _{p is the} same, but in the first version (successive PDPs), the total time required to measure the speed is two times longer than in the second version.

After the radiation of each probe radio pulse, signals are reflected from the irradiated object by the receiving radar antenna. In both PDPs, the time interval for the reception of radar signals has a fixed and identical position relative to the probe pulses. The beginning of the interval of reception of radar signals is delayed relative to the beginning of the emitted pulse for a time t ₀ > τ, and the end for a time t ₁ <T _p if sequential PDPs are used (T _c = NT _p ) and t ₁ <T _p / 2 if nested DCP (T _s = T _p / 2). Moreover, the values of t ₀ and t ₁ must be constant for all pulses in a pair of AFP. The duration of the reception interval determines the length of the DP within which the object or its part is located.

Moreover, if the proposed method is used to measure the aircraft’s own speed and the object of observation is a portion of the underlying surface, then as long as possible the reception interval is selected to increase the accuracy of the speed measurement. At the same time, in order to save computational operations, it is impractical for the resulting DP corresponding to the reception interval to go beyond the antenna spot on the underlying surface. In the observation and measurement of air objects rate value t ₀ and t _and is selected so that the reception was in the range of only the object which velocity is measured.

Received radar signals are transferred from a high pulse filling frequency f _a (n) and f _b (n) to zero frequency with simultaneous conversion to a complex form. Next, the received signals are converted from analog form to digital using a two-channel analog-to-digital converter (ADC) having a conversion time (cycle) Δt, the value of which is chosen no more than the duration of the radio pulse τ. For each reception interval receive M complex samples of the signal with numbers m = 0, 1, 2, ..., M-1, where M = (t ₁ -t ₀ ) / Δt. Moreover, the first samples (m = 0) correspond to time instants t ₀ .

The obtained digital samples are stored in the form of two-dimensional arrays separately for each PP. In this case, the radar signals for the PDP, in which the frequency of filling of the radio pulses increases from pulse to pulse (f _a (n)), are stored in the array s _a (m, n). The radar signals of another PPP from a pair in which the filling frequency decreases from pulse to pulse (f _b (n)) are recorded in the array s _b (m, n).

After recording all the samples of the radar signals (for n = 0, 1, 2, ..., N-1 and m = 0, 1, 2, ..., M-1) in the arrays s _a (m, n) and s _b (m , n) for one pair of PDPs determine the value of the radial velocity V _r .

In accordance with the property of the PDP signals described above, the true value of the speed will be such that, after the phase correction of the input radar signals, the output signals in the PD for the forward and reverse PDPs will be identical with accuracy to noise. The difference between the initial random phases of the signals reflected from the radar object for the forward and reverse PDPs does not matter for the operation of the method.

будет иметь свой экстремум (F_b ^∗(m,k) - массив комплексно сопряженный массиву F_b(m,k)).To determine the desired speed V _r on a predetermined interval V _min ÷ V _{max of} all possible values of speed V _r find a value of speed V for which the module of the correlation function

will have its extremum (F _b ^∗ (m, k) is an array complex conjugate to the array F _b (m, k)).

This value of speed V is taken as the true value of speed V _r (closest to it). The extremum of the function C (V) on the interval V _min ÷ V _{max is} unique.

To calculate the value of the function C (V) for each value of V, when searching for its extremum, output arrays of complex samples of signals F _a (m, k) and F _b (m, k) are used. Moreover, the output arrays F _a (m, k) and F _b (m, k) for each value of V are obtained using the discrete Fourier transform of the samples of signals recorded in the rows of the intermediate arrays u _a (m, n) and u _b (m, n ):

где k=0, 1, …, N-1.

where k = 0, 1, ..., N-1.

In this case, the intermediate arrays u _a (m, n) and u _b (m, n) necessary for obtaining output arrays are preliminarily obtained by phase correction of the input signal arrays using the expressions u _a (m, n) = s _a (m, n) e ^{iφ (n)} and u _b (m, n) = s _b (m, Nn-1) e ^{iφ (Nn-1)} .

The phase coefficients φ (n) used in the above expressions are calculated for the selected value of velocity V by the formula

φ (n) = - 4πT _p (V / s) (n + 1/2) (f _o + Δf (n- (N-1) / 2)), where c is the speed of propagation of radio waves in space.

The search for an extremum is performed step by step using some algorithm for searching for an extremum of a function (sequential search, the method of successive approximations, etc.). It should be borne in mind that the extremum search method affects the accuracy of determining the speed.

The output arrays F _a (m, k) and F _b (m, k) used in the processing process contain M private DPs with N range samples in each. For all ADC readings, their private DPs are constructed, which can partially overlap in range if the conversion cycle is Δt <τ.

The accuracy of the considered method of measuring the speed is proportional to the duration of the PDP used to irradiate the object, that is, the value of NT _p , which is similar to the accumulation time in the DIS. At the same time, the method uses correlation processing of arrays of output signals F _a (m, k) and F _b (m, k), which are long-range profiles of high-resolution objects, which increases the accuracy of determining the speed.

When measuring the aircraft’s own speed, the objects of observation are sections of the underlying surface, which have a large extent (hundreds and thousands of meters). Their DPs consist of a large number of elements (hundreds and thousands with a range resolution of about one meter), which improves the accuracy of speed measurements.

The considered method of measuring the radial velocity of an object can be used in a multifunctional radar with digital signal processing.

Figure 1 shows the timing diagrams of one pair of probing signals of the type of PDP (a - sequential PDP, b - embedded PDP). Figure 2 presents the functional diagram of the radar that implements the proposed method. Figure 3 presents the algorithm for processing radar signals in the radar signal processor.

The diagram of figure 2 shows the radar transceiver 1 (PRLS), frequency synthesizer 1 (MF), two ADCs (3 and 4, respectively), buffer random access memory 5 (BOSU) and radar signal processor 6 (RLS).

PPLS 1 with the first output is connected to the first input of the midrange 2, and the second and third outputs are connected to the inputs of the ADC 3 and ADC 4, respectively, which are connected by their output groups to the first and second groups of inputs of the BOZU 5, the group of outputs connected to the group of inputs of the PRLS 6, a group of outputs connected to a group of radar outputs, the first and second outputs connected respectively to the first input of the radar station 1 and to the second input of the midrange 2, while the third output of the radar station 6 is connected to the control input of the BOZU 5.

A radar sensing system generates and emits radar-sensing radar signals in the direction of the observation object, receives radar signals reflected from the radar object and their analog processing and can be performed as an appropriate device (see, for example, [2, p. , the structural diagram of which is given in the appendix.

MF is designed to generate signals at a given intermediate frequency, necessary for the formation and processing of radar signals such as PDP. MFs are used in PRLS instead of the reference quartz oscillator (laser in the diagram in the appendix is highlighted with a dark background), which forms a signal of intermediate frequency (f _pr ) of constant value. The midrange generates N frequencies with a tuning discrete Δf and an average frequency f _pr , while the selection of the required frequency is made according to control codes received from external devices.

To build the midrange, any of the currently existing integrated digital frequency synthesizers can be used, for example, ADF4117, ADF4360-5, etc. (see, for example, www.analog.com).

Two ADCs convert the quadrature of the radar signals at zero frequency from the output of the radar signal to the digital form. BOSU is designed to record digital samples of signals from the ADC outputs into arrays s _a (m, n) and s _b (m, n), which are then used to determine the radial velocity of the object. The nodes of the ADC and BOZU in figure 2 replace the same nodes in the radar scheme from the application (marked with a dark background). PRLS calculates the value of the radial velocity of the object using the algorithm shown in Fig.3.

With the technical implementation of the method, the ADP201V5 digital signal processing module with the ADMDDC2WB-L analog input submodule manufactured by Instrumental Systems CJSC can be used to implement the functions of the ADC, BOSU and PRLS.

Two ADCs installed in the ADMDDC2WB-L submodule have a maximum clock frequency of 70 MHz. BOSU and PRLS are implemented in the ADP201V5 module, which has a large amount of RAM and high performance. (The module includes 5 TigerSHARC ADSP-TS201S processors with a total peak performance of 15 Gflops and two Xilinx Virtex4 FPGAs.)

The system depicted in figure 2, operates as follows. For each measurement of the radial velocity value, the radar-frequency radar detector generates and radiates two AFPs in the direction of the object — one with increasing frequency of filling of the radio pulses, the other with decreasing (their mutual order can be any). In this case, the second PPP (PPP 2) can be emitted after the end of the first (PPP 1) or superimposed on the first (diagrams figa and figb, respectively).

To obtain PDP 1 and PDP 2, an MF is used that generates reference signals with frequencies selected from the grid f _pr (n) = f _pr + Δf (n- (N-1) / 2). The choice of the necessary frequency is carried out according to the control commands received in the midrange from the radar. In this way, an PPD is formed with a given discrete and a linear frequency modulation sign, as well as a sequence (sequential PDPs or embedded PDPs).

Signals reflected from the radar object with the help of radar receiving, receiving, amplifying and transferring to zero frequency with simultaneous conversion to a complex form. At the same time, only the total antenna receive channel (Σ) and the corresponding receive path are used in the PRLS (see the appendix). Radar signals are transferred to the zero frequency using coherent detectors (QGD, see the appendix), using the midrange output signals with a frequency f _pr (n) as reference.

Quadrature radar signals at zero frequency from the outputs of the radar control signals are fed to the inputs of two ADCs. Digital readings of the radar signals from the outputs of the ADC are fed to the inputs of the BOSU. Arrays of digitized radar signals are recorded in the BOSU.

After the radiation of a pair of PDPs by the command received from the radar in the BOSU, the recording of digital samples of the radar signals is stopped. After that, the transmission of the signal array generated in the BOSE to the PRLS begins.

The array of signals formed in the BOSU contains arrays s _a (m, n) and s _b (m, n), which are then used to determine the speed of an object in the radar control system according to the algorithm shown in Fig. 3. Simultaneously with the transmission and processing of this array of signals in the radar control system or with some delay, a new cycle of work begins, including the emission and reception of radar signals for a new pair of PDPs and the formation of a new array of signals in the BOZU to determine the next speed reference.

Information sources

1. Radio control systems / Ed. V.A. Weitzel. M .: Drofa, 2005 (analogue).

2. Multifunctional radar systems / Ed. B.G. Tatarsky. M.: Bustard, 2005.

3. Borkus MK, Black A.E. Correlation meters of ground speed and drift angle of aircraft. - M .: Owls. radio, 1973.

4. Cook C., Bernfeld M. Radar signals (theory and application). M .: Sov. Radio, 1971.

Claims (1)

A method for radar measuring the radial speed of an object, including radiation through a transmitting antenna of a radar station (radar) in the direction of the irradiated object of the first and second sequences of sounding radio pulses, each of the sequences containing N radio pulses having a duration τ, the same and constant initial phases, the repetition period T _p, discretely varying from pulse to pulse frequency radio pulses filling increments Δr = 1 / τ linearly, but with different signs, so that adioimpulsy one of the sequences (either first or second) have increasing frequency filling _{f a (n) = f o} + Δf (n- (N- 1) / 2) and RF pulses other sequences have a decreasing fill rate f _b (n) = f _o -Δf (n- (N-1) / 2), where n are the pulse numbers of the sequences (n = 0, 1, 2, ..., N-1); f _o is the average frequency of filling of the radio pulses (f _o = (f (0) + f (N-1)) / 2), while the beginning of the second sequence is delayed relative to the beginning of the first sequence for a time T _c = NT _p (the second sequence is emitted after the end of the first) or T _c = T _p / 2 (each n-th pulse of the second sequence is emitted after the n-th pulse of the first), and after the emission of each radio pulse through the radar receiving antenna in the selected time interval, radar signals reflected from the irradiated object are received, moreover start of this delay interval ayut respect to the emitted pulse at time t _0> τ, and the slot end is delayed by the time t ₁ <T _p, if T _c = NT _p, or t ₁ <T _p / 2, where T _c = T _p / 2, this value t _{0 is} chosen the same for all pulses of both sequences, the received signals from a high pulse filling frequency f _a (n) and f _b (n) are transferred to the zero frequency with simultaneous conversion to a complex form, these signals are digitized with a time discrete Δt, not exceeding the duration of the radio pulse τ, and record the obtained samples in arrays of complex input signals s _a (m, n) for a sequence with increasing pulse filling frequency f _a (n) and s _b (m, n) for a sequence with decreasing pulse filling frequency f _b (n), where m are sample numbers (m = 0 , 1, 2, ..., M-1), and the number of samples is selected from the condition M = (t ₁ -t ₀ ) / Δt, and the first samples (m = 0) correspond to times t ₀ , then using the resulting arrays s _a (m, n) and s _b (m, n) determine the value of the radial velocity V _r , by choosing this value of the velocity V _r from the interval V _min ÷ V _{max of} possible values of the desired speed V _r according to the extremum of the correlation module the functions

, where F _b * (m, k) is the complex conjugate array F _b (m, k), while F _a (m, k) = Ф {u _a (m, n)} and F _b (m, k) = Φ {u _b (m, n)}, where k = 0, 1, 2, ..., N-1; Ф {} is the discrete Fourier transform, u _a (m, n) = s _a (m, n) e ^{iφ (n)} and u _b (m, n) = s _b (m, Nn-1) e ^{iφ (Nn -1)} , and the values of the phase coefficients φ (n) are calculated for each selected value of the velocity V according to the formula φ (n) = - 4πТ _p (V / c) (n + 1/2) (f _o + Δf (n- ( N-1) / 2)), where c is the speed of propagation of radio waves in space.

Images