RU2534217C1 - Radar method of detecting low-visibility unmanned aerial vehicles - Google Patents

Abstract

FIELD: radio engineering, communication.

SUBSTANCE: initial base probing frequency should be selected equal to 150 MHz and adjustment should be performed up to 6 GHz. After analysing reflections at different frequencies and detecting spectral response exceeding the threshold at one of the frequencies, radiation is switched from frequency adjustment mode to single-frequency mode which corresponds, based on frequency, to presence of spectral response from a low-visibility unmanned aerial vehicle (UAV). After switching to the detected estimated resonance frequency fr, results of detecting a low-visibility UAV in the corresponding range gate are repeatedly verified. The conformity of Doppler frequency Fd of spectral response in consecutively generated spectra of the same range gate, as well as the exceeding of a set detection threshold by the spectral response are verified. If in three consecutive spectra the spectral response from the UAV exceeds the threshold and its Doppler frequency Fd remains unchanged, a low-visibility UAV is detected in the corresponding range gate.

EFFECT: scanning the entire frequency range and high accuracy of detection.

3 dwg

Description

The invention relates to methods for radar detection of airborne objects (AT), in particular to methods for detecting unmanned aerial vehicles (UAVs) with low radar signature.

A known method of detecting air objects, including UAVs, which consists in emitting into space using a radar station (radar) of pulsed sounding signals, reflecting them from the VO, receiving the reflected signals by the radar antenna system, filtering the reflected signals by frequency to highlight reflections from moving VO against the background of reflections from stationary local objects, comparing filtered reflections with a threshold and in case of exceeding the established threshold - making a decision that a moving VO is detected [1-3].

This method is used in most radars of the old park and has the disadvantage that reliable detection is possible only in the case of reflections of electromagnetic waves (EMW) from typical objects with an effective scattering area (EPR) of the order of units of square meters. In the case of reflection of electromagnetic radiation from UAVs, whose EPR can be from tenths to thousandths of a square meter, the power of the reflected signals is not enough to exceed the threshold, and the detection of such objects is impossible.

There is also a method for detecting subtle BOs (including UAVs), which, in contrast to the method described above, assumes the accumulation of reflections from BOs obtained in different periods of radar pulse repetition [4]. Coherent accumulation (addition) of reflected signals allows you to exceed the detection threshold even in the case of a low reflectance of VO. To ensure coherent addition of the reflected pulse signals, the pulse repetition rate F is increased _and or, which is the same, the pulse repetition period T _{and is} reduced. The method requires that at a given speed of viewing the space, the minimum number of pulses N _{and min} received after reflection from the HE is sufficient to detect the HE with a given probability. When choosing a repetition rate, use the expression [4, p. 71-72, 89-90]

$$F_{\mathrm{and}}\ge \frac{N_{\mathrm{and}\text{min}}\Delta \beta \Delta \epsilon }{T_{\mathrm{about}bs}{\Theta }_{\beta 0.5}{\Theta }_{\epsilon 0.5}},\text{(one)}$$

where Δβ and Δε are the magnitudes of the sectors of the survey of space in azimuth β and elevation ε T _review - period of the review of space; Θ _β0.5 and Θ _ε0.5 are the width of the antenna pattern (BOTTOM) in the azimuthal and elevation planes at half power level.

This detection method is better than the typical one, but it does not allow to detect subtle UAVs efficiently, since no method has been established for the required number of accumulated pulses N _{and min} , and with an increase in F _, there is an ambiguity in determining the range to the detected object. In addition, in modern radars, the number of accumulated pulses does not exceed 10-100, which is clearly not enough to detect low-reflecting UAVs.

There is another method of detecting VO (UAVs) [4], in which the accumulation of the required number of pulses is achieved not only by increasing the repetition frequency F _and , but also by reducing the angular speed of rotation of the radar antenna. If you know the allowable number of pulses, the addition of energy of which ensures the detection of VO (UAV) with a probability not lower than the required, then with the number of revolutions of the radar antenna of the circular view per minute equal to n _a (taking into account only the azimuthal rotation of the antenna), according to [4, p .90] inequality must be satisfied to ensure accumulation

$$n_a\le \frac{60F_{\mathrm{and}}{\Theta }_{\beta 0.5}}{N_{\mathrm{and}\text{min}}\Delta \beta }.\text{(2)}$$

This means that the larger the required number of accumulated reflected pulses, the lower the antenna rotation speed should be. Therefore, the method of detecting VO (UAVs) involves reducing the speed of viewing space by slowing the speed of rotation of the antenna in sectors where the presence of weakly reflecting VOs, including UAVs, is assumed. Otherwise, the detection method adheres to traditional principles.

It should be emphasized that in modern radars this method is quite actively used. As an example, one can use the 9C18M1 all-round radar, in which the periods or rates of one round-trip antenna revolution (full view of the space) are 4.5 s, 6 s and 6-18 s at the lowest pace [5]. The time of irradiation of HE can vary in this radar from 0.02 to 0.053 s. The width of the bottom in azimuth is 1.6 °. The number of pulses that accumulate incoherently in accordance with the principle of operation of the 9C18M1 radar is from 16 to 84. The pulse repetition rate can be 0.8 kHz or 1.6 kHz. It is obvious that when detecting a type of UAV, it is recommended to use the lowest viewing rate with an accumulation of 84 pulses. However, this is the main drawback of the method - at a low pace of the review, the viewing time of the space increases significantly and the updating of radar information is rare, which negatively affects the results of the detection of HE.

Also known are detection methods that use, along with the Rayleigh and quasi-optical scattering region, the resonance scattering region [6, p.15-30]. If the product of the wave number k = 2π / λ, where λ is the wavelength, by the size of the L0 is much larger than 1 (kL >> 1), then the scattering region of the electromagnetic waves incident on the object is called quasi-optical. For kL << 1, the scattering region is recognized as Rayleigh. And only in the case kL≈1 is the scattering region considered resonance, in which the EPR can increase significantly [6].

Another known method for detecting VOs having a low reflectivity is based on this effect. In accordance with the principles of operation of the devices described in [7, pp. 74-82], as well as using traditional methods of correlation-filter processing [1-3,8], a known method for detecting HE, including weakly reflecting UAVs, is as follows. Select a frequency f ₁ corresponding to a wavelength λ ₁ commensurate with the size of the largest typical BO, generate high-frequency electromagnetic oscillations at a frequency f ₁ , modulate high-frequency electromagnetic oscillations at a frequency f ₁ in time, resulting in the formation of a pulsed probing radar signal, after which the pulse repetition period T _and similarly form the next pulse sounding radar signal and thus create a pack of N pulse sounding radar signals at a frequency f ₁ , double the probing frequency with a frequency multiplier, form a packet of pulsed sounding signals at a frequency of 2f ₁ with the same repetition period T _and then similarly form a packet of pulsed sounding signals at a frequency of 3f ₁ with a repetition period T _and and so on, i.e. successively increase the carrier frequency of a packet of sounding signals from N radio pulses up to the formation of a packet of sounding signals of N radio pulses at a frequency of Kf ₁ , where K is an integer corresponding to a wavelength commensurate with the size of the smallest BO, amplify the generated sounding radio pulses in power and emit they are put into space using the radar antenna system, the reflected signals are received sequentially with the help of the radar antenna system, digitized and recorded in random access memory the device of the amplitude of the received reflected signals of each repetition period, and the sampling period of the analog-to-digital conversion is chosen not less than 10-30 times shorter than the duration of the probe radio pulse, the entire set of reflected signals is divided into serial signals connected by the boundaries, but disjoint and equal in length to the range gates, range strobes are numbered within each repetition period from 1 to M, and each range strobe is equal in width to the duration of the probe radio pulse and τ _and , detect all recorded reflected signals with a digital phase detector to obtain quadrature components of the reflected signals, i.e. they convert the received digitized reflected signals into a complex form, carry out the coordinated processing of the digitized received signals within each range gate by convolution with a complex conjugate digitized sounding signal of the same repetition period, find the peak of the reflection response in each range gate to the maximum of its amplitude, record in the complex in the form of the values of the reflection response peaks of each repetition period in the random access memory, form for each k-th of K sounding frequencies I (for each kth burst of radio pulses) digital arrays of reflection response peaks of the same range gates and receive for reflections at each kth sounding frequency M arrays with N elements in each array, the Fourier transform operation is performed with elements of each array, in the result is a corresponding spectral array for each array, in which, in the presence of a VO at a range corresponding to the range strobe used, a spectral response will be generated whose frequency position corresponds to the Doppler frequency of the reflected VO signal, the spectral responses of the reflected signals are compared with a threshold value, and if the spectral response exceeds the threshold, the frequency of this spectral response is fixed as the frequency of its maximum component in the random access memory, and then a decision is made on the presence of the corresponding VO range, and in case of exceeding the threshold in the m-th array at all K sounding frequencies, VO is considered to be the typical type, and in case of exceeding the threshold only at one of the frequencies, the probe tion considered VO low reflection, unobtrusive, how is an unmanned aerial vehicle, and a Doppler frequency F _d of the spectral response is determined radial velocity detected at the formula V _r = F _d λ _k / 2, where λ _k - the wavelength of the probing signal k-th carrier frequency (i.e. in the k-th packet of radio pulses).

The described method for detecting inconspicuous UAVs is more effective than those indicated earlier. However, it also has disadvantages. Firstly, a multiple increase in the sounding frequency does not allow one to analyze reflections at all frequencies without exception, i.e. there are frequency ranges at which radiation and reflection analysis are not carried out, but at which resonance scattering can occur. Secondly, the decision to detect UAVs at one of the frequencies is not specified, which can lead to errors, since the appearance of a spectral response at one of the K frequencies in one of the M range gates can be the result of random fluctuations, surge surges in the equipment, short-term flight of a shell, meteorite, etc. Thus, the known method for detecting subtle UAVs needs to be improved.

The objective of the invention is the development and improvement of the known method for the resonant detection of subtle UAVs, providing a view of the entire frequency range (enumerating all wavelengths commensurate with the dimensions of the HE and their design elements) and clarifying the decision made on one implementation of the radiation of a packet of radar signals.

To solve this problem, it is proposed to carry out sequential tuning of the frequency of radio pulses (signals) from one packet to another (from the previous to the next) with a tuning step of Δf = 10 MHz. In this case, it is proposed to choose the base initial sounding frequency equal to 150 MHz (λ = 2 m), and carry out the tuning up to 6 GHz (λ = 5 cm). Another suggestion is that after analyzing the reflections of radio pulses at different sensing frequencies and identifying the occurrence of a spectral response exceeding the threshold at one of the frequencies, the radiation is transferred from a mode with a frequency tuning to a single-frequency mode, corresponding in frequency to the presence of a spectral response from a subtle (low reflectance) UAV. After switching to the detected estimated resonant frequency f _p (the wavelength λ _p corresponds to it), the results of detecting an unobtrusive UAV in the corresponding range gate are re-checked. The fact of matching (equality) of the Doppler frequency F _{d of the} spectral response in successively formed spectra of the same range gate, as well as the fact that the spectral response exceeds the established detection threshold, is also verified. If in several consecutive spectra (for example, in three) the spectral response from the UAV exceeds the threshold and its Doppler frequency F _d remains unchanged, then a decision is made to detect an imperceptible UAV in the corresponding strobe with a radial velocity V _r equal to V _r = F _d λ _p / 2.

Due to the proposed improved method, the main disadvantages of the prototype are eliminated, namely: viewing the entire frequency band from 150 MHz to 6 GHz in steps of 10 MHz eliminates the passage of resonance arising from irradiation of the UAV, and the transition to the radiation mode of the signals at the found resonant frequency f _p allows by comparison spectra to verify the reliability of the decision upon detection.

Let us explain the essence of the resonance scattering effect. There is a significant difference between the emerging resonant diffraction field and the diffraction fields of the quasi-optical and Rayleigh scattering regions. In the high-frequency approximation, the scattered electromagnetic field consists of the following components [6, p.21-30]: mirror reflections; the field in the discontinuity sections of the smooth surface of the VO (rib, wing edge) and in the discontinuity sections of the derivative function that describes the scattering surface; scattering at the light-shadow boundary, i.e. creeping waves; traveling waves arising from an oblique incidence of electromagnetic waves on weakly convex surfaces; reflections from concave parts of the surface of VO.

The main components of the scattered field in the resonance region are specular reflections, edge waves, and surface waves. In this case, creeping waves also play an important role, the contribution of which increases especially at the junction of the Rayleigh and resonance regions. When concentrated VOs are irradiated from directions where the specular and edge scattering are weak, creeping waves make a decisive contribution to the total backscattering. In this case, the total scattering field is formed mainly from the diffracted creeping electromagnetic wave and the field of the specularly scattered electromagnetic wave. The scattering diagram in the resonance region weakly depends on the angle of the VO in comparison with the quasi-optical region, but is also multi-lobe [6].

The intensity of reflection from each conductive object depends on its shape, size and wavelength of the probe signal. At low frequencies (kL << 1), most objects do not reflect electromagnetic waves, but only refracts them. With increasing frequency and transition to the resonance region (kL≈1), the EPR of the irradiated object sharply increases [7, pp. 81-83]. At L≈λ / 2, the ESR maximization is observed. The experimental dependences of the change in the EPR of the object (ball, cone, rod) on the change in the product kL given in [7, Figs. 3.22–3.24] confirm the above. In other words, if an object with a size of about λ / 2 is irradiated by a plane computer, then it acts as a parallel circuit with a resonance of currents, in which, with a change in frequency, a rapid change in the phase shift between current and voltage occurs. Such a rapid phase change is observed in metallic objects only near resonant sizes.

It is also known that modern UAVs can be either with a metal frame, or completely from composite and plastic materials [9, 10]. Therefore, their equivalent transverse radar size, determined by metallized elements (control board, engine, optical lens mount, antenna), can differ by tens or more times. It is especially difficult to detect UAVs if there are few metal parts in its design and equipment and their dimensions are small. In this case, the reflected electromagnetic waves can be fixed by the radar receiver only at certain angles and under the conditions of resonance reflection. And since the true transverse size of the reflecting element of the UAV is unknown, it is necessary to search for resonance by smoothly or stepwise changing the sounding frequency in steps of no more than tens of MHz.

To test the operability of the proposed method, special experiments were organized and conducted in anechoic chamber conditions. The investigated Orlan-3 UAV at the attacking angle (zero heading angle) was installed on the radio-transparent platform. A miniature radar with spaced receiving and transmitting antennas was used. During the experiment, the radiation frequency was gradually changed in increments of 10 MHz, and after receiving the reflected signals and digitizing them, the level of the received signal was fixed and a graph of the EPR versus frequency was plotted. The frequency varied from 300 MHz (λ = 1 m) to 3 GHz (λ = 10 cm). The experiment was repeated 3 times. EPR measurement results are shown in FIGS. 1-3. It can be seen from them that with an average EPR value of the order of σ _cf ≈0.03 m ² under resonance conditions, the EPR value reaches 0.16 m ² , which is 5 times more than the average EPR σ _cf. Resonance occurs at a frequency of the order of 2780 MHz, which corresponds to a wavelength of λ _p ≈ 10.8 cm. This means that the size of the reflecting element of the Orlan-3 UAV design at zero heading angle is approximately 5.4 cm. The characteristics formed in different experiments, indicates the reliability and stability of the obtained experimental results.

Thus, the proposed method for detecting subtle UAVs should consist of the following sequentially performed operations:

1. Packs of radio pulses of duration τ _and with a repetition period T _and are sequentially formed, and the number of radio pulses in each packet is equal to N (for further rapid Fourier transform, it is advisable to choose the number N equal to 2 ^S , where S = 8, 9, 10) the frequency of the radio pulses of the first packet is set to 150 MHz (λ = 2 m), and the carrier frequency of each subsequent next packet is increased by Δf = 10 MHz, the carrier frequency of the radio pulses is tuned until it reaches 6 GHz (λ = 5 cm).

2. They amplify the generated radio pulses by power and sequentially radiate them into space using the radar antenna system.

3. Consistently receive the reflected signals using the radar antenna system, digitize them using an analog-to-digital converter, and record the amplitudes of the received reflected signals of each repetition period into RAM (RAM), and the sampling period of the analog-to-digital conversion is selected as 10 -30 times shorter duration of the probing signal τ _and .

4. The entire set of digitized reflected signals recorded in RAM within each repetition period is divided into consecutive, interconnected, but disjoint and equal in length range gates, number range gates within each repetition period from 1 to M, and the duration of the gates is chosen equal to the duration sounding radio pulse τ _and .

5. All recorded reflected signals are detected using a digital phase detector to obtain quadrature components of the reflected signals, i.e. the received digitized reflected signals are converted into a complex form.

6. Within each m-th range gate, coordinated processing of the digitized received signals is carried out by convolution with the digitized complex conjugate probing radio pulse of the same repetition period.

7. The peak of the response of the reflections in each range gate is determined by the criterion of the maximum of its amplitude and the values of the response peaks of each m-th repetition period of each k-th burst of radio pulses are recorded in complex form.

8. For each k-th of K bursts of radio pulses, digital arrays of reflection response peaks of the same range gates number m are generated and for each k-th burst of radio pulses M arrays with N elements in each array are obtained.

9. Conduct the operation of the Fourier transform (fast Fourier transform) with the elements of each array of response peaks and obtain, for each array, the corresponding spectral array in which the VO spectral response is generated when the VO is actually located in the corresponding range gate (if in the mth range gate was in, then in the corresponding m-th spectral array arises its spectral component - the spectral response).

10. The spectral responses of the reflected signals in each spectral array are compared with a predetermined threshold value (level), and if the threshold is exceeded, the frequency of the corresponding spectral response of the m-th array of the k-th burst of radio pulses, which is taken as the Doppler frequency of the corresponding VO, is fixed in RAM and at the same time decide on the detection of an appropriate range of HE.

11. If the threshold is exceeded by the spectral response of the m-th array according to the results of the reflection analysis of all bursts of radio pulses, the detected VO is considered to be a typical VO with EPR of the order of square meters, and if the threshold is exceeded only on one of the probe frequencies (only in the spectral array of one separate RF pulse burst) further alter the radiation mode similar packs radiation from a single carrier mode resonance frequency f _p, equal to the carrier frequency burst on which is received one-time reflections (not repeat eesya for other packets) excess spectral response threshold.

12. In strict accordance with the operations described above, spectral arrays of reflections are generated and analyzed for packs of radio pulses emitted at a frequency f _p , and if the fact that the threshold is exceeded in three consecutive m-arrays (coincidence in range) and frequency coincidence of these spectral responses that exceed the threshold, they decide to detect an imperceptible UAV at the appropriate range.

13. The radial velocity of the kth burst of radio pulses of the VO or subtle UAV detected during the analysis of reflections is calculated from the Doppler frequency Fd of the corresponding spectral response that exceeded the threshold by the formula V _r = F _d λ _k / 2, where λ _k is the wavelength of the probe signal in k-th pack of radio pulses.

As follows from the description and essence of the method, it is really devoid of the disadvantages inherent in the prototype. The method provides an analysis of the entire frequency range at which modern UAVs reflect electromagnetic waves in the resonance scattering region. In addition, the method eliminates the adoption of random, unconfirmed decisions about the detection of UAVs. The method can be recommended for use in advanced radar search and detection of subtle BO, including radar with a cylindrical or conformal phased antenna array.

Information sources

1. Handbook of radar / Ed. M.I. Skolnik. Per. from English M .: Sov. Radio, 1967. Volume 1. The basics of radar. 456 s

2. Theoretical Foundations of Radar / Ed. POISON. Shirman. M .: Sov. Radio, 1970 .-- 560 p.

3. Okhrimenko A.E. Basics of radar and electronic warfare. Part 1. Basics of radar. M .: Military Publishing, 1983.- 456 p.

4. Guide to the basics of radar technology / Ed. V.V. Druzhinina. M., Military Publishing, 1967.768 p.

5. Radar target detection 9S18M1. Technical description. Book 1. General information. EP 1.005.029 TO, 1983. 152 s.

6. Radar characteristics of aircraft / Ed. L.T. Tuchkova. M .: Radio and communications, 1985.236 s.

7. Nebabin V.G., Sergeev V.V. Methods and techniques of radar recognition. M .: Radio and communication, 1984. p. 74-82 (prototype).

8. Radio-electronic systems. Directory. Fundamentals of construction and theory / Ed. POISON. Shirman. M .: Radio engineering, 2007.510 s.

9. Mosov S.P. Unmanned reconnaissance aircraft of the countries of the world: the history of creation, combat experience, current status, development prospects. Monograph. Kiev: Publ. Rumb House, 2008.160 s.

10. Vasily N.Ya. Unmanned aerial vehicles. Minsk: OOO Potpourri, 2003.272 p.

Claims (1)

The radar method for detecting stealth unmanned aerial vehicles, which consists in the formation of K packs of radio pulses with the same N number of radio pulses in each packet, with the same duration τ _and each radio pulse and the same T equal to _{and the} pulse repetition period, each radio pulse packet their carrier frequency, which is different from the others, is amplified by the generated radio pulses in power and sequentially radiate them into space using an antenna radar system station, the reflected signals are received using the antenna system of the radar station, they are digitized using an analog-to-digital converter, and the amplitudes of the received reflected signals of each repetition period are recorded in the random access memory, the sampling period of the analog-to-digital conversion is selected as 10 30 times less than the duration of the probing signal τ _and , share the entire set of digitized reflected signals recorded in random access memory Within each repetition period, the range gates, consecutively disjoint and of equal duration in length, number the range gates within each repetition period from 1 to M, and the strobe duration is chosen equal to the duration of the probe radio pulse τ _and all the recorded reflected signals are detected using a digital phase detector to obtain quadrature components of the reflected signals, i.e. the received digitized reflected reflected signals are converted into a complex form, within each mth range gate, the digitized received signals are coordinated by convolution with a digitized complex conjugate probing radio pulse of the same repetition period, the peak of the reflection response in each range gate is determined by the criterion of its maximum amplitude and write in complex form the values of the response peaks of each m-th repetition period of each k-th packet of radio pulses in the random access memory o, form for each kth of K bursts of radio pulses digital arrays of reflection response peaks of the same range gates by number m and obtain for each kth burst of radio pulses M arrays with N elements in each array, a conversion operation is performed with elements of each array of response peaks Fourier and get as a result for each array the corresponding spectral array in which the spectral response of the air object is formed when the air object is actually in the corresponding range gate, comparing t the spectral responses of the reflected signals in each spectral array with a predetermined threshold value (level) and, if the threshold is exceeded, the frequency of the corresponding spectral response of the m-th array of the k-th burst of radio pulses, which is taken as the Doppler frequency of the corresponding airborne object, is fixed in the random access memory and at the same time decide on the detection of an air object at an appropriate range, if the threshold is exceeded by the spectral response of the m-th array according to the results For analysis of the reflections of all bursts of radio pulses, the detected airborne object is considered to be an ordinary typical airborne object with an effective scattering area of the order of units of square meters, and the radial velocity of the airborne object detected during the analysis of reflections of the kth packet of radio pulses is calculated from the Doppler frequency F _{d of the} corresponding spectral response that exceeds the threshold by the formula V _r = F _d λ _k / 2, where λ _k is the wavelength of the probe signal in the k-th packet of radio pulses,
characterized in that the carrier frequency of the radio pulses of the first packet is set equal to 150 MHz, and the carrier frequency of each subsequent packet is increased relative to the frequency of the previous packet by 10 MHz, and the carrier frequency of the packets of radio pulses is tuned until it reaches 6 GHz, as well as when the threshold is exceeded by the spectral response of the m-th array as a result of the analysis of reflections of only one separate packet of radio pulses, i.e. only at one of the sensing frequencies, the further radiation mode is changed to the radiation mode of similar packs with a single carrier resonant frequency f _p equal to the carrier frequency of the packet, the reflections of which result in a one-time excess of the threshold value by the spectral response after processing the reflected signals belonging to additionally formed radio pulse packets at the carrier frequency f _p, and the order of processing the reflected RF pulse does not change, the fact of exceeding a check corresponding spectral array obtaining the spectral responses of the set threshold value and in case of coincidence of numbers spectral arrays in which exhibit excess of the threshold level, the three consecutive received packets reflected on the resonance frequency f _p radio pulses, and when the frequencies of these exceeded the threshold spectral responses make the final decision about the detection on the corresponding range stealth unmanned aerial vehicle.

Images