• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    Since the detection method implements a waveform of the FMCW type, the emission waveform is produced according to a recurrence pattern (10) of given period Tr covering a given transmission frequency band of given width. B, each pattern being cut according to a given number P of sub-patterns (21, 22, 23, 24) of duration Tr / P covering an excursion frequency band AF = B / P, said sub-patterns being spaced between they have a frequency difference equal to AF. The radar performs: a first remote compression process performing a low-resolution distance compression at the scale of each recurrence pattern (10) from a B / P fraction of said corresponding width B transmission band; the frequency band covered by each of said sub-patterns (21, 22, 23, 24); one of a Doppler treatment over a given number N of successive recurrences so as to form P ambiguous distance-Doppler maps at low-resolution distance, said maps being segmented according to different speed domains; a second differentiated-resolution distance compression process according to the speed domain to which the relative speed of said target relative to said radar belongs.
    公开号:FR3055049A1
    申请号:FR1601224
    申请日:2016-08-11
    公开日:2018-02-16
    发明作者:Pascal Cornic；Bihan Patrick Le；Jean Michel Quellec
    申请人:Thales SA；
    IPC主号:
    专利说明:

    055 049
    01224 ® FRENCH REPUBLIC
    NATIONAL INSTITUTE OF INDUSTRIAL PROPERTY © Publication number:
    (to be used only for reproduction orders)
    ©) National registration number
    COURBEVOIE © Int Cl 8 : G 01 S 13/536 (2017.01)
    PATENT INVENTION APPLICATION
    A1
    Date of filing: 11.08.16. ©) Applicant (s): THALES— FR. © Priority : @ Date of availability of the request: 16.02.18 Bulletin 18/07. @ Inventor (s): CORNIC PASCAL, LE BIHAN PATRICK and QUELLEC JEAN MICHEL. (56) List of documents cited in the preliminary search report: See the end of this brochure References to other related national documents: ®) Holder (s): THALES. O Extension request (s): © Agent (s): MARKS & CLERK FRANCE General partnership.
    MULTIPLE RESOLUTION FMCW RADAR DETECTION METHOD AND RADAR IMPLEMENTING SUCH A METHOD.
    FR 3 055 049 - A1
    The detection method using an FMCW type waveform, the emission waveform is produced according to a recurrence pattern (10) of given period Tr covering a band of emission frequencies of given width B, each pattern being cut out according to a given number P of sub-patterns (21, 22, 23, 24) of duration Tr / P covering a frequency band of excursion AF = B / P, said sub-patterns being spaced between them with a frequency difference equal to AF. The radar performs:
    a first distance compression processing carrying out distance compression at low resolution on the scale of each recurrence pattern (10) from a fraction B / P of said emission band of width B, corresponding to the band of frequency covered by each of said sub-patterns (21,22, 23, 24);
    - one of a Doppler processing on a given number N of successive recurrences so as to constitute P distance-Doppler maps ambiguous at low distance resolution, said maps being segmented according to different speed domains;
    - A second distance compression processing with differentiated resolution according to the speed range to which the relative speed of said target belongs to said radar.
    22 23 24
    MULTI-RESOLUTION FMCW RADAR DETECTION METHOD AND RADAR IMPLEMENTING SUCH A METHOD
    The present invention relates to a radar detection method, using an emission wave of the FMCW type, with multiple resolution. It also relates to a radar implementing such a method. The invention is particularly applicable in the field of radars for motor vehicles. More generally it can be applied to FMCW radars.
    Speed cameras for cars have been introduced for driving assistance functions, more oriented towards comfort, such as adapting to cruising speed for use on motorways, known as ACC (Adaptive Cruise Control), or "Stop and Go" in urban driving. They use millimeter waves, in particular the band 76-81 GHz.
    Thanks to the evolution of technologies, current applications also target anti-collision type safety functions, and it is even envisaged in the relatively short term to achieve a fully autonomous vehicle, the perception of the environment being ensured by the association a number of sensors, based on different technologies: radar, video, infrared in particular.
    Because of its all-weather capabilities, the radar remains a predominant sensor in this context and its detection and discrimination capabilities must be extended to guarantee the overall reliability of the system. With regard to anti-collision, the radar sensor must in particular be able to distinguish among the fixed objects it detects, those which correspond to elements of road infrastructure, from those which correspond to vehicles immobilized on the track which potentially constitute a risk of collision. In this context, it is particularly fundamental that it does not generate false alarms that can lead to braking or an emergency avoidance maneuver, without any real cause, in particular when the vehicle is traveling at high speed. This imposes increased sensitivity and capacity for discrimination, making it possible to understand the situation in front of the vehicle at a long distance, typically greater than 200 m. It may also be necessary to detect roadsides.
    In this context, the distance resolution must be very large on fixed objects, which results in a large number of distance boxes to cover the range of the radar and therefore very significant digital processing needs to perform the processing in time. real. In the case of a beamforming radar by calculation, which must process several angular directions simultaneously, this need is multiplied by the number of beams to be processed.
    Furthermore, with regard to automobile radars, the cost of the sensor is extremely constrained, and the computing resources are consequently limited. For an automotive application, it is therefore necessary to find solutions to optimize the use of processing resources according to the context.
    A technical problem to be resolved is to obtain sufficient discrimination capacities, while limiting the impact on the processing volume. To date the problem has not been resolved or is unsatisfactory.
    Auto radars use differentiated waveforms for short range and long range, the two modes being exclusive. The waveforms are essentially frequency modulated, called FMCW, or combined with frequency hopping then called, FSK FMCW. FMCW waveforms can alternate ramps of different frequencies over time, either to optimize the emission band to the desired range, or to resolve the distance / speed ambiguity problems inherent in this type of radar. This results in a loss of waveform efficiency since the different emission patterns share the radar integration time.
    In urban mode, at reduced speed, distance resolution is preferred. It is typically less than a meter, which corresponds to a large "instantaneous" emission band, typically several hundred megahertz. However, the distance domain is small, and the number of distance boxes to be processed remains modest.
    On high-speed motorways, speed resolution is preferred, which implies a high Doppler resolution, typically of the order of 25 Hz, corresponding to a discrimination of 5 cm / sec for a radar operating at 76 GHz.
    The latter case corresponds to operation in ACC mode in which the separation of the vehicles in front of the radar carrier is effected by speed. The distance resolution is relatively low, typically several meters, and does not allow sufficient obstacle discrimination to automatically initiate high-speed emergency braking. Here again, the number of distance boxes to be processed remains modest due to the low resolution, even if the distance domain is greater.
    Neither of these two operating modes therefore makes it possible to perform an anti-collision function at high speed, which requires both a high distance resolution and a high speed resolution.
    An object of the invention is in particular to solve the aforementioned problem. To this end, the subject of the invention is a method of radar detection of a target using a waveform of the FMCW type, the waveform on transmission being produced according to a recurrence pattern of given period. Tr covering a transmission frequency band of given width B, each pattern being cut according to a given number P of sub-patterns of duration Tr / P covering a excursion frequency band AF = B / P, said sub-patterns being spaced apart from each other by a frequency difference equal to AF, said method performs at least:
    a first distance compression processing effecting distance compression at low resolution on the scale of each recurrence pattern from a fraction B / P of said transmission band of width B, corresponding to the frequency band covered by each of said sub-patterns;
    Doppler processing on a given number N of successive recurrences so as to constitute P ambiguous distance-Doppler maps at low resolution in distance, said maps being segmented according to at least two speed domains, one of which corresponds to fixed targets;
    a second distance compression processing from said distance maps, the resolution of which depends on the speed range to which the expected speed of said target belongs.
    In said first processing, said compression consists for example, for each sub-pattern, of a distance compression by direct demodulation of the signal emitted by the received signal, followed by a Fourier transform on M / P points according to a synchronous demodulation leading to each pattern to the constitution of M / 2P distance boxes, M / P being an integer.
    Said Doppler processing performs, for example, on said N successive recurrences, a Fourier transform on the Doppler axis for each of said distance boxes and for each sub-pattern, resulting in the constitution of said P distance-Doppler maps, each of dimension M / 2P distance boxes by N Doppler boxes.
    Said resolution is for example maximum for fixed targets. Said expected speed corresponding to a fixed target, the second processing performs for example compression according to the maximum resolution by carrying out the coherent integration of said P distance-Doppler maps according to each of the distance filters centered on said resolution sub-patterns -,
    C being the speed of light. Said coherent integration is for example carried out after compensation for the delays and frequency differences between said sub-patterns.
    In a possible embodiment, said expected speed corresponding to a moving target, said second processing performs a coherent integration on all of said distance-Doppler maps to give a new distance-Doppler map, said coherent integration being carried out in such a way that '' a coherent integration is carried out on each grouping of distance-Doppler maps corresponding to sub-patterns covering adjacent frequency bands.
    In one possible embodiment, said expected speed being high, above a given threshold, the second processing performs the non-coherent integration of at least two distance-Doppler maps of the same distance resolution.
    The order of said sub-patterns is for example changed from one broadcast burst to the next.
    The period Tr of said recurrence pattern is for example adjusted as a function of the speed of the carrier of said radar and of the emission frequency in order to comply with the condition of unambiguous Doppler on targets at maximum approach speed.
    A formation of beams by calculation being carried out on reception on each of said distance-Doppler maps at low resolution in distance, the second processing is for example carried out for each beam.
    In a particular implementation, said distance resolution is a function of the direction of the beam in reception.
    The distance measurement is for example lifted by measuring the Doppler frequency.
    The invention also relates to an FMCW type radar capable of implementing such a method. This radar being for example able to equip a motor vehicle.
    Other characteristics and advantages of the invention will become apparent with the aid of the description which follows, made with reference to the appended drawings which represent:
    - Figure 1, a FMCW type radar waveform used in the prior art;
    - Figure 2, a FMCW type radar wave according to the invention;
    - Figure 3, a FMCW type radar waveform according to the invention where the sub-patterns are presented in a different order;
    - Figure 4, two successive radar emission bursts of a wave according to the invention;
    - Figure 5, an illustration of the distance filters associated with the sub-patterns of a wave according to the invention;
    - Figure 6, a radar wave according to the invention where the sub-patterns are arranged in a particular order;
    - Figure 7, an example of decomposition of distance-Doppler maps according to speed domains.
    Figure 1 recalls the prior art. More particularly, it illustrates the modulation of the transmitted frequency, this modulation being of the FMCW type.
    We generally consider an automobile radar mounted on a vehicle, the beam of which is oriented towards the front thereof. This radar implements a linearly frequency modulated waveform according to the FMCW principle, well known to those skilled in the art, in accordance with FIG. 1.
    A first processing is carried out on the short time, on the scale of the recurrence 10 of duration Tr, aiming to separate the signals received by a direct demodulation followed by a digitization of the signals and an FFT on P points, corresponding to the duration of frequency ramp 1.
    This operation performs compression on the distance axis. At the end of this processing, the signals are distributed in P / 2 ambiguous distance boxes, the beat frequency f b of a signal received at distance D and Doppler
    F d written:
    f b = 2 - ^ + F d (1)
    Or :
    B is the transmitted frequency band;
    C is the speed of light;
    D is the distance from the target;
    T is the duration of the frequency ramp;
    F d is the Doppler frequency.
    The distance frequency is given by:
    f iisc = (2) and the distance resolution AD is determined, according to the classical relation, by:
    Furthermore the Doppler frequency F d is given by the relation F d = - £
    At where V r is the radial speed of the target and λ is the wavelength of the emission signal.
    Following this “short time” processing, a second “long time” processing is carried out, consisting for each ambiguous distance box, in performing a second FFT (or DFT) on N points, corresponding to the duration NT r of a Doppler burst 20.
    At the end of this processing, a distance-speed (or distance-Doppler) map is thus obtained comprising for P / 2 ambiguous distance boxes, N unambiguous speed filters. For each cell of the ambiguous distance-speed map that gave rise to a detection, the estimation of the unambiguous Doppler frequency F d makes it possible, from equation (1), to deduce unambiguously the distance frequency of each signals detected by:
    fdist - fb - Fd
    The Doppler resolution AF d is given by AF d = -, where N is the number of recurrences in the Doppler burst.
    The term distance-Doppler or distance-speed will be used interchangeably hereinafter, these two terms being equivalent because the measurement
    Doppler corresponds to a speed measurement.
    In such a radar, the processing requires N fast Fourier transforms (FFT) on P points for distance compression and P / 2 FFT on N points for Doppler compression, which implies significant processing means.
    For example, for a maximum detection distance of 384 meters and a distance resolution of 1.5 meters, P = 512.
    If we impose that there is no Doppler ambiguity, for a maximum speed V rmax of 130 km / h, or 36 m / sec, and a wavelength λ of 3.9 mm, the duration maximum of the recurrence is in this case Tr = 27 psec.
    By considering a Doppler burst duration typically of the order of 40 ms and choosing a recurrence number N in power of 2, we arrive at N = 2048 for a recurrence duration Tr = 19.5 psec.
    It is therefore necessary to perform 2048 FFT on 512 points and 256 FFT of 2048 points, for each radar reception channel, during the duration of the Doppler burst, ie 40 ms, the processing having to be carried out in real time. For a beamforming radar by calculation comprising Q beams, it is necessary for each element of the distance-speed map of dimension P / 2xN, to additionally perform angular compression by an FFT on Q points. The volume of computation to be carried out in real time is thus very large with regard to the technological resources available, in particular if it is desired to use low-cost processing means.
    FIG. 2 illustrates the treatment carried out by the method according to the invention. To carry out the treatment, we start from an emission waveform, of the FMCW type, with sub-patterns as illustrated in FIG. 2. More precisely, a waveform is produced on the transmission according to a recurrence period pattern Tr, each pattern consisting of P sub-patterns 21, 22, 23, 24 of duration Tr / P corresponding to a ramp of excursion frequency AF = B / P, the patterns being spaced between them by a frequency difference AF, the assembly thus covering the total band B. In the example of FIG. 2, P = 4.
    Having laid down the principle of cutting the recurrence pattern into P sub-motives, before engaging in radar processing, the choice is made of the FMCW waveform as a function of the speed of the radar carrier and the traffic situation. Then in a second step, the maximum distance resolution of the radar is determined.
    Beforehand, the radar recurrence frequency is therefore chosen as a function of the speed of the carrier to guarantee unambiguous Doppler for the targets of interest having the maximum approach speed.
    On motorways, typically, this maximum speed corresponds to the speed of the carrier or the speed of approaching fixed obstacles, typically 130 km / hour in France. We can of course consider another maximum speed value.
    On the road, this speed corresponds to the relative maximum speed of two vehicles facing each other, typically 180 km / hour in France. Other speed values are of course possible.
    Thus, the maximum value of the waveform repetition period is determined so as to respect the Nyquist criterion for the maximum relative speed, which corresponds to the relation:
    4L v rmax where Vrmax is the maximum relative speed of the targets considered, λ the wavelength of the radar and Tr the period of recurrence.
    The repetition frequency MTr can thus be automatically adjusted between two patterns, as a function of the speed of the carrier and the frequency of emission F to comply with the condition of unambiguous Doppler on targets at maximum approach speed.
    The maximum distance resolution of the radar is also determined so as not to generate a migration from distance to distance from space to targets on targets at maximum approach speed during the coherent integration time corresponding to the duration of the Doppler burst NT r , typically 40 ms .
    For example, for a maximum speed Vrmax of 130 km / h, or 36 m / sec, the maximum distance resolution AD is 1.5 meters, corresponding to the relative displacement of the target during the burst of duration 40 ms.
    The maximum useful band emitted by the radar is in this case = -, or in r 2ΔΏ 'this example: B = 100 MHz.
    As an example, table 1 below indicates, in a motorway situation, the maximum theoretical values of the duration of the recurrence and of the useful transmission band according to the carrier speed.
    Carrier speed 50km / h 70km / h 90km / h 110km / h 130km / h Maximum relative speed of 50 70 90 110 130 highway targets,corresponding to the approach speed of fixed targets km / h km / h km / h km / h km / h Maximum duration ofrecurrence Tr 70 ps 50 ps 39 ps 32 ps 27 ps Maximum distance resolutionAD on fixed targets 0.6 m 0.8m 1 m 1.2m 1.5m Maximum useful band B 250 180 150 125 100 on fixed targets MHz MHz MHz MHz MHz
    Table 1
    In the same way, table 2 below indicates in road situation, the 5 maximum theoretical values of the duration of the recurrence and the useful emission band for the fixed targets and the moving targets, according to the carrier speed, in considering a transmission frequency of 76 GHz. We consider here a vehicle arriving in the opposite direction with a maximum speed of km / h.
    Carrier speed 10 km / h 30 km / h 50 km / h 70 km / h 90 km / h Maximum relative speed onfixed targets 10 km / h 30 km / h 50 km / h 70 km / h 90 km / h Maximum relative speed on 100 120 140 160 180 moving targets km / h km / h km / h km / h km / h Maximum duration ofrecurrence Tr 35 ps 29 ps 25 ps 22 ps 19 ps
    Distance resolutionmaximum AD on fixed targets 0.1m 0.3 m 0.5 m 0.8m 1 m Maximum useful band onfixed targets 1.5 GHz 500MHz 300MHz 180MHz 150MHz Distance resolutionmaximum AD on targetsmobiles 1.1 m 1.3m 1.6 m 1.8m 2 m Maximum useful band onmoving targets 140MHz 115MHz 95 MHz 80 MHz 75 MHz
    Table 2
    These parameters make it possible to set according to different speed ranges 5 of the carrier and according to the conditions of use of the radar, typically on road or on highway:
    - The total band to be transmitted to obtain the maximum distance resolution on fixed targets;
    - The maximum useful band on moving targets;
    - The maximum duration of the radar recurrence.
    Once these choices have been made, the total transmission band B and the duration of the recurrence T r are fixed.
    The recurrence period T r is then divided into P sub-patterns of duration 15 T / P, each pattern corresponding to a ramp of excursion frequency
    AF = B / P of slope B / T, the patterns being spaced apart by a frequency difference AF, the whole thus covering the total band B over a time T. For the following description P will be taken equal to 4.
    The frequency ramps can follow one another continuously as illustrated in FIG. 2.
    FIG. 3 illustrates an example where the sub-patterns 21, 22, 23, 24 are placed in any order.
    FIG. 4 illustrates the case of two consecutive bursts 41, 42 where the order of the sub-patterns has changed between the two bursts. Preferably, the order of succession of the sub-patterns is modified from one Doppler burst to the next, so as to reduce the probability of interference between radars of the same type, as illustrated in FIG. 4
    There may be a dead time in the transmission, from one recurrence to the next (case where the ramp time is less than the recurrence period, ie T <Tr) or more generally between two successive patterns . However, these dead times must be of equal duration and the continuity of the transmission phase must be respected.
    From the waveforms and sub-patterns thus generated, the treatment applied in reception is presented below. The latter notably performs the following operations:
    - Carry out for each sub-pattern, a first processing in distance, said to be low resolution, consisting of a compression by distance by direct demodulation of the signal emitted by the received signal, followed by a fourier transform (FTT) on M / P points (M / P being integer) according to a conventional synchronous demodulation processing of an FMCW radar, this processing resulting for each reason in the constitution of M / 2P distance boxes;
    - Carry out on a sequence of N successive recurrences a Fourier transform (FFT or DFT) on the Doppler axis for each distance box and for each sub-pattern, leading to P ambiguous distance-speed matrices, of dimensions M / 2P distance boxes by N speed boxes, if applicable, apply a beam forming treatment by calculation on the distance-Doppler maps thus obtained;
    - From the P different ambiguous distance-speed maps thus obtained and for each beam, apply a second distance processing with optimized resolution and adjusted according to the speed of the expected moving targets;
    - Detect the targets present in each of the beams;
    - Remove the distance-speed ambiguity on the targets detected in the distance-speed maps, from the unambiguous Doppler frequency.
    We will now describe in more detail the processing applied in reception covering all these operations.
    In a first step, a first distance processing is carried out, consisting of low-resolution distance compression. More precisely, for each sub-pattern and at each recurrence 10 of the same Doppler burst 20, a distance compression processing is carried out by synchronous demodulation followed by an FFT (or DFT) over the duration T / P of the sub-pattern, resulting to a separation of the different targets according to M / P ambiguous distance boxes at low resolution -.
    For example, for a total band of 100 MHz and a cutting of the waveform into four sub-patterns (P = 4), the distance resolution at the output of the FFT is 6 meters, corresponding to a band of 25 MHz . The signal received over the short time of the ramp of index p after synchronous demodulation, p varying from 0 to P-1 is written:

    2DB, 2v r CT + λ (3)
    Or :
    - B is the total transmission frequency band;
    - P is the number of ramps constituting the complete pattern 30 (see Figure 3);
    - F is the lowest frequency of the complete pattern;
    - T is the duration of the complete pattern;
    - D is the distance from the target to the radar;
    - v r is the speed of movement of the target;
    - t c is the “short” time corresponding to a duration of analysis on a single sub-reason;
    - T r is the period of recurrence of the complete pattern corresponding to the "long"time;
    - n is the index of the ramp being analyzed;
    - It is the speed of light;
    - k p is the integer between 0 and P-1 corresponding to the frequency offset k p B / P of the ramp of index p with respect to F;
    - λ is the wavelength of the emission signal considered to be constant relative to the Doppler effect.
    The indices p and n indicate the rank of the sub-patterns or the ramps, we will use the terms of index or rank afterwards.
    We notice :
    φ Ορ = -2π. 2 (J) + p v rT / F) Çp + t | a phase at the origin for the sub-reason of index p and:
    2.BD 2V f b = - h F d with F d = the beat frequency, identical for all
    CT Λ sub-patterns 21, 22, 23, 24.
    The signal received over the short time of the ramp of index p after synchronous demodulation, p varying from 0 to P-1 is then written:
    s (tç) = e ~ j <P0P ' e ~ 2jnfb te (4) whose Fourier transform on M points can be written for the filter of index m corresponding to the maximum of energy:
    sin (7Γ (f b - y))
    In the above expression, the only term dependent on the subscript p is
    Thus, we can more simply express the response at the output of the Fourier transform for the sub-motif of rank p in the form:
    ç = / Z e ~ J <Pop ° p, m lx m c where K m is independent of the index p of the sub-motif.
    By generalizing to all the recurrences of the Doppler burst of index n we obtain:
    S p , m (n) = K m e ~ ^ p ^ (5) with:
    φ Ορ (η) = - 2n . 2 {D ^ + ^ / P > {F + k p B -} (6) where D (n) is the distance from the target to the nth recurrence, that is:
    D (n) = D q + v r nT r (7)
    D o being the distance at the origin of the times.
    For the sub-motif of index p, we can thus write the phase of the signal at the nth recurrence:
    2 (D 0 + v r n R ) (F + k p .Î) 2 (pv r T / P) (f + k p .j) <Ρορ (η) = ~ 2π · --- 2ττ-- 10 which can be simplified in the form:
    <Po P (n) = ~ - 2π ~ 2nF d nT r (8)
    Λρ r by considering that _ 2v r F ~ 2v r ^ F + k p . ^ D C ~ C 'and by posing
    We note that φ Ορ (ή) contains:
    - a Doppler term, common to all the sub-patterns, -2nF d nT r , and; a phase term, different for each pattern: _ 2n ^ f20
    In a second step, an FFT (or DFT) is performed on each distance cell corresponding to the instrumented domain and for each sub-pattern, along the Doppler axis over the duration of the Doppler burst corresponding to the long time NT r .
    Similarly to the Fourrier transform on the distance axis, we can simply express the response at the output of the Fourrier transform on the Doppler axis for the sub-pattern of rank p in the form:
    Um, n lP = K m , n e “* p) (W)
    K'm, n is independent of the index p of the sub-motif.
    The index m corresponds to the distance filter of index m, centered on the ambiguous distance -, of resolution -.
    The index n corresponds to the speed filter of index n, centered on the Doppler frequency - of resolution -.
    NT r NT r
    The subscript p corresponds to the sub-motif p of the waveform.
    At this stage, the treatment produced P matrices or "distance-Doppler maps"
    Ambiguous PCs, with distance resolution PAD = - and Doppler resolution âF d = including M / 2P x N elements U mnp .
    In a third step, these distanceDoppler maps are separated into at least two sub-maps in the Doppler domain, the first corresponding according to the carrying speed to the possible domain of fixed targets, and the second or seconds corresponding to the domain of mobile targets according to or not a breakdown into sub-speed domains.
    The estimation of the Doppler bounds in the domain of fixed targets is obtained by knowing the speed of the carrier V, the aperture angle and the pointing angle of the radar beam and the wavelength λ.
    Indeed, for a radar beam with an angular aperture in azimuth ΔΘ lighting the road along the axis of movement of the vehicle and pointed in the direction θ = 0, the Doppler domain of the corresponding fixed targets is between:
    _ 2Vcos (4Θ / 2). p _ 2V dmin ~ χ dmax ~
    Typically, for Δθ = 3 °, V = 36 m / sec and λ = 3.9 mm, we have:
    Fdmin = 18492 Hz
    Fdmax = 18518 Hz
    The Doppler band corresponding to fixed obstacles is limited to 25 Hz, corresponding to a single filter if the duration of the Doppler burst is 40 ms, corresponding to a Doppler resolution of 25 Hz.
    For an angular aperture radar beam in Azimuth ΔΘ illuminating the road along the axis of movement of the vehicle and pointed relative to the axis θ = 0 by an angle a greater than ΔΘ, the Doppler domain of the corresponding fixed targets is between:
    „_ 2Vcos (α-Δθ / ΐ). „_ 2Vcos (α + Δθ / 2 ~) '' dmin J θ * '' iimax -
    All other things being equal, for a 10 ° depointing, the Doppler band is then 168 Hz, corresponding to 7 filters.
    This example shows that the Doppler domain in which it is necessary to perform distance processing with the maximum resolution is very low, and can be isolated easily. In particular, fixed targets can be easily isolated.
    The field of moving targets can itself be broken down into different fields of relative speeds over which the rest of the processing will be differentiated as will be described later.
    This processing suite includes a second distance compression processing performed from the P distance-Doppler maps obtained following the first distance compression processing. This second treatment is a function of the expected relative speed of the target. In particular, one type of processing is applied to fixed targets and another type of processing is applied to moving targets.
    We begin by describing the second distance processing, applied to the elements of the distance-Doppler maps corresponding to the fixed targets.
    In this second processing, the radar aims to compress the signal according to the maximum resolution:
    CC
    Δΰ = - = 2B 2PâF
    This treatment is applied to a first limited part of the distance-Doppler map, corresponding to the fixed targets, or to a number of Doppler filters Nf, such that Nf is much less than N (Nf "N). It consists in carrying out a coherent summation, that is to say in amplitude and in phase, on the distance axis of the P ambiguous distance-Doppler maps.
    For this, with reference to FIG. 5, each distance filter 51, 52, 53, 54
    PC PC centered on m— and in width - is broken down into P adjacent filters 531, 532, 533, 534 in width—, centered on the distance:

    2B p varying from 0 to P-1.
    P is equal to 4 in the example of FIG. 5, each filter being broken down into four filters.
    In order to carry out the coherent summation making it possible to compress the signals in distance filters 531, 532, 533, 534 of resolution C / 2B, it is necessary to compensate for the phase differences between the P distance-Doppler maps which result from the frequency shift and the time delay between the different sub-patterns 21, 22, 23, 24.
    The phase compensation to be applied for each sub-pattern of rank p, relating to the distance filter 53 of rank m and to the speed filter of index n is:
    - + 4π γ- ~ for the frequency offset and;
    λρ p — T
    - 2π For this 9 U 'concerns the time delay.
    These phase terms correspond respectively to compensation with respect to a distance and to a target speed situated respectively in the center of the distance filter 531, 532, 533, 534 and in the center of the speed filter considered.
    We thus obtain a base of P unit vectors V m , n , p such that:
    4πΒ
    Â — T
    IZ = p a p y m, n, p c
    The compression on the distance axis is carried out by carrying out the sum of the projections of the P vectors U mnv on the P vectors V mirliP according to the following relation:
    W m , n, p
    P-l ^ m, n, p ^ m, n, p p = 0 either:
    P-l. ™ = Σ p = 0
    Km, n e
    4nD
    J (m, p + 2πV -A — T P NT r
    At the end of this processing, a single distance-Doppler map of dimension M / 2 distance boxes by Nf speed boxes is obtained. If the radar comprises several reception channels, a beamforming treatment by calculation is then applied to this distance-speed map (or distance10 Doppler).
    We will now describe the second distance compression processing applied to the elements of the distance-Doppler maps corresponding to the moving targets. The processing example described is applied to high relative speed targets.
    In this second processing, the aim is to compress the signal according to an intermediate resolution AD = (Q being an integer multiple of 2) adapted to their relative speed, then a non-coherent summation of the resulting distance-Doppler maps is carried out.
    The compression treatment in principle is identical to the treatment described for fixed targets, but decomposes the distance filter not into P filters of width £ -, but into Q filters of width -.
    It is applied to a second limited part of the distance-Doppler map, corresponding to high speeds, for example on N ve Doppler filters, such as N ve < N. It can moreover be carried out on different speed domains with different AD resolutions across the entire mobile target domain.
    We perform partial compression in distance to go from one
    PC OC P P resolution - to a resolution - by combining - covering patterns - bands of
    2B 2B Q Q adjacent frequencies of width y, the whole covering a band | ·.
    This operation is performed on all distance-Doppler maps from low-resolution distance processing obtained from all the sub-patterns, by grouping maps from the patterns whose bands are adjacent and cover a band |.
    This gives Q ambiguous distance-speed maps of dimension MQ / 2 distance boxes by N ve speed filters. An inconsistent summation of the Q distance-speed maps, that is to say in power, is then carried out, which gives a gain in frequency diversity linked to the difference in the carrier frequencies of the sub-patterns used.
    For example, considering the case where the waveform is broken down into four sub-patterns (P = 4), the assembly covering a band of Frequency B as illustrated in FIG. 6, it is possible to carry out a treatment producing Q = 2 distance-Doppler maps with a distance resolution of β
    - by combining on the one hand the patterns 21, 22 of rows 1 and 4, and on the other hand the patterns 24, 23 of rows 2 and 3.
    We then perform a quadratic summation of the two maps thus obtained with a frequency diversity of |.
    This step is described in more detail below. The objective is to decompose, in the P distance-speed maps resulting from the low-resolution distance processing, each distance filter 51, 52, 53, 54 centered on m— and of width
    PC oc - in Q adjacent width filters centered on the distances:
    œ (m- ±) + j (l + q)) C 2B q varying from 0 to Q-1.
    We consider an integer number of emission sub-patterns corresponding to adjacent frequency sub-bands of width, the whole covering a continuous band of width from the frequency F of radar emission.
    Noting:
    j the ranks of these sub-patterns, for j varying from 0 to - 1;
    - The wavelength corresponding to the frequency
    and 'T
    - and the delay of the sub-pattern of row j compared to the sub-pattern of the first row (index 0);
    at the end of the Doppler processing, the signal relating to the pattern of rank j corresponding to the distance D o and to the Doppler F <j is written:
    JT _ J7 a ii P u m, n, j ~ '' m.nP and the phase compensation to be applied for each sub-pattern of row j, relating to the original distance filter of row m and to the speed filter of row n is:
    47îD + —for the frequency offset and;
    That
    - 2π - N J r as regards the temporal delay of the sub-pattern j with respect to the origin of the times of the recurrence in progress.
    These phase terms correspond respectively to compensation with respect to a distance and to a target speed situated respectively in the center of the distance filter and in the center of the speed filter considered.
    We thus obtain a base of unit vectors V m , n, j such that:
    z 47rD m g
    V = e λ ι v m, nj c a · n T + 2n— ^)
    The compression on the distance axis is carried out by realizing the P / Q sums of the Q vectors U min j on the P vectors V m> n j according to the following relation:
    Wm, n, q ~ 'U m , n, j ^ m, n, jj = 0 either:
    m, n, q ï -l ^ Cn, n i
    7 = 0. z 47TO 0. ^ a i F dK A nD -Ji-T ^ + Zn-Ap—) j (
    At the end of this processing, a distance-Doppler matrix of dimension MQ / 2 distance boxes is obtained by N ve Doppler filters. The operation is repeated on the remaining PP / Q sub-patterns to form Q distanceDoppler matrices of dimension MQ / 2 distance boxes by N ve Doppler filters. Finally, we obtain a new distance-Doppler map with increased distance resolution.
    If the radar comprises several reception channels, a beamforming treatment by calculation is then applied to these Q distance-speed cards. A non-coherent summation of the different matrices thus obtained is then produced for each beam to obtain a single matrix comprising MQ / 2 boxes distances by N ve Doppler filters.
    In a following step, for the elements of the distance-Doppler maps corresponding to the moving targets of the lowest relative speeds, the non-coherent summation of the P distance-Doppler maps obtained by the low-resolution distance compression processing is carried out.
    This treatment is applied to the last limited part of the distanceDoppler map, corresponding to low speeds, for example on Nvf Doppler filters, such as Nvf <N, with N = Nf + Nv e + Nvf.
    If the radar comprises several reception channels, a beamforming treatment by calculation is previously applied to the P distance-speed maps. A non-coherent summation of the different maps thus obtained is then carried out for each beam to obtain a single map comprising M / 2 distance boxes by Nvf Doppler filters.
    For all the distance-speed maps obtained, target detection then clearance of distance-speed ambiguity using the unambiguous Doppler frequency and registration of the targets detected on the distance axis for each beam.
    FIG. 7 symbolically represents the shape of the decomposition into distance / Doppler filters of the field instrumented by the radar after the treatments described above. The abscissa axis represents the distance, the origin 70 corresponds to the position of the radar. The ordinate axis represents the relative speed of the targets relative to the radar. A distance / Doppler 80 filter is characterized by a width in distance and a width in speed.
    The decomposition is carried out according to distance boxes whose dimension is adjusted according to the relative speed of the targets. The width in speed remains for example constant. For this purpose, the speed domain is divided into several zones or domains 71, 72, 73, 74, 75 and for each of these zones corresponds a width in distance of the distance / speed filter. In addition to the fixed target domain, several speed domains are taken into account.
    A first speed domain 71 corresponds to mobile targets at low speed, this domain corresponds to ACC type applications. These are targets approaching the vehicle carrying the radar, rolling in front of it but slower. A second speed domain 72 covers the moving targets at higher speed. A third speed domain 73 covers moving targets at even higher speed. A fourth domain 74 corresponds to fixed targets and a fifth domain corresponds to moving targets at high speeds in approach in the opposite direction. This case corresponds to crossing vehicles and does not exist on motorways in particular.
    The distance-Doppler map in FIG. 7 shows that the smallest distance boxes 81 are assigned to fixed targets, that is to say that the distance resolution is maximum for fixed targets and lower for mobile targets. By adjusting the width of the distance boxes to the speed, the volume of processing and calculations to be performed is significantly reduced. Thus, for the first domain 71, filters of wider width are used, which limits the number of filters and therefore the number of calculations.
    Advantageously, the instrumented range and the distance resolution can also be adjusted in a differentiated manner as a function of the direction of the beam considered and of the distance to further limit the volume of calculations.
    Finally, we present an example of application where we consider the vehicle carrying the radar traveling on a highway at a speed V = 130 km / h.
    According to Table 1 presented above, the maximum useful band on fixed target is 100 MHz, corresponding to a distance resolution of 1.5 m, and the maximum duration of the recurrence is 27 ps for a Doppler burst duration of 40 ms .
    One can choose for example a duration of recurrence of 19.5 ps to carry out the Doppler treatment by an FFT on N = 2048 points. The waveform can be divided according to Figure 6 into four 25 MHz bandwidth sub-patterns, all covering a band of 100 MHz.
    Considering a 3 ° beam opening in azimuth 10 ° offset from the vehicle axis, the range of fixed targets occupies seven speed filters on which processing is carried out at maximum distance resolution. We can for example cut the remaining speed domain into two sub-bands typically covering:
    from 0 to 65 km / h for the first;
    and from 65 to 129.6 km / h for the second;
    on which a distance compression treatment is carried out respectively with a resolution of 6 m and 3 m, corresponding to a frequency band of 25 MHz and 50 MHz.
    Compared to a radar carrying out a compression in distance with the maximum resolution of 3 m on the whole of the speed domain, the number of distance boxes is approximately halved on half of the domain and by four on the other half, this which corresponds overall to a reduction in the computation volume of 60%.
    By automatically adapting the waveform and the associated processing according to the context, in particular according to the speed of the carrier and the conditions of implementation, the invention advantageously makes it possible to optimize simultaneously the performance of the radar and the resources. used to detect both moving and stationary objects across the entire radar range. It allows maximum distance resolution on fixed obstacles and lower distance resolution on mobile obstacles which are separated by Doppler processing, which minimizes the computational load. In “low resolution distance” mode, the probability of target detection is further favored by frequency diversity.
    The invention also makes it possible to reduce the risks of mutual disturbances between radars of the same type, due to the diversity of the waveforms capable of being used at a given time by these radars. It also makes it possible to optimize the efficiency of the waveform, because it is not necessary to interleave different waveforms to treat the whole of the distance-speed domain. Finally, it does not increase the volume of the electronics or the complexity of the radar. In fact, the method according to the invention can be implemented by existing radars without modifying their reception circuits or their processing circuits. In particular, the treatment is simple and easy to implement.
    A radar implementing the method according to the invention can advantageously equip a motor vehicle, for detecting fixed or mobile obstacles present in the environment of the vehicle. In particular, the radar can be used for an anti-collision function.
    权利要求:
    Claims (16)
    [1" id="c-fr-0001]
    1. A method of radar detection of a target using an FMCW type waveform, characterized in that the emission waveform being produced according to a recurrence pattern (10) of given period Tr covering an emission frequency band of given width B, each pattern being cut according to a given number P of sub-patterns (21, 22, 23, 24) of duration Tr / P covering a frequency band of excursion AF = B / P, said sub-patterns being spaced apart by a frequency difference equal to AF, said method performs at least:
    a first distance compression processing carrying out distance compression at low resolution on the scale of each recurrence pattern (10) from a fraction B / P of said emission band of width B, corresponding to the band of frequency covered by each of said sub-patterns (21, 22, 23, 24);
    Doppler processing on a given number N of successive recurrences so as to constitute P ambiguous distance-Doppler maps at low resolution in distance, said maps being segmented according to at least two speed domains (71, 72, 73, 74, 75) of which one corresponds to the fixed targets;
    a second distance compression processing from said distance maps, the resolution of which depends on the speed range (71, 72, 73, 74, 75) to which the expected speed of said target belongs.
    [2" id="c-fr-0002]
    2. Method according to claim 1, characterized in that in said first processing said compression consists, for each sub-pattern (21, 22, 23, 24) in distance compression by direct demodulation of the signal transmitted by the received signal, followed of a Fourier transform on M / P points according to a synchronous demodulation leading for each pattern to the constitution of M / 2P distance boxes, M / P being an integer.
    [3" id="c-fr-0003]
    3. Method according to claim 2, characterized in that said Doppler processing performs, on said N successive recurrences, a Fourier transform on the Doppler axis for each of said distance boxes and for each sub-pattern, resulting in the constitution of said P distanceDoppler cards, each of dimension M / 2P distance boxes by N Doppler boxes.
    [4" id="c-fr-0004]
    4. Method according to any one of the preceding claims, characterized in that said resolution is maximum for fixed targets.
    [5" id="c-fr-0005]
    5. Method according to claim 4, characterized in that, said expected speed corresponding to a fixed target, the second processing performs compression according to the maximum resolution by achieving coherent integration of said P distance-Doppler maps according to each of the distance filters ( 531, 532, 533, 534) centered on said sub-patterns (21, 22, 23, 24) of resolution -, C being the speed of light.
    [6" id="c-fr-0006]
    6. Method according to claim 5, characterized in that said coherent integration is carried out after compensation for delays and frequency differences between said sub-patterns.
    [7" id="c-fr-0007]
    7. Method according to any one of the preceding claims, characterized in that, said expected speed corresponding to a moving target, said second processing performs a coherent integration across all of said distance-Doppler maps to give a new distance-Doppler map , said coherent integration being carried out in such a way that a coherent integration is carried out on each grouping of distance-Doppler maps corresponding to sub-patterns (21, 22, 23, 24) covering adjacent frequency bands.
    [8" id="c-fr-0008]
    8. Method according to any one of the preceding claims, characterized in that, said expected speed being high, the second processing performs the non-coherent integration of at least two distance-Doppler maps of the same distance resolution.
    [9" id="c-fr-0009]
    9. Method according to any one of the preceding claims, characterized in that the order of said sub-patterns (21, 22, 23, 24) is changed from a transmission burst (41) to the following (42) .
    [10" id="c-fr-0010]
    10. Method according to any one of the preceding claims, characterized in that the period Tr of said recurrence pattern (10) is adjusted as a function of the speed of the carrier of said radar and of the transmission frequency to respect the condition of no -Doppler ambiguity on targets at maximum approach speed.
    [11" id="c-fr-0011]
    11. Method according to any one of the preceding claims, characterized in that a beam formation by calculation being carried out on reception on each of said distance-Doppler maps at low resolution in distance, the second processing is carried out for each beam.
    [12" id="c-fr-0012]
    12. Method according to any one of the preceding claims, characterized in that said distance resolution is a function of the direction of the beam on reception.
    [13" id="c-fr-0013]
    13. Method according to any one of the preceding claims, characterized in that the ambiguity on the distance measurement is removed by measuring the Doppler frequency.
    [14" id="c-fr-0014]
    14. FMCW type radar, characterized in that it is capable of implementing the method according to any one of the preceding claims.
    [15" id="c-fr-0015]
    15. Radar according to claim 14, characterized in that it is able to equip a motor vehicle.
    [16" id="c-fr-0016]
    16. Radar according to claim 15, characterized in that it is used for an anti-collision function.
    1/5
    Long time: Doppler treatment over the duration of the burst (N recurrences)
    类似技术:
    公开号 | 公开日 | 专利标题
    FR3055049A1|2018-02-16|MULTI-RESOLUTION FMCW RADAR DETECTION METHOD AND RADAR USING SUCH A METHOD
    EP3315994B1|2021-03-10|Multibeam fmcw radar, in particular for a motor vehicle
    EP0863409B1|2004-08-04|Method and device for continuous wave frequency modulation radar detection with range-velocity ambiguity resolution
    EP1851568B1|2016-05-25|Method for detecting in bistatic mode by passively operating on non-co-operating radioelectric transmissions
    EP3279690B1|2020-05-13|Method for measuring the height of a target relative to the ground by a moving radar, and radar implementing such a method
    EP2831615B1|2017-01-18|Device for active and passive electromagnetic detection with a low likelihood of interception
    FR2461991A1|1981-02-06|FREQUENCY MODULATION AND FREQUENCY-MODIFIED RADAR INSTALLATION, ABOVE THE POTENTIAL SIGNALS, HAVING THE ROLE OF WARNING TO A DRIVER WHICH HE HAS TOO CLOSE TO ANOTHER VEHICLE
    EP0681190A1|1995-11-08|Method and systemfor discrete radar detection
    FR2901366A1|2007-11-23|METHOD FOR DETECTING REFLECTORS OF ELECTROMAGNETIC IMPLUSION
    FR2905765A1|2008-03-14|DEVICE AND METHOD FOR ESTIMATING THE DIMENSIONS OF A PARKING PLACE, MOTOR VEHICLE COMPRISING SUCH A DEVICE.
    EP1851689B1|2009-04-01|Surface acoustic wave transponders
    WO2015078682A1|2015-06-04|Anticollision radar, especially for an aircraft when taxiing, and anticollision system
    WO2001090774A1|2001-11-29|Method and device for measuring the speed of a moving object
    EP2574957A1|2013-04-03|Method for estimating the unambiguous Doppler frequency of a moving target, in particular marine, and radar implementing said method
    EP3321711A1|2018-05-16|Receiving device for electronic scanning antenna capable of operating in radar and resm mode, and radar provided with such a device
    FR3064800A1|2018-10-05|METHOD FOR GUIDING AN AIR TARGET, IN PARTICULAR IN A VERTICAL LANDING PHASE, AND RADAR SYSTEM USING SUCH A METHOD
    FR3027405A1|2016-04-22|DETERMINING THE DIRECTION OF MOVEMENT OF A VEHICLE
    FR3109825A1|2021-11-05|Radar imaging method, and radar implementing such a method.
    EP1522871B1|2007-02-21|Radar with synthetic direction finding channels
    CA3007909A1|2017-06-22|Method for spectral estimation of the clutter of a haline liquid medium
    EP3819670A1|2021-05-12|Processing method for coherent mimo radar system processing ddma waveforms
    WO2021058674A1|2021-04-01|Radar imaging method, and radar implementing such a method
    FR3082950A1|2019-12-27|METHOD AND SYSTEM FOR DETERMINING AN ANGLE OF ARRIVAL OF A RADIOELECTRIC SIGNAL
    EP2810093B1|2018-07-18|Method of ambiguity resolution in distance measurements
    WO1997019365A1|1997-05-29|Radar device for a vehicle, particularly a motor vehicle
    同族专利:
    公开号 | 公开日
    EP3282283A1|2018-02-14|
    US10551488B2|2020-02-04|
    FR3055049B1|2018-07-27|
    EP3282283B1|2020-08-19|
    US20180045819A1|2018-02-15|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    EP0863409A1|1997-03-04|1998-09-09|Thomson-Csf|Method and device for continuous wave frequency modulation radar detection with range-velocity ambiguity resolution|
    US20100277361A1|2007-09-12|2010-11-04|Thomas Focke|Motor vehicle fmcw radar having linear frequency ramps of different slopes that are set apart, which are associated with different angular ranges|
    FR2987683A1|2012-03-02|2013-09-06|Thales Sa|RADAR WITH LOW PROBABILITY OF INTERCEPTION|
    US20160131742A1|2013-06-25|2016-05-12|Robert Bosch Gmbh|Angle-resolving fmcw radar sensor|
    US20160116570A1|2014-10-22|2016-04-28|Texas Instruments Incorporated|Method to "zoom into" specific objects of interest in a radar|
    DE19942665B4|1998-09-07|2014-02-13|Denso Corporation|FM CW radar apparatus for measuring the distance to a target and the relative velocity of the target|
    FR2954516B1|2009-12-18|2012-09-21|Thales Sa|RADAR DETECTION METHOD, IN PARTICULAR FOR AEROPORT RADARS USING A DETECTION AND OBSTACLE REMOVAL FUNCTION|DE102015226443A1|2015-12-22|2017-06-22|Robert Bosch Gmbh|Radar sensor, corresponding operating method and vehicle|
    DE102016202936A1|2016-02-25|2017-08-31|Robert Bosch Gmbh|Device for determining operating data for a radar sensor|
    DE102017200706A1|2017-01-18|2018-07-19|Robert Bosch Gmbh|Multiple subsampled chirp sequence radar|
    US10473755B1|2017-02-21|2019-11-12|National Technology & Engineering Solutions Of Sandia, Llc|Waveform warping for general frequency-modulated radar waveforms|
    IL259190D0|2018-05-07|2018-06-28|Arbe Robotics Ltd|System and method of fmcw time multiplexed mimo imaging radar using multi-band chirps|
    US11119185B2|2018-06-07|2021-09-14|GM Global Technology Operations LLC|Resolving doppler ambiguity in multi-input multi-output radar using digital multiple pulse repetition frequencies|
    US10890663B2|2018-06-28|2021-01-12|Zoox, Inc.|Loading multi-resolution maps for localization|
    KR20200002463A|2018-06-29|2020-01-08|삼성전자주식회사|Method and device to operate radar|
    US11125869B2|2018-10-16|2021-09-21|Infineon Technologies Ag|Estimating angle of human target using mmWave radar|
    US11061126B2|2019-03-26|2021-07-13|The Boeing Company|Cooperative frequency-modulated continuous-waveform radar systems|
    KR20210065259A|2019-11-26|2021-06-04|삼성전자주식회사|Radar apparatus and method of operating the same|
    法律状态:
    2017-07-28| PLFP| Fee payment|Year of fee payment: 2 |
    2018-02-16| PLSC| Search report ready|Effective date: 20180216 |
    2018-07-26| PLFP| Fee payment|Year of fee payment: 3 |
    2019-07-30| PLFP| Fee payment|Year of fee payment: 4 |
    2020-07-24| PLFP| Fee payment|Year of fee payment: 5 |
    2021-07-27| PLFP| Fee payment|Year of fee payment: 6 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1601224|2016-08-11|
    FR1601224A|FR3055049B1|2016-08-11|2016-08-11|MULTI-RESOLUTION FMCW RADAR DETECTION METHOD AND RADAR USING SUCH A METHOD|FR1601224A| FR3055049B1|2016-08-11|2016-08-11|MULTI-RESOLUTION FMCW RADAR DETECTION METHOD AND RADAR USING SUCH A METHOD|
    EP17181250.6A| EP3282283B1|2016-08-11|2017-07-13|Fmcw radar detection method with multiple resolution and radar using such a method|
    US15/649,356| US10551488B2|2016-08-11|2017-07-13|Multi-resolution FMCW radar detection method and radar implementing such a method|
    [返回顶部]