Method for focusing the radar detection for a relative movement

Abstract

The invention relates to a method for a radar system for detecting the surroundings of a motor vehicle, comprising - transmission means for the directed emitting of transmission signals, - reception means for the directed receiving of transmission signals reflected off objects and - signal processing means for processing the received signals, wherein - the frequency of the emitted transmission signals is modulated such that it contains a sequence of K linear ramps with at least approximately the same gradient and duration, hereinafter referred to as frequency ramps, - in the signal processing means, a signal with substantially the current transmission frequency or a constant offset thereto is mixed with the transmission signals reflected off objects and received by the reception means, - in the signal transmission means, the output signal of the mixture is, if necessary after suitable preprocessing, scanned I times during each of the K frequency ramps, - in the signal processing means, after preprocessing, a two-dimensional discrete time-frequency transformation over these I·K scanning values is completely or only partially determined, characterised in that, preferably dependent on the vehicle movement, the preprocessing of the I·K scanning values includes a frequency displacement of the signal formed from the I scanning values of the respective frequency ramp such that the frequency of the signals formed by the I scanning values remains unchanged over the K frequency ramps for objects with a defined radial relative movement, as a result of which in particular blurring of power peaks generated by such objects is counteracted in the two-dimensional time-frequency transformation.

Description

[0001] Method for focusing radar detection for relative motion

[0002] The invention relates to a radar method and radar system for use in driver assistance systems in motor vehicles. The radar system comprises a method according to the invention for focusing the radar detection for a relative movement.

[0003] State of the art

[0004] Motor vehicles are increasingly being equipped with driver assistance systems that use sensor systems to detect the surroundings or the current traffic situation and, based on the detected traffic situation, derive automatic vehicle reactions and / or instruct the driver, particularly by issuing warnings. A distinction is made between comfort and safety functions.

[0005] FSRA (Full Speed ​​Range Adaptive Cruise Control) plays an important role as a comfort feature in current development. The vehicle regulates its own speed to the desired speed set by the driver, provided the traffic situation permits; otherwise, the own speed is automatically adjusted to the traffic situation.

[0006] Safety functions now exist in a wide variety of forms. One group includes functions that reduce braking or stopping distances in emergency situations, all the way up to autonomous emergency braking. Another group is lane change functions: These warn the driver or intervene in the steering if the driver intends to make a dangerous lane change, i.e., if a vehicle in the adjacent lane is either in the blind spot (known as BSD - Blind Spot Detection) or rapidly approaching from behind (LCA - Lane Change Assist).

[0007] Nowadays, the driver is no longer just assisted, but the driver's task is increasingly carried out autonomously by the vehicle, i.e. the driver is increasingly replaced; this is referred to as autonomous driving.

[0008] Radar sensors are used for systems of the type described above, often in combination with sensors of other technologies, such as camera sensors. Radar sensors have the advantage, among other things, that they operate reliably even in poor weather conditions and can directly measure not only the distance of objects but also their radial relative velocity via the Doppler effect. The transmission frequencies typically used are 24 GHz, 77 GHz, and 79 GHz.

[0009] The above-mentioned functions require not only a long sensor range but also high measurement accuracy, resolution, and separation capability for distance and relative speed. High resolution and separation capability for distance and relative speed are also important because they can at least partially compensate for the poor angular resolution and separation capability of automotive radar sensors (resulting from their small size). However, there is the problem in today's radar systems that simultaneous high resolution of distance and relative speed is only fully possible with comparatively little relative movement between the object to be measured and the radar system. At higher relative speeds, "blurring" or "smearing" can occur in the radar detection. The term "blurring" or "smearing" is understood in particular to mean that, for example,Point-like objects in the radar image expand to several detection cells and thus appear like extended objects.

[0010] Task, solution and advantages of the invention

[0011] The object of the invention is to achieve high distance and relative speed resolution for relatively moving objects using a vehicle radar sensor. Stationary objects are of particular interest for sensors facing the direction of travel.

[0012] This problem is fundamentally solved with the aid of a radar method and radar system according to claims 1-16. The invention illustrates how radar detection of objects with a defined relative motion can be focused.

[0013] The advantages of the invention arise from improved radar detection.

[0014] In the method according to the invention for a radar system for detecting the surroundings of a motor vehicle, the radar system comprises transmitting means for emitting transmitted signals, receiving means for receiving transmitted signals reflected by objects, and signal processing means for processing the received signals, wherein the frequency of the emitted transmitted signals is modulated in such a way that it contains a sequence of K linear ramps with at least approximately the same gradient and duration, which is referred to below as frequency ramps. In the signal processing means, a mixture takes place between a signal with essentially the instantaneous transmitted frequency or a constant offset to this and the transmitted signals reflected by objects received by the receiving means. Furthermore, in the signal processing means, the output signal of the mixture is optionallyAmplification, bandpass filtering or the like) during each of the K frequency ramps is sampled I times and in the signal processing means, after pre-processing, a two-dimensional discrete time-frequency transformation is determined fully or only partially over these IK sample values, wherein preferably depending on the vehicle movement, the pre-processing of the I ■ K sample values ​​includes a frequency shift of the signal formed from the I sample values ​​of the respective frequency ramp in such a way that the frequency of the signals formed by the respective I sample values ​​over the K frequency ramps remains unchanged for objects with a defined radial relative movement, thereby counteracting a blurring or smearing of the power peaks generated by such objects in the two-dimensional time-frequency transformation.

[0015] The method can be used expediently for a radar system whose detection range includes the direction of travel, wherein the defined radial relative movement is the negative of the vehicle's own movement, so that for stationary objects in the direction of travel the frequency of the signals formed by the respective I sampling values ​​remains constant over the K frequency ramps.

[0016] Furthermore, the method can realize the frequency shift by multiplication with a rotating complex unit vector.

[0017] Preferably, the frequency shift is realized by multiplication by a rotating complex unit vector, wherein the respective I sample values ​​of the frequency ramps are equidistant in time and the rotational speed of the complex unit vector is constant during each frequency ramp, but changes over the frequency ramps. According to a preferred embodiment, the rotational speed of the complex unit vector can change over the frequency ramps proportionally to the integral of the speed of the defined radial relative movement, i.e., in particular, proportional to the integral of the vehicle's own speed.

[0018] Preferably, the center frequency and the time interval between the frequency ramps are at least approximately constant, and a linear change in the rotational speed of the complex unit vector over the frequency ramps is used, which corresponds to the possibly simplifying assumption of a constant speed of the defined radial relative movement during the acquisition of the I ■ K sample values, thus in particular corresponding to a constant vehicle speed.

[0019] Alternatively, the center frequency and the temporal spacing of the frequency ramps can also change at least approximately linearly, wherein the relative change in the temporal spacing is at least approximately twice as large as the relative change in the center frequency and the signs of these changes are opposite, and a linear change in the rotational speed of the complex unit vector over the frequency ramps is used, which corresponds to the possibly simplifying assumption of a constant speed of the defined radial relative movement during the acquisition of the IK sample values, i.e. in particular a constant vehicle speed.

[0020] Preferably, the phase of the complex unit vector is point-symmetric over both the I samples and the K frequency ramps, i.e. zero in the center in each case, so that there is no change in the position of the power peaks generated by objects in the two-dimensional discrete time-frequency transformation.

[0021] Conveniently, the first stage of the two-dimensional discrete time-frequency transformation can be performed over the respective I samples per frequency ramp, preferably with a fast Fourier transform for efficient implementation of the discrete Fourier transform, and the frequency shift can be implemented in combination with the window function used for the transformation. According to a particular embodiment of the invention, the window function can be changed iteratively from frequency ramp to frequency ramp by multiplying by the same constantly rotating complex unit vector.

[0022] Furthermore, frequency shifts corresponding to different radial relative movements can be calculated using at least partially identical sample values ​​of several two-dimensional discrete time-frequency transformations.

[0023] Advantageously, preferably depending on the vehicle movement, the preprocessing of the IK sample values ​​can include a phase shift of the signal formed from the I sample values ​​of the respective frequency ramp such that the phase of the signals formed by the respective I sample values ​​over the K frequency ramps for objects with a defined radial relative movement has a purely linear change, thereby preventing the power peaks generated by such objects from being smeared or blurred in the two-dimensional time-frequency transformation in the dimension generated by the K frequency ramps.

[0024] Preferably, the method can be applied to a radar system whose detection range includes the direction of travel, wherein the defined radial relative movement is the negative of the vehicle's own movement, so that for stationary objects in the direction of travel, the phase of the signals formed by the respective I sample values ​​over the K frequency ramps has a purely linear change.

[0025] Furthermore, the phase shift can be realized by multiplication with a rotating complex unit vector.

[0026] It is expedient to combine the phase shift with the window function used in the transformation for the dimension generated by the K frequency ramps.

[0027] In addition, the invention also encompasses a radar system for detecting the surroundings of a motor vehicle, which focuses the radar detection for a relative movement, in particular using a method according to the invention. The radar system comprises transmitting means for the directed emission of transmitted signals, receiving means for the directed reception of transmitted signals reflected by objects, and signal processing means for processing the received signals. The frequency of the emitted transmitted signals is modulated such that it includes a sequence of K linear ramps with at least approximately the same gradient and duration (frequency ramps). In the signal processing means, a mixture takes place between a signal with essentially the instantaneous transmitted frequency or a constant offset thereto and the transmitted signals reflected by objects received by the receiving means.In the signal processing means, the output signal of the mixture is sampled once during each of the K frequency ramps, if necessary after suitable preprocessing, and in the signal processing means, after preprocessing, a two-dimensional discrete time-frequency transformation is fully or partially determined using these IK sample values. Furthermore, the radar system is characterized in that, preferably depending on the vehicle movement, the preprocessing of the IK sample values ​​includes a frequency shift of the signal formed from the I sample values ​​of the respective frequency ramp such that the frequency of the signals formed by the I sample values ​​remains unchanged across the K frequency ramps for objects with a defined radial relative movement, thereby counteracting smearing / floating, i.e., a kind of widening, of the power peaks generated by such objects in the two-dimensional time-frequency transformation.

[0028] Brief description of the drawings

[0029] Fig. 1 shows an exemplary embodiment of a radar system.

[0030] Fig. 2 shows the frequency of the transmission signals, which represent so-called frequency ramps, with a constant frequency position.

[0031] Fig. 3 shows the magnitude spectrum after the two-dimensional discrete Fourier transformation for four objects without application of the method according to the invention, wherein the objects have no relative acceleration to the radar system.

[0032] Fig. 4 shows the magnitude spectrum after the two-dimensional discrete Fourier transformation for four objects using the method according to the invention, wherein the objects have no relative acceleration to the radar system.

[0033] Fig. 5 shows the magnitude spectrum after the two-dimensional discrete Fourier transformation for four objects which have a relative acceleration to the radar system, whereby only the first step of the method according to the invention is applied, i.e. only a frequency shift of the received signals.

[0034] Fig. 6 shows the magnitude spectrum after the two-dimensional discrete Fourier transformation for four objects which have a relative acceleration to the radar system, whereby both steps of the method according to the invention are applied, i.e. both frequency and phase shift of the received signals.

[0035] Fig. 7 shows the frequency of the transmitted signals with a linearly changing frequency position.

[0036] Example

[0037] Considered is an exemplary design of a radar system, which is roughly illustrated in Fig. 1. The radar system has a transmitting antenna TX0 for transmitting transmitted signals and M=4 receiving antennas RX0-RX3 for receiving transmitted signals reflected from objects. The antennas are implemented as patch antennas on a flat circuit board 1.1 using planar technology. This circuit board is oriented with respect to the horizontal and vertical directions in the vehicle as shown in the figure and faces the direction of travel. All antennas (transmitting and receiving antennas) have the same beam characteristics in elevation and azimuth. The four receiving antennas (and thus their phase, i.e., radiation centers) each have the same lateral, i.e., horizontal, distance d = X / 2 = 1.96 mm from each other, where X = c / 76.5 GHz = 3.92 mm is the mean wavelength of the transmitted signals in the used frequency band of 76-77 GHz and c=3*10 8 m / s is the speed of light.

[0038] The transmission signals emitted by the transmitting antenna are obtained from the high-frequency oscillator 1.2 in the 76-77 GHz range, whose frequency can be varied via a control voltage vcontrol. The control voltage is generated in the control means 1.7, which may contain, for example, a phase-locked loop or a digital-to-analog converter, which are controlled so that the frequency response of the oscillator corresponds to the desired frequency modulation.

[0039] The signals received by the four receiving antennas are also mixed down to the low-frequency range in the real-value mixers 1.3 with the signal from oscillator 1.2. The received signals then pass through the bandpass filters 1.4 with the transfer function shown, the amplifiers 1.5, and the analog / digital converters 1.6. They are then further processed in the digital signal processing unit 1.8.

[0040] In order to measure the distance of objects, the frequency hx of the high-frequency oscillator and thus of the transmitted signals is changed very quickly and linearly (in Tch=51.2ps by Bch=600MHz, with the center frequency fc=76.5GHz), as shown in Fig. 2. This is referred to as a frequency ramp (often also referred to as a "chirp"). The frequency ramps are repeated periodically at a fixed interval of To=70ps; in total, there are K=512 frequency ramps, all of which have the same frequency response, i.e., the same frequency gradient, the same frequency position (i.e., in particular, the same start and center frequency), and the same duration. In recent years, this type of modulation has become increasingly widespread and established in radar for detecting the surroundings of motor vehicles. It allows for a high sensor range and speed resolution (due to long data acquisition times) as well as a high distance resolution (through the use of high modulation bandwidth).

[0041] During each frequency ramp k=0, ... , K-1 the received signals from each of the M=4 A / D converters m=0,... ,M-1 are sampled 1=2048 times each at intervals of T s =25ns (i.e., at 40MHz), with sampling always beginning at the same time relative to the start of the ramp (see Fig. 2); the resulting digital samples with index i=0, ... ,1-1 are denoted by s(i,k,m). Signal sampling only makes sense in the time domain where received signals from objects arrive in the distance range of interest - after the start of the ramp, at least the propagation time corresponding to the maximum distance of interest must be waited for (for a maximum distance of interest of 200m, this corresponds to 1.33ps); it should be noted that here and in the following, distance always refers to the radial distance, and relative velocity to its radial component.

[0042] As is known from the state of the art and can also be easily derived, the scanning signal s(i,k,m) in the case of a single point-like object at a distance r(k) represents a sinusoidal oscillation over the index i, which can be described in very good approximation as follows: s(i,k,m) = A(m) / (r(k) / (meter)) A 2 • sin(2ir i / lr(k) / (meter) Bch / 150MHz + cp(k)+cpo(m)) , (1 ) ie the frequency of the oscillation is proportional to the object distance r(k), which changes slightly with a radial relative movement of the object to the sensor over the K=512 frequency ramps k=0,... ,K-1. A relative movement also affects the phase position <p(k) der sinusförmigen Schwingung aus; bei einer Bewegung mit konstanter radialer Geschwindigkeitskomponente v ergibt sich: cp(k) = 2TT-k-2T Dvfc / c , (2) ie, the phase position changes linearly over the frequency ramps k, where the rate of change of the phase is proportional to the radial relative velocity v of the object. Due to the linearity of the receivers, the sampling signal s(i,k,m) results in the case of multiple and / or extended objects as a linear superposition of sinusoidal functions of the above form.

[0043] If the change in object distance r(k) is neglected in equation (1), i.e., a constant frequency is assumed for all frequency ramps, then the optimal filtering of the signal form (1) per receive channel m corresponds to a two-dimensional discrete Fourier transform (DFT), which can be implemented very efficiently in two stages using two one-dimensional fast Fourier transforms (FFT). This, including suitable signal windowing (in both dimensions), has become established as the standard evaluation method for the modulation form considered here, shown in Fig. 2.

[0044] According to this two-dimensional DFT, power peaks S(j,l,m) appear in the resulting spectrum, the respective positions of which correspond to the mean distance r and relative velocity v of the corresponding object - see Fig. 3, which shows the magnitude spectrum |S(j,l,m) / A(m)| in dB, independent of the receiving channel m, for four point objects with the same radar cross-section, azimuth angle of approximately 0° and the following distances and relative velocities: [n=50m, v=-50m / s], [r2=100m, V2=0m / s], [r3=149.25m, V3=-50m / s] and [r4=150m, V4=-50m / s]; the signals of the objects are still superimposed by receiver noise, which lies in the spectrum significantly below the power peaks of the objects labeled with the object numbers. Negative relative velocities mean objects approaching relatively to the vehicle; the vehicle's own speed is assumed to be 50 m / s, so that objects 1, 3 and 4 are stationary.The moving object 2 has no relative velocity, traveling at 50 m / s both in absolute and absolute terms. The dimension j=0, ... , J-1 resulting from dimension i (sampling value indices) is called distance gates, and the dimension l=-L / 2, ... , L / 2-1 resulting from dimension k (frequency ramps) is called Doppler gates, since the position of the power peaks in dimension j essentially results from the object distance and in dimension l from the radial relative velocity (which is mapped via the Doppler effect) - it can be neglected here that the power peak position also exhibits a very small dependence on the other of the two physical quantities, distance and relative velocity.It should be noted that the velocity cannot be unambiguously calculated from the Doppler gate of the power peak, since in the present design only an unambiguous range of 28 m / s is realized across the K=L=512 Doppler gates. Ambiguities can be realized, for example, by varying the spacing TD of the frequency ramps from radar cycle to radar cycle (see also below). According to Fig. 3, the number of range gates is only J=801 and thus significantly smaller than the number I=2048 of samples. This is because, on the one hand, the samples are real-valued, so their spectrum is symmetrical, i.e., no additional information is contained in the upper half of their DFT, and, on the other hand, the upper transition range of the analog bandpass filter 1.4 according to Fig. 1 has a frequency bandwidth of 8.75 MHz (corresponding to the range of 448 frequency support points of the DFT).With the modulation bandwidth Bch=600MHz used here, the distance gate width 150MHz / Bch-1 m is equal to 25cm, so that the J=801 distance gates allow a maximum range of 200m.

[0045] As can be seen from Fig. 3, the two stationary objects 3 and 4 with [T3=149.25 m, V3=-50 m / s] and [r4=150 m, V4=-50 m / s] could not be separated, but merged into a single power peak, even though their separation is three times the range gate width, which would lead one to expect separability (typically, a difference of about two range gates is needed to separate two equally powerful point targets). The reason for this can be seen by comparing stationary object 1 with [n=50 m, v1=-50 m / s] and moving object 2 with [r2=100 m, V2=0 m / s]: the power peak for the moving object with zero relative velocity has the expected sharp shape, whereas it is significantly broadened for the stationary object with high relative velocity – this broadening of the power peaks leads to their merging in stationary objects 3 and 4.The broadening is explained by the fact that at a relative velocity of -50 m / s, the object moves by approximately 1.79 m during the entire data acquisition time of 512-70 ps = 35.84 ms, thus passing through more than seven range gates. The broadening occurs not only in the distance dimension but also in the Doppler dimension, because the object is not in a range gate the entire time, but only for a reduced time. This can be imagined as a Doppler window with a window of reduced width – the spectrum of such a window is broadened, and thus also the shape of the power peak, which corresponds to this spectrum.In addition to the reduced object separation capability, broadening the power peak has two further disadvantages: first, the possible detection range is reduced (because the level decreases due to energy dissipation), and second, the measurement of distance and relative velocity becomes less accurate (because noise superimposed on a blurred power peak causes greater errors). This example in Fig. 3 shows that simultaneous high resolution (and thus separation capability) in distance and relative velocity has its limits when the object being measured is moving relative to the object.

[0046] The two-dimensional DFT used neglects the fact that in relation (1 ) the frequency of the received signal changes slightly over the frequency ramps with the object distance r(k); thus, the higher the relative velocity of the object, the more it deviates from optimal filtering. Assuming a constant relative velocity v (which is a mostly permissible simplification during the short data acquisition time), the object distance changes from frequency ramp to frequency ramp by V-TD (To=70ps is the distance between the frequency ramps) and thus, according to relation (1 ), the time-discrete frequency (related to the sample index i, i.e., not related to the continuous time t) by

[0047] Af = 1 / l (wT D ) / (meter) Bch / 150MHz . (3)

[0048] This change Af of the time-discrete frequency of the received signal from frequency ramp to frequency ramp can be compensated by a corresponding inverse frequency shift, i.e. the frequency of the received signal of the kth frequency ramp k=0, ... , K-1 is shifted by the frequency -k-Af compared to the frequency of the received signal of the first frequency ramp k=0, which can be realized according to the frequency shift theorem of the Fourier transform by multiplication in the time domain (i.e. over the I samples with index i=0, ... ,1-1 of the respective frequency ramp k) with a rotating complex unit vector pi(i,k) = exp(-j.-2TT-ik-Af) (4) (where "exp" denotes the exponential function and the symbol j the imaginary unit - not to be confused with the running variable i for the samples); the rotational speed of the complex unit vector, which is constant within a frequency ramp, increases linearly over the frequency ramps.After multiplying the sample values ​​s(i,k,m) with the complex unit vector pi(i,k) according to ref. (4) using the frequency change Af according to ref. (3), objects with the radial relative velocity v have the same reception frequency in all frequency ramps, so that their power peak is sharp, ie focused, in the two-dimensional spectrum.

[0049] The ramp index k is defined asymmetrically (it starts at zero, so it is not zero for the "medium" frequency ramp). Thus, when the frequency is shifted with the rotating complex unit vector pi(i,k) according to equation (4), the received frequency of each frequency ramp is shifted to that of the first frequency ramp, so that the result of the distance measurement would be the distance at the first frequency ramp, i.e., at the beginning of data acquisition. Typically, however, one would like to determine the distance in the middle of data acquisition, which can be achieved by replacing k with a (point-)symmetric k-(K-1) / 2 (which is zero for the "medium" frequency ramp (K-1) / 2).The index i of the sample values ​​is also defined asymmetrically; thus, the complex unit vector pi(i,k) according to equation (4) has a linearly changing phase at the "average" index (1-1 ) / 2 over the frequency ramps k, which would lead to a slight Doppler shift, i.e. a slightly distorted measurement of the relative velocity; to avoid this, i is replaced by a (point-)symmetric i-(l-1 ) / 2. This results in the rotating complex unit vector pi(i,k) for the frequency shift being pi(i,k) = exp(-i-2TT (i-(l-1 ) / 2) (k-(K-1 ) / 2) Af) (5) with the frequency change Af according to equation (3).

[0050] For this Af, a relative velocity v must be defined, i.e., the broadening of the power peaks in the two-dimensional DFT can only be completely avoided for a relative velocity v of objects; for objects with a different relative velocity, the broadening will be more pronounced the further their relative velocity is from the relative velocity defined for the design of Af. For the sensor considered here, which faces the direction of travel, the precise detection of stationary objects is essential, particularly with regard to autonomous braking functions. Braking must not be erroneous, which could happen, for example, if two stationary objects (stationary vehicles, guardrail posts, buildings, ...) to the right and left of the vehicle's lane merge and are therefore assumed to be in the vehicle's lane.In this context, it is important to understand that in radar systems, pure angular separation is very poor; in the radar system shown here, with only one transmitting and four receiving antennas, it is only possible - if at all - if the two objects have a very large angular difference. Even radar systems available today with significantly more transmitting and receiving antennas can usually only separate two objects if they have a significant angular difference and a very similar backscatter cross-section. Therefore, one relies on very good separation via range and / or Doppler (Doppler separation also helps with stationary targets because the radial relative velocity depends on the cosine of the azimuth angle); if there is only one object in a range-Doppler cell, the angle determination is generally accurate enough.In addition to the false braking described above, the overlooking of real stationary obstacles is of course also critical, e.g. a stationary vehicle next to a guardrail or under a bridge; here too, good separation ability over distance and / or Doppler is important in order to be able to determine the azimuth angle and thus the position relative to the vehicle's own lane with sufficient accuracy. Therefore, the effect described above of smearing / blurring of the power peaks of stationary objects is very disadvantageous in both the distance and Doppler dimensions. Therefore, for the sensor considered here, which looks in the direction of travel, it is advantageous to choose the relative velocity v for the design of Af and thus the rotational velocities of the rotating complex unit vector pi(i,k) as that of stationary targets at an azimuth angle of 0°, i.e. as the negative of the vehicle's own velocity v. ego ; from ref. (3) we get

[0051] Af = 1 / l (-veg o T D ) / (meter) Bch / 150MHz . (6)

[0052] If the frequency shift defined in this way is applied to the above example with the 4 objects, i.e. if the sample values ​​s(i,k,m) are multiplied by the complex unit vector pi(i,k) according to ref. (5) using the frequency change Af according to ref. (6), the magnitude spectrum |S(j, I, m) / A(m)| in dB according to Fig. 4 is obtained after the subsequent two-dimensional DFT. All three stationary objects 1, 3 and 4 now show sharp power peaks, so that the two objects 3 and 4 can also be clearly separated with a distance difference of only 0.75m. The level of the power peaks of these objects has also increased by almost 6dB (because the power is now focused in a sharp peak).

[0053] However, the power peak of the moving object 2 is now broadened and the level has decreased by almost 6 dB (leading to a reduced detection range). In the original spectrum shown in Fig. 3, no frequency shift was performed, meaning that the focus was effectively on the relative velocity zero - and object 2 has a relative velocity zero; with the focus on relative velocity -Vego in the spectrum shown in Fig. 4, the relative velocity of object 2 is now v egoaway from it, which leads to the same broadening for the moving object as for the stationary objects in the original spectrum shown in Fig. 3. However, from a functional point of view, this is less critical for the forward-facing sensor because, on the one hand, optimal separation in distance and Doppler is less important for moving objects (several moving objects with very similar distance and relative speed rarely occur), and on the other hand, a large detection range is not required for an object moving at the same speed (which is the case for object 2). However, if one still wants to have a focused power peak for moving object 2 as well, one would have to calculate a second two-dimensional DFT - in this case, the multiplication for the frequency shift would be omitted (since there is zero frequency shift due to zero relative speed).In the general case, if one wants to focus on several relative velocities, one has to calculate both the multiplication for the frequency shift and the two-dimensional DFT several times; for an object, one then uses the spectrum where the relative velocity of the object is closest to the relative velocity on which one focused.

[0054] So far, we have considered the case where the relative velocity v used to focus the power peaks is constant during the data acquisition time, or can be assumed to be. Especially at high relative accelerations and longer data acquisition times, this assumption would lead to blurred power peaks. The change Af of the time-discrete frequency of the received signal from frequency ramp to frequency ramp is then no longer constant (as in ref. (3)), but changes over the frequency ramps k=0, ... ,K-1 with the changing relative velocity v(k):

[0055] Af(k) = 1 / l (v(k) T D ) / (Meter) Bch / 150MHz . (7)

[0056] For the rotating complex unit vector pi(i,k) for the frequency shift, the sum over Af(k) is required; for this, the sum over the distance change v(k) o included in the above formula must be calculated. The sum over the time-discrete values ​​v(k) o (TD is the distance between the time-discrete values) corresponds to the integral int[v(t)](k) over the time-continuous v(t) at the times t(k) corresponding to the frequency ramps k when transitioning to continuous time; the center of the data acquisition time can be used as a reference point for the integration, for example.This results in the rotating complex unit vector pi(i,k) being: pi(i,k) = exp(-i-2TT (i-(l-1 ) / 2) / l int[v(t)](k) / (Meter) Bch / 150MHz) ; (8) it should be noted that in this relationship it is further assumed that the relative velocity is constant during a ramp, which is expressed by a constant rotational velocity of pi(i,k) during each frequency ramp (this assumption is also valid at high relative acceleration due to the very short time of the frequency ramps).

[0057] After multiplying the sample values ​​s(i,k,m) by the unit vector pi(i,k), all frequency ramps have the same reception frequency, so there is no broadening of the power peaks in the spectrum in the distance dimension. However, it must also be considered that the relative velocity changes with relative acceleration, which means that several Doppler gates can be crossed over the data acquisition time, thus still leading to a broadening of the power peaks in the Doppler dimension. Fig. 5 shows this effect for the four objects above with an additional relative acceleration of 10 m / s. 2(by emergency braking of the own vehicle). A changing relative speed means that the phase <p(k) der Empfangssignale über die Frequenzrampen hinweg nicht mehr linear ändert, so dass Bez. (2) nicht mehr gültig ist; die Änderung ist nun proportional zur nicht konstanten Relativgeschwindigkeit, und damit muss man den Term k-To-v in Bez. (2) durch das Integral int[v(t)](k) ersetzen (analog zur obigen Herleitung für pi(i,k) nach Bez. (8)): cp(k) = 2TT-2-int[v(t)](k) fc / c . (9) Der nichtlineare Anteil in der Phase <p(k) ergibt sich aus Differenz zwischen dieser Bez. (8) und der ursprünglichen Bez. (2), und man kann ihn durch Multiplikation der Abtastwerten s(i,k,m) mit einem entsprechenden zweiten komplexen Einheitsvektor P2(k) kompensieren, welcher sich von Frequenzrampe zu Frequenzrampe ändert, pro Frequenzrampe aber konstant ist: p2(k) = exp(-i-2TT-2 (int[v(t)](k)-k T DVav) fc / c) , (10) where Vav is the average velocity during the data acquisition time. Fig. 6 shows the magnitude spectrum |S(j, l,m) / A(m)| when applying this second complex unit vector p2(k) for the relative acceleration 10 m / s 2 ; the stationary objects now show sharp power peaks again (note that the two objects 3 and 4 are separated in distance dimension in both Fig. 5 and Fig. 6, which is not obvious due to the viewing angle).

[0058] After multiplying the sample values ​​s(i,k,m) by both complex unit vectors pi(i,k) and p2(k), a sharp power peak is obtained in the spectrum in both dimensions (i.e., distance and Doppler), provided the object exhibits the relative motion assumed for the design of pi(i,k) and p2(k). As explained above, for the forward-looking sensor considered here, stationary objects around azimuth 0° are most important, so the relative motion to be considered is the inverse, i.e., negative, of the generally non-constant proper motion Vego(t); thus, the two complex unit vectors pi(i,k) and p2(k) are: pi(i,k) = exp(i-2iT-(i-(l-1 ) / 2) / l-int[v e go(t)](k) / (Meter)-Bch / 150MHz) , (11 a) p2(k) = exp(i-2TT-2 (int[Vego(t)](k)-k TD Vego,av) fc / c) (11 b)

[0059] It should be noted that in Relation (11 b) one could in principle also omit the subtraction of k-To-Vego.av; this would imply a Doppler shift such that stationary objects at azimuth 0° would be at Doppler 0.

[0060] As an alternative to the modulation form shown in Fig. 1, one can use the one shown in Fig. 7, in which the individual frequency ramps have a bandwidth Bch=150MHz reduced to 25%, but their frequency position (e.g. characterized by their center frequency) is linearly distributed over the bandwidth B s =600MHz. This linear change in the center frequency leads to a distance-dependent component in the Doppler dimension of the two-dimensional DFT, over which a modulation bandwidth B s=600MHz corresponding high range resolution is achieved. As shown in DE 10 2020 210 079 B3, for relatively moving objects, a strong broadening of the power peak in the Doppler dimension can be counteracted by a linearly changing temporal spacing To(k) of the frequency ramps k=0,...,K-1; relatively speaking, this change is twice as strong in magnitude as the change in the center frequency and has the opposite sign, so for the example in Fig. 2 it is -2,600MHz / 76.5GHz = -1.57% (decrease by 1.57% over the K frequency ramps). However, even with this form of modulation, there is the effect that the received frequency changes slightly over the frequency ramps due to relative movement; However, since the individual frequency ramps have a significantly smaller bandwidth Bch=150MHz than for the first modulation form, the effect of blurring / smearing of the power peaks in distance and Doppler dimensions is significantly smaller (by a factor of 4).This smaller effect can also be completely eliminated, as above, by frequency shifting, i.e., by multiplying the sampled values ​​by the complex unit vector pi(i,k) according to the formulas above. By additional multiplication by the second vector p2(k) according to the formulas above, the effect of a non-constant relative velocity, i.e., a relative acceleration, can be compensated. The above relationships were derived, at least in part, under the assumption of temporally equidistant frequency ramps, which is not exactly the case here—however, the deviation from this assumption is negligible.

[0061] The two-dimensional DFT is usually determined first with an FFT for the distance dimension and then with an FFT for the Doppler dimension. The reason for this is that the data from the frequency ramps are acquired one after the other, and as soon as data from a new frequency ramp is available, this first FFT can be determined using the samples from the respective frequency ramp. Furthermore, after this first FFT for the distance dimension, the resulting data can be highly compressed without significant loss of information (see EP 3 152 587 B1). The FFT for the Doppler dimension can only be determined once the data for all frequency ramps are available (i.e., after the entire data acquisition); if one were to start with the FFT for the Doppler dimension, the determination of the two-dimensional DFT could only begin after the entire data acquisition.If the FFT is determined first for the distance dimension, i.e. over the dimension i of the sample values ​​s(i,k,m), then one only needs to multiply by the portion of the complex unit vectors that depends on dimension i, i.e. by pi(i,k) - P2(k) does not depend on dimension i and can therefore also be multiplied after the first FFT. It is advantageous to first form the product between the window function of the first FTT (for the distance dimension) and the vector pi(i,k) once and to apply the modified window function to the received signals of all M=4 receiving antennas. In the case of a relative velocity assumed to be constant, i.e. for a pi(i,k) according to equation (5), it can be particularly useful toWhen using hardware-accelerated processing units, it can be advantageous to generate this modified window function iteratively over the frequency ramps k by multiplying it with the vector exp(-j-2iT-(i-(l-1 ) / 2)-Af) that is independent of k (this is also a rotating complex unit vector). It should be noted that multiplying the real-valued samples s(i,k,m)) with the complex unit vector pi(i,k) results in a complex-valued signal, so that the potential advantages of a real-valued input signal can no longer be utilized in the FFT.As mentioned above, the multiplication with the complex unit vector p2(k), which depends only on the frequency ramp variable k, can only be performed after the first FFT, which is conveniently realized by a single multiplication of vector p2(k) with the window function of the second FFT (for Doppler dimension); this modified window function can then be used for the signals to all M=4 receiving antennas and all J=801 range gates.

[0062] It should be emphasized that the method according to the invention is not limited to the above sequence for implementing the two-dimensional DFT. One can also start with the FFT for the Doppler dimension, in which case the multiplication by both complex unit vectors pi(i,k) and p2(k) must first be implemented, since both depend on the frequency ramp variable k.

[0063] So far, a forward-facing sensor has been considered. Of course, the method according to the invention can also be used for sensors with a different orientation. However, the relative motion underlying the design of the complex unit vectors pi(i,k) and p2(k) may then be defined differently; for example, for a rearward-facing sensor, it is not stationary objects that are of greatest importance, but rapidly approaching vehicles. To ensure that the radar system is robust with respect to interference from other radar systems, modulation parameters are preferably varied, particularly analogously to the approaches described in WO 2008 / 040341 ​​A1, DE 102009 016 480 A1, and EP 2 629 113 B1, e.g.:

[0064] - average spacing of frequency ramps from cycle to cycle (as mentioned above, also allows for easy resolution of speed ambiguities);

[0065] - Modulation bandwidth Bch (magnitude and / or sign) from cycle to cycle;

[0066] - temporal spacing To(k) of the frequency ramps by superimposing a random or pseudorandom mean-free component varying over k, typically in the range up to a few microseconds; for relatively moving objects, the reception phase then exhibits a component that varies slightly over the frequency ramps, but which is still so small that the resulting effects are negligible according to the DFT (noise and level reduction of the power peak);

[0067] - Frequency position of the frequency ramps (i.e., their center frequency) by superimposing a random or pseudorandom mean-free component that varies over k; this variation of the frequency position can also be achieved by always using the same frequency ramps, but varying the time from which the samples of the received signal are obtained; the resulting phase variation of the received signals, which is proportional to the distance gate, can be compensated by appropriate general phase correction according to the first one-dimensional DFT for distance dimension;

[0068] - Phase position of the individual transmission signals by an additional phase modulator in the transmission means, whereby the phase position is varied randomly or pseudo-randomly via the frequency ramps, which is to be compensated again on the receiving side, preferably in the digital signal processing means.

[0069] In the radar system considered in Fig. 1, there are M=4 receiving antennas and associated receiving channels m=0,...,M-1. Following the two-dimensional DFT, a digital beamformer, e.g., again in the form of a DFT or FFT, is preferably calculated in each range Doppler gate (j,l); thus, a three-dimensional Fourier transformation is performed. Power peaks are then determined in the three-dimensional spectrum. The azimuth angle of an object results from the position of its power peak in the third dimension, which is derived from the dimension m of the receiving channels; distance and relative velocity are derived from the other two dimensions according to the relationships above. In order to have more channels available for angle formation, it is preferable to use not only several receiving antennas, but also several transmitting antennas, and to evaluate the signals from all combinations of transmitting and receiving antennas to realize many virtual receiving channels.If all or some of the transmitting and / or receiving antennas are not operated simultaneously, then several preferably similar sequences of frequency ramps of the types described above are nested.

[0070] Concluding remark

[0071] It should be noted that the inventive considerations and designs presented in the above application example can, of course, be easily transferred to general measurements and parameter interpretations, i.e., they can also be applied to other numerical values. Therefore, general parameters are also provided in the formulas and figures.

Claims

Claims 1 . Method for a radar system for detecting the surroundings of a motor vehicle with - transmitting means for transmitting signals, - Receiving means for receiving transmission signals reflected from objects and - signal processing means for processing the received signals, wherein - the frequency of the transmitted signals is modulated in such a way that it contains a sequence of K linear ramps with at least approximately the same gradient and duration, which are referred to below as frequency ramps, - in the signal processing means, a mixture takes place between a signal with essentially the instantaneous transmission frequency or a constant offset to this and the transmission signals received by the receiving means and reflected by objects, - in the signal processing means, the output signal of the mixture is sampled l times during each of the K frequency ramps, if necessary after suitable preprocessing, - in the signal processing means, after pre-processing, a two-dimensional discrete time-frequency transformation is determined fully or only partially over these I ■ K sample values, characterized in that preferably depending on the vehicle movement, the pre-processing of the IK sample values ​​includes a frequency shift of the signal formed from the I sample values ​​of the respective frequency ramp such that the frequency of the signals formed by the respective I sample values ​​over the K frequency ramps remains unchanged for objects with a defined radial relative movement, which in particular counteracts a blurring / smearing of the power peaks generated by such objects in the two-dimensional time-frequency transformation.

2. Method according to claim 1, characterized in that a detection range of the radar system includes the direction of travel and the defined radial Relative movement is the negative of the vehicle's own movement, so that for stationary objects in the direction of travel the frequency of the signals formed by the respective I sample values ​​remains constant over the K frequency ramps. Method according to one of the above claims, in which the frequency shift is realized by multiplication with a rotating complex unit vector. Method according to claim 3, in which the respective I sample values ​​of the frequency ramps are equidistant in time and the rotational speed of the complex unit vector is constant during each frequency ramp, but changes over the frequency ramps. Method according to claim 4, in which the rotational speed of the complex unit vector changes over the frequency ramps proportional to the integral of the speed of the defined radial relative movement, i.e. in particular changes proportionally to the integral of the vehicle's own speed.Method according to claim 5, in which the center frequency and the time interval of the frequency ramps are at least approximately constant and a linear change in the rotational speed of the complex unit vector over the frequency ramps is used, which corresponds to the possibly simplifying assumption of a constant speed of the defined radial relative movement during the acquisition of the IK sample values, that is to say in particular a constant vehicle speed.Method according to claim 5, in which the center frequency and the time interval of the frequency ramps change at least approximately linearly, the relative change in the time interval being at least approximately twice as large as the relative change in the center frequency and the signs of these changes being opposite, and a linear change in the rotational speed of the complex unit vector over the frequency ramps is used, which leads to the possibly simplifying assumption of a constant speed during the acquisition of the IK sample values. speed of the defined radial relative movement, i.e. in particular constant vehicle speed. Method according to claim 6 or 7, in which the phase of the complex unit vector is point-symmetrical over both the I sample values ​​and the K frequency ramps, i.e. in each case equal to zero in the middle, so that there is no change in the position of the power peaks generated by objects in the two-dimensional discrete time-frequency transformation. Method according to one of the above claims, in which the first stage of the two-dimensional discrete time-frequency transformation is carried out over the respective I sample values ​​per frequency ramp, preferably with a fast Fourier transformation for the efficient realization of a discrete Fourier transformation, and the frequency shift is realized in combination with the window function used for the transformation.Method according to claims 6 or 7 and 9, wherein the window function is changed iteratively from frequency ramp to frequency ramp by multiplication by the same constantly rotating complex unit vector. Method according to one of the above claims, wherein several two-dimensional discrete time-frequency transformations with frequency shifts corresponding to different radial relative movements are calculated over at least partially identical sample values.Method according to one of the above claims, wherein preferably depending on the vehicle movement, the preprocessing of the IK sample values ​​includes a phase shift of the signal formed from the I sample values ​​of the respective frequency ramp such that the phase of the signals formed by the respective I sample values ​​over the K frequency ramps for objects with a defined radial relative movement has a purely linear change, thereby preventing the power peaks generated by such objects from being smeared / blurred in the two-dimensional time-frequency transformation in the dimension generated by the K frequency ramps. Method according to claim 12, characterized in that the detection range of the radar system includes the direction of travel and the defined radial relative movement is the negative of the vehicle's own movement, so that for stationary objects in the direction of travel, the phase of the signals formed by the respective I sample values ​​exhibits a purely linear change over the K frequency ramps. Method according to claim 12 or 13, in which the phase shift is realized by multiplication by a rotating complex unit vector. Method according to claim 12, 13, or 14, characterized by a combined realization of the phase shift with the window function, which is used in the transformation for the dimension generated by the K frequency ramps. Radar system for detecting the surroundings of a motor vehicle with - transmitting means for the directed emission of transmission signals, - Receiving means for the directed reception of transmission signals reflected from objects and - signal processing means for processing the received signals, wherein - the frequency of the transmitted signals is modulated in such a way that it contains a sequence of K linear ramps with at least approximately the same gradient and duration, which are referred to below as frequency ramps, - in the signal processing means, a mixture takes place between a signal with essentially the instantaneous transmission frequency or a constant offset to this and the transmission signals received by the receiving means and reflected by objects, - in the signal processing means, the output signal of the mixture is sampled l times during each of the K frequency ramps, if necessary after suitable preprocessing, - in the signal processing means, after preprocessing, a two-dimensional discrete time-frequency transformation is determined fully or only partially over these I ■ K samples, characterized in that, preferably depending on the vehicle movement, the preprocessing of the IK sample values ​​includes a frequency shift of the signal formed from the I sample values ​​of the respective frequency ramp such that the frequency of the signals formed by the respective I sample values ​​remains unchanged over the K frequency ramps for objects with a defined radial relative movement, thereby counteracting in particular a blurring / smearing of the power peaks generated by such objects in the two-dimensional time-frequency transformation.

Images