Method for determining the true velocity of a target object detected by radar measurement of an FMCW radar within a single measurement cycle using range migration

Abstract

A method is proposed for determining the true velocity of a target object detected by radar measurement of an FMCW radar within a single measurement cycle, comprising the steps of: generating at least one reference signal for at least one specified reference velocity, generating an error measure that quantifies the effect of the deviation of the range migration on the frequency response of the Doppler window, and using the error measure to determine into which of the ambiguous intervals existing due to the sampling rate the true velocity of the target object falls.

Description

The present invention relates to the field of radar speed detection in the automotive field. In this case, the problem of velocity ambiguity (velocity ambiguity) within a measurement cycle is solved.Modern vehicles (cars, transporters, trucks, motor cycles etc.) have a multiplicity of sensors which make information available to the driver and control individual functions of the vehicle in a partially or fully automated manner. An important sensor principle for detecting the environment in the area of the autonomous and semi-autonomous driving vehicles is radar technology. Radar sensors in the vehicle sector typically comprise a plurality of transmitting and receiving elements which form virtual channels of the radar sensor (Rx / Tx antenna pairs). Down-conversion into the baseband signal, filtering and then digitizing the baseband signal thus obtained take place in each reception channel. Preprocessing the baseband signals for each reception channel can be used to detect and locate a target object, also referred to as target, in a field of view of the radar sensor.Most radar sensors currently used in the vehicle sector operate as multi-pulse radar sensors (also referred to as chirp sequence radar sensors), in which a plurality of frequency-modulated pulses are emitted at short intervals. The preprocessing can in this case comprise, in particular, a Fourier transformation or optimum filtering of the baseband data for a single radar pulse (fast-time processing), a further Fourier transformation for a plurality of pulses (so-called slow-time processing), beamforming, signal power detection, for example based on constant-force-alarm-rate methods, and a peak determination.One challenge for multi-pulse radar sensors with a constant pulse repetition time (PRT) is that they are limited to measurements of a relative speed (or a relative radial speed within the so-called main interval. Only within this main interval between -vm / 2 and + vm / 2 can a maximum unique speed be measured. This means that velocities at distance vmare mapped to the same value within the main interval. Current radar sensors which are used in the automotive sector typically have values of vm in the range from 20 to 60 m / s, so that ambiguous speed measurements can occur. This means that measured speeds can deviate by multiples of vmfrom the actual speed.One approach to solving these problems is to use different pulse repetition times, either between pulses or from a coherent processing interval (CPI) to follow. This allows the maximum speed to be increased. A disadvantage of using different pulse repetition times between successive intervals is that a latency time is increased and an association between detections at multiple intervals becomes necessary. This is prone to errors, in particular in the case of multiple and / or rapidly moving targets, as occurs regularly in the vehicle sector. Approaches in which the pulse repetition time is varied between pulses within the same interval are not suitable for fast Fourier transform (FFT) processing and may have side lobes in the velocity spectrum.A disadvantage of previously known methods is the comparatively complicated processing and the associated need for computing power. In addition, the previous approaches usually result in a loss of resolution.In WO 2020 / 201 337 A1, therefore, a device for determining a relative speed of a target object by means of model adaptation is proposed. However, since there is still a need for improvement in this solution, the object of the invention is to provide an improved apparatus and a method for determining a relative speed of a target object.This object is achieved by the features of the independent claims. Advantageous embodiments are the subject of the dependent claims.A method is proposed for determining a true speed of a target object detected by means of radar measurement of an FMCW radar within a single measurement cycle, comprising the steps of: generating at least one reference signal for at least one specified reference speed, and generating an error measure which quantitates the effect of the deviation of the range migration on the frequency response of the Doppler window, and using the error measure to determine in which of the ambiguous intervals present on the basis of the sampling rate the true speed of the target object falls.In one embodiment, in the case that a single reference speed is provided, the error measure is used in such a way that a signal adjacent to the peak value is generated at each boundary of a speed interval which is a multiple of the maximum detectable speed of the radar measurement and the associated error measure is calculated, wherein the speed interval at which the error measure is the smallest is determined as the interval in which the true speed of the target object falls.In one embodiment, in the case that a plurality of reference speeds are provided, the error measure is used in such a way that an associated error measure of the measured speed of the target object relative to the reference speeds is calculated at each limit of a speed interval of each of the reference speeds, and wherein a weighting of each error measure per reference speed is carried out, and wherein the speed interval at which the error measure is the smallest is determined as the interval in which the true speed of the target object falls.In one embodiment, the error measure serves as a measure of the true speed of the target object.In one embodiment, the sign of the speed of the detected target object is determined on the basis of the true absolute speed by dividing the fast Fourier transformation function into two sub-functions, wherein the sub-functions each correspond to a half range gate.In one embodiment, the fast Fourier transform function is divided into two sub-functions, and when one range gate is processed one after another, in the case that a target object is detected, the Doppler bin is stored at a first range gate and the sign of the result of the difference of the two sub-functions is checked in a subsequent range gate, wherein in the case of a positive value, it is detected that the target object is moving toward the radar sensor and in the case of a negative value, it is detected that the target object is moving away from the radar sensor.In one embodiment, double peak values of the measured radar signal are detected by evaluating an energy ratio of the two peak values in the case that two closely adjacent peak values are detected along a Doppler dimension after the complete FFT split into four sub-functions and the energy values of the two peak values are close to one another, and detecting double peak values in the case that both energy ratios are greater than a predefined threshold value.In one embodiment, the specified reference speed is equal to zero.In one embodiment, scanning signals of signals from radar measurements are aligned with scanning signals of the radar reference signal.Further features and advantages of the invention are evident from the following description of exemplary embodiments of the invention, on the basis of the figures of the drawing which shows details according to the invention, and from the claims. The individual features can each be realized individually or severally in any combination in a variant of the invention.Preferred embodiments of the invention are explained in more detail below with reference to the attached drawings. FIG. 1 shows the signal spectrum after a Doppler FFT of a stationary and a moving target object. FIG. 2 shows an error curve over increasing speeds. FIG. 3 illustrates determining the interval in which the true velocity lies, according to an embodiment of the present invention. FIG. 4 shows the signal spectrum after a Doppler FFT of a stationary and a very fast moving target object. FIG. 5 illustrates determining the sign of the relative velocity, according to an embodiment of the present invention.In the following descriptions of the figures, identical elements or functions are provided with identical reference numerals.For driver assistance systems, including Advanced Driver Assistance Systems (ADAS) and Automated Driving (AD) systems, a radar sensor may provide distance measurements and speed measurements of objects in its vicinity. Among many modulation schemes, the FMCW (Frequency Modulation Continuous Wave) is the most widely used waveform in the automotive field. Under FMCW modulation, a sequence / sequence is emitted at so-called chirps. The signal reflected after encountering an object (referred to as target object below) is detected by the radar sensor and converted into a baseband signal. Then, an analog-to-digital converter will sample each chirp to obtain the analog-to-digital converter samples. This yields a two-dimensional data block x[n t, m t]. One dimension runs along the samples in the same chirp and is referred to as the fast-time dimension. The other dimension, referred to as the slow-time dimension, runs across the chirps but along the same location of the analog-to-digital converter samples.Often, a two-dimensional FFT is applied to the two-dimensional data block x[n t, m t] so that the frequency response S[n f, m f] can be generated. When a reflection center is present in the view area, a high response, i.e., a high-energy value at a frequency location, is achieved in the two-dimensional spectrum, so that the distance (range) and the speed of the target object from the two-dimensional location of the high response can be obtained.In order to achieve a specific clear measurement of the maximum frequency, the sampling frequency (sample rate) should be at least twice greater than the highest frequency component to be examined in the measured signal according to the Nyquist sampling principle. However, this is difficult to maintain in terms of speed dimension for many automotive applications. For this reason, there will often be an alias effect, which means that a signal from a high speed will fold back into a unique interval. By the simple FFT-based method for peak detection, the true velocity cannot be determined. It will only output the folded velocity value in the unique interval.A universal approach to solving the velocity ambiguity problem is to set two or more different pulse repetition frequencies in successive radar cycles (measurement cycles), wherein the pulse repetition time / pulse repetition frequency remains constant within each cycle. This results in different, unique speed limits in these measurement cycles. Thereafter, the so-called Chinese Reducer Theorem is applied to determine the true velocity from the ambiguous estimates of velocity from the different configurations of the pulse repetition frequencies.In case the target object has a radial velocity component, the distance will experience a shift from its original location at each chirp as soon as a new radar cycle (also referred to as measurement cycle below) starts. This effect is called range migration. At low speed, e.g. |v|<20 m / s, or if the total chirp time in a measurement cycle is very short, the influence of the range migration is generally negligible. However, if the total chirp time in a measurement cycle is long enough to provide sufficient speed resolution and the target object has a speed that falls within ambiguous intervals, the effect of range migration will provide artifacts that are sufficiently strong to be measured.The signal processing normally used, based on simple FFT (fast Fourier transformation), in radars used in the automotive sector ignores the effect of range migration. In the present invention, however, this is used to determine the true speed (=actual relative speed of the target object to the radar) of a target object detected by the radar used. That is, the true velocity is determined from ambiguous velocity measurement within a measurement cycle by employing the effect of range migration.The steps described below are carried out in this case. First, a signal model is created. Here, a reference signal for a specified reference speed of the target object is generated. In particular, the generating is done by modifying the Doppler window (weighting the signals in the Doppler dimension) as described in detail below.In a further step, an error measure is generated in order to quantify the effect of the deviation of the range migration on the frequency response of the Doppler window.In a further step, this error measure is used to determine into which of the ambiguous intervals present on the basis of the sampling rate the true speed of the target object falls.As already described, a signal model is created in a first step. To simplify the description, only one Doppler dimension is considered. However, the method is also applicable to radar systems with multiple antenna channels, such as MIMO radars, by extracting the distance Doppler map at an angular position.The signal model of a point target object which is subject to the effect of range migration (for simplification, with the signal amplitude being ignored), by weight and before FFT is: the superimposition frequency f R, which corresponds exclusively to the distance (short=range), being calculated as:The shift of the superposition frequency f RM[ mt] at the mt-th chirp caused due to the range migration is estimated via:Then, a DFT (Discrete Fourier Transform) is applied along a so-called fast-time dimension nt of the time signal x[n t, m t] resulting in:The term W R[ ·] also has a dependence on the chirp index mt under the influence of the range migration. By converting the W R and w D to the term w DR one obtains:It can be seen that the effect of range migration can be considered a modification of the original Doppler window. More specifically, the main lobe of the frequency response of the Doppler window is flattened under the effect of range migration, as shown in Figure 1 (dotted curve).As already described, an error measure is generated in a second step. This error measure is an index of the artifacts of the range migration.For example, a = [a 0 a 1 a 2 a 3 a 4]T is set as the signal values in the neighborhood of the peak value of the original frequency response of w D[ m t]. In this case, the true (real) signal (radar measurement signal) y=[y 0 y 1 y 2 y 3 y 4]T will have a different amplitude (and thus a different scaling factor k) and additional noise n i y i= k·a i+ n iThe noise can be further decomposed into a general noise level n and a frequency-specific variation term e i: y i= k·a i+ n+e i.Thus, the signals y = k·a + n·1 + e adjacent to the peak value can be obtained, whereinLet the signals adjacent to the peak be switched to y = A·k n+ e.The value of k n= [ k n] T can be determined by means of least squares, for example.However, the direct use of the least squares method can result in a k n that fits well into an arbitrary y.It is therefore proposed to add some restrictions for k n to the optimization problem. For example, such a restriction can be in the form of a loss function (engl. The following are added: where q k and q n are predefined weighting factors in order to restrict the values k and n. It should be noted that the constraint of k close to 1 is based on the normalized version of the vectors y=y / ||y|| and a=a / ||a|| involved in the calculation. The constraint term n 2 denotes the expectation of the noise level compared to the signal of the target object. Therefore, the factor q n can be adjusted according to the expected measured signal-to-noise ratio.The minimizer of L(k n) is similar to the unrestricted least squares solution: k̂ n= ( A T A) +-1( A T y) +, where the new matrices (A T A) + and (A T y) + are computed as: andThe residual error is still to be obtained without the constraint term as:The error sources of ||e|| can be categorized into two classes: the deviation of the main lobe from model A and other signal noise. In detail, the influence of the deviation of the shape of the main lobe will be dominant for target objects with good SNR (signal-to-noise ratio). This is believed to result primarily from the flattening effect of range migration. It can also be observed through numerical calculations that the flattening effect of the range migration becomes stronger as the relative velocity v becomes higher. Thus, ||e|| may be used as an error measure to reflect the underlying true velocity of the target object. In FIG. 2, an example of the error amount in which the error monotonously increases with increasing speed is shown.In a further step, the true speed is now determined. Here, the velocity value estimated by spectral analysis is referred to as v f. The true velocity v will be a value that is a shift of multiple maximum unique velocities v max i.e., v = v f+ h · v max, where h is the integer (integer) describing the true interval in which the true velocity v lies. The error measure of the range migration artifacts (as defined above) may be used to set thresholds and determine the value of h. Two cases can be distinguished for this:In the first case, a single reference speed and a plurality of limit values are predefined. In this case, the matrix A is generated for an ideal response of the Doppler window, i.e. at a velocity of the target object of v=0. Then, at each boundary v i= h·v max a neighborhood y i to be observed is generated at a defined SNR and the error ||e i|| is calculated. The error limit can be used as a limit to determine the speed interval within which the true measurement of y v is to fall.As indicated in FIG. 2, in the case where the error ||e v|| calculated with the observation vector y v is within two reference errors, i.e. ||e i|| < || e v|| < || e i+1||, the true speed v is likewise between the two corresponding speeds v i and v i+1, which were used to obtain the errors ||e i|| and ||e i+1||. That is, v i< v < v i+1. This solves the speed ambiguity.In the second case, a plurality of reference speeds are specified, as shown in FIG. 3. All interval boundaries, i.e. v max, 2 v max, 3 v max etc., are selected as reference speeds. By evaluating the error measure of the observation vector y v( neighborhood of the peak value as a result of the Doppler FFT for a signal observed at the velocity v) with all reference velocities, the two best reference velocities (here v max and 2 v max) can be used to calculate a weighted sum position v w. The hypothesis h opt having the smallest distance between v opt= v f+ h opt · v max and v w is considered to be the true velocity indicated by y v in FIG. 3.For example, a model matrix A h is set at a boundary h·v max for each speed. During application, the observation vector y v for each reference speed is calculated from: ||e h|| = || y v- A h ·k̂ n,h||If it is provided that the two smallest errors ||e 1st|| and ||e 2nd|| correspond to the interval index h 1 and h 2 a weighted sub-interval h w can be calculated: wherein the weights w 1st( for h 1st) and w 2nd( for h 2nd) are evaluated with the corresponding error ||e 1st|| and ||e2nd| via:The parameter b is a predefined constant that avoids being divided by zero and regulates the influence of the error ||e h||. S is a normalization factor which ensures that w 1st+ w 2nd= 1, i.e.:The weighted sum h w yields a sub-interval value which is between h 1st and h 2nd and closer to that having the smallest error.Subsequently, the true interval h opt can be obtained by the following equation:In a further embodiment, the sign of the speed is determined by dividing the Doppler FFT into a plurality of subfunctions FFT 1, FFT 2 (shown in FIG. 5 ). In automotive applications, both positive and negative speeds are often of interest. The method described above may determine the best hypothesis h opt at all positive intervals, or at all negative intervals. However, it cannot determine the best hypothesis h opt at positive and negative intervals at the same time, since the distance window W R[ n f] applied is usually an even function (i.e., symmetric about W R[0] due to the condition of zero group delay). Also, at the range gate RG (range gate) where the highest Doppler spectral response is expected, the response of the range window is:At a particular chirp index mt, the frequency shift f RM+ and f RM- due to range migration of positive and negative speeds has the following property:Due to the symmetry of W a[ n f]:Consequently, the following relationship applies at the range gate:In summary, a target moving away from the sensor causes the same signal as a target moving toward the sensor at the same absolute speed. In order to determine the sign of the speed, two methods are proposed below which are based on the knowledge of the true absolute speed, as determined in the method described above.In a first method, after the true absolute velocity v opt= v f+ h opt · v max is determined at the range gate with the highest frequency response, two rotated integrations are performed:The range gate index G(mt) on a chirp is not constant, but depends on the chirp index. A first integration S D,1 starts at the range gate at which the shift of the range gate per chirp due to the velocity is. Samples from the following range gates are then used on the mt-th chirp:The second integration S D,2 starts at the range gate and fetches the samples.The above integrations can be effectively applied to two Doppler FFTs on f D both along the assumed respective range migration lanes across the chirps, i.e. different from the classic Doppler FFT performed along the chirps in a range gate.The sign of the speed can be obtained by: |S D,1| - | S D,2|Another option is to evaluate the result ||S D,0|| at f D obtained by classic Doppler FFT, but along the chirps at the range gate The sign is obtained by the sign of |S (D,1)|-| S (D,0)| or |S (D,0)|-| S (0,2)|.The first method cannot be applied when using a radar system that processes data only range gate by range gate, in particular when the data of a range gate no longer exists after processing. Here, data of multiple range gates required for the first method is not available. Therefore, in a second method, the Doppler FFT is performed in the same range gate across the chirps. But the complete FFT S[n f, m f] will be performed in two shorter FFTs S 2,0 and S 2,1 as follows:When a target object is detected in a range gate (denoted as RG1), its Doppler bin m D( Doppler gate) is stored. In a subsequent range gate (denoted as RG2), the sign of |S 2,0[ RG2, m D]| - | S 2,1[ RG2, m D]| is checked. A positive value indicates that the target object is moving toward the radar sensor.The selection of RG2 depends on the frequency response of the range window. Advantageously, RG2 is to be chosen at RG1+offset (offset), the offset corresponding to the distance between the peak value and the first valley in the frequency response.In a further embodiment, double peak values are determined by dividing the Doppler FFT into a plurality of sub-functions. Based on the above methods, knowledge of a correct Doppler bin of the target object is to be expected, which is typically realized via a processing chain by peak detection, CFAR, etc. However, in case the true velocity is very high, e.g. over 60 m / s, double peaks appear around the true Doppler bin, as seen in Fig. 4. Methods such as peak detection cannot be used to determine the correct Doppler location of the target object.The double peaks appear only when the effect of the range migration is significant, i.e. for target objects with high relative speeds to the radar, i.e. e.g. >60m / s. In this case, the main lobe of the frequency response of the range window at the range gate, e.g. at which a peak has been detected, will be in the average chirp samples of the range gate. The preceding and trailing chirps primarily sample the side lobes of the frequency response of the range window. Consequently, the result of the Doppler FFT on the samples of the middle chirps will have a significantly higher response than the result of the Doppler FFT on the samples of the preceding and trailing chirps.To make use of this fact, the full Doppler FFT is split into four short FFTs in a range gate:Unlike decimation in time FFT, the simple partitioning cannot take advantage of all of the symmetry of the twiddle factors (IF). However, the rotating factors have phases of a plurality, so that the complex multiplication therewith is only a reversal of sign or setting of output values to zero, or exchanging real and imaginary components, and does not introduce any further complex multiplication.In summary, the following steps may be performed to handle Doppler peaks, as shown in FIG. 5 : 1. in a range gate, if two peaks are detected very close to each other (e.g., |m f,2- m f,1| ≤ 5) along the Doppler dimension and after the full Doppler FFT, and if their energy values are also very similar (e.g., within an absolute difference of ≤ 2dB), they are held for double peaks and processed further. 2. the following energy ratio for the double peaks is determined: 3. If both and greater than a threshold TΔS2, the peaks at m f,1 and m f,2 are determined as double peaks due to distance migration. The true Doppler location is found midway between m f,1 and m f,2.The choice of the limit TΔS2sollte take into account both SNR levels and the shape of the Doppler window.In a further embodiment, scanning signals of signals from radar measurements are aligned with scanning signals of the radar reference signal. Thus, the peaks of the main lobes are superposed on each other. This enables a higher measurement accuracy. The effectiveness of the fitting formula k̂ n= ( A T A) +-1( A T y) + suggests that the data a=[a 0 a 1 a 2 a 3 a 4]T for calculating the model A and the data in signal y=[y0y1y2y3y3y4]Tare aligned with each other, thus, for example, both a2 and y2 coincide with the peak value of the frequency response window. However, during actual application, the theoretical peak value of the real observed data may be at any location between two adjacent sampling points. By curve fitting the data vector y and performing re-sampling of the fitted curve, the data vector y and the model vector a can be adjusted.For example, B-spline may be applied to perform the curve fitting: where B(x) is the spline basis function.Assuming that the original scanning position y 0 in the neighborhood vector y corresponds to x s, then the optimum scanning position x opt, to be matched with the model vector a, is supposed to satisfy: -0.5 ≤ x opt- x s ≤ 0.5For the model vector where the center a2in arepresents the peak in the response, the resampled data vector y shall also have a peak at y(x opt+ 2). In addition, due to the symmetry of the frequency response window of the main lobe, the optimum resample point shall be: wherein the first term p 0| y(x+2)-0.5(y(x+1)+y(x+3))| expresses that the true peak value (or the valley in the case of Doppler peaks) has the greatest difference from its adjacent sample points. The second and third terms p 1| y(x+1)-y(x+3)|+p 2| y(x)-y(x+4)| limit the symmetry shape of the main lobe.p0, p1, p2 are three predetermined weights which control the contribution of the difference of the peak values and the symmetry constraint for the optimization. More specifically, when p1=p2=0, the optimization degrades to the case of searching the highest (or lowest in the case of double peak values) value on the fitted curve as the location for (x opt+ 2).The method is used in the field of semi-autonomous to autonomous driving. For this purpose, a vehicle is equipped with a corresponding sensor system and actuator system. It also has at least one control device on which the described method or methods are implemented as a computer program.The detection of distances, angles and relative speeds to target objects, i.e. objects located in the range of vision of the radar sensor of the vehicle, is necessary in order to avoid rear-end collisions, for example.Reference numerals:B e Effective bandwidth at each chirp c Speed of light N, M Number of analog-to-digital converter samples in a chirp; Number of chirps Δt analog-to-digital converter sampling intervals within a chirp R of radial far edge of the target object at the beginning of the measurement cycle RG range gate W R[ n f] frequency response of the distance window at the n f- th range gate w D[ m f] frequency response of the distance window at the m f- th Doppler bin W R[ n t] nt-th coefficient of the distance window in the time domain w D[ m t] mt-th coefficient of the Doppler window in the time domain w DR[ n f, m t] impulse response of the Doppler window mt chirp index f C carrier frequency f R superposition frequency f RM[ mt] shift of the superposition frequency due to range migration at the mt-th chirp v true speed of the target object v f ambiguous speed estimation (vmaxunique speed interval converted by the spectral analysis-if the true speed lies within this interval, v= vf x [n t, m t] ( clomplexes) time signal of the nt-th sample in the mt-th chirp S [n f, m f] DFT sample at the n f- th range gate and the m f- th Doppler bin T p pulse repetition time h i: index of the ambiguous intervals of interest of velocity ||e|| error measure y v observation vector (neighborhood of the peak value as a result of the Doppler FFT for a signal observed at velocity v)References included in the specificationThis list of documents cited by the applicant has been produced in an automated manner and is only included for the better information of the reader. The list is not part of the German patent application or utility model application. The DPMA does not take any adhesion for any faults or omissions.Patent Literature citedWO 2020 / 201 337 A1

[0007]

Claims

Method for determining a true speed of a target object detected by means of radar measurement of an FMCW radar within a single measurement cycle, comprising the steps of: - generating at least one reference signal for at least one specified reference speed, and - generating an error measure (||e||) which quantitates the effect of the deviation of the range migration on the frequency response of the Doppler window, - using the error measure (||e||) in order to determine in which of the ambiguous intervals present on the basis of the sampling rate the true speed of the target object falls.Method according to claim 1, wherein in the event that a single reference speed is provided, the error measure (||e||) is used in such a way that a signal (y i) adjacent to the peak value is generated at each boundary of a speed interval which is a multiple of the maximum detectable speed (v i= h·v max) of the radar measurement and the associated error measure (||e i||) is calculated, wherein the speed interval at which the error measure (||e||) is the smallest is determined as the interval in which the true speed of the target object falls.Method according to Claim 1, wherein, in the case where a plurality of reference speeds are provided, the error measure (||e||) is used in such a way that an associated error measure (||e 1st||; || e 2nd||) of the measured speed of the target object with respect to the reference speeds is calculated at each boundary of a speed interval of each of the reference speeds, and wherein a weighting of each error measure (||e 1st||; || e 2nd||) per reference speed is carried out, and wherein the speed interval at which the error measure (||e 1st||; ||e 2nd||) when the interval in which the true speed of the target object falls is determinedMethod according to one of the preceding claims, wherein the error measure (||e||) serves as a measure for the true speed of the target object.Method according to one of the preceding claims, wherein, based on the true absolute speed, the sign of the speed of the detected target object is determined by dividing the fast Fourier transformation function into two sub-functions, wherein the sub-functions each correspond to a half-range gate.Method according to one of the preceding claims, wherein the fast Fourier transformation function is divided into two sub-functions, and if one range gate is processed after the other, in the case that a target object is detected, the Doppler bin (m D) is stored at a first range gate (RG1) and the sign of the result of the difference of the two sub-functions |S 2,0[ RG2, m D]| - | S 2,1[ RG2, m D]| is checked in a subsequent range gate (RG2), wherein in the case of a positive value it is detected, the target object is moved toward the radar sensor and it is detected in the case of a negative value that the target object is moved away from the radar sensor.Method according to one of the preceding claims, wherein double peak values of the measured radar signal are detected by evaluating an energy ratio ( Δ S n f, m f 2) of the two peak values in the case that two closely adjacent peak values are detected along a Doppler dimension after the complete FFT split into four sub-functions, and the energy values of the two peak values are close to one another, and detecting double peak values in the case that both energy ratios ( Δ S n f, m f 2) are greater than a predefined threshold value.The method of any preceding claim, wherein the specified reference speed is zero.Method according to one of the preceding claims, wherein scanning signals of signals from radar measurements are aligned with scanning signals of the radar reference signal.

Images