Method for obtaining the offset of a radar sensor

Abstract

A method for detecting misalignment of radar sensors (4, 6) arranged on a vehicle (1), wherein - determining a Doppler spectrum with respect to the beam transmitted and received by the radar sensor; - determining an angle of incidence of at least one frequency band with respect to the Doppler spectrum at least in a partial interval - comparing the determined angle of incidence with the expected angle of incidence with respect to the frequency band; - identifying misalignment of the radar sensor based on the deviation between the measured angle of incidence and the expected angle of incidence.

Description

Field of the Invention

[0001] The present invention relates to a method for detecting the misalignment (Dejustage) of a radar sensor arranged on a vehicle. Furthermore, the present invention relates to a sensor arrangement, a computer program, and a machine-readable storage medium. Background Art

[0002] Due to the increasing level of automation in vehicles, the number of radar sensors installed in vehicles continues to increase. In particular, each vehicle usually has multiple radar sensors installed. For example, distance adjustment and object recognition can be performed through radar sensors.

[0003] In the case of an increasing number of radar sensors, the manual adjustment overhead for each radar sensor is large and cost-intensive. Therefore, currently used radar sensors are only placed in the corresponding holders without adjustment.

[0004] Calibration of the radar sensor at the tape end is also increasingly being cancelled. Calibration is performed by measuring the orientation of the radar sensor. Alternatively, there is an increasing requirement for the radar sensor to determine its own orientation within the shortest possible time during normal driving operation. Thereby, it is also possible to identify a change in the orientation of the radar sensor in its holder - for example, after a parking push (Parkrempler).

[0005] Methods are known for performing simultaneous estimation of the own speed of a vehicle with a radar sensor, the orientation of the radar sensor relative to the driving direction, and the misalignment of the radar sensor relative to the nominal position. The estimation is performed using targets or objects classified as stationary. Here, the angle of incidence of the reflected radar beam can be compared with the measured relative speed. If the stationary object is not directly in the driving direction, its relative speed with respect to the radar sensor is reduced by the cosine value of the corresponding angle.

[0006] Due to the low speed resolution and angle resolution, time-consuming averaging is required in the known methods. In addition, the classification of objects increases the computational overhead for analyzing the measurement data of the radar sensor.

[0007] Furthermore, dedicated radar systems are also known that can determine the direction of movement in addition to the speed of movement. However, in this way, it is not possible to estimate the misalignment of physically separate radar sensors, for example, for driver assistance functions. Summary of the Invention

[0008] The task underlying the present invention can be considered to be to provide a fast and accurate method for determining the misalignment of a radar sensor.

[0009] This task is solved by means of the technical solution according to the present invention. Advantageous configurations of the present invention are respectively preferred embodiments.

[0010] A method for detecting a misalignment of a radar sensor arranged on a vehicle is proposed. Here, a Doppler spectrum for the beam transmitted and received by the radar sensor is determined. For at least one frequency bin in the Doppler spectrum, the angle of incidence is determined at least in a partial interval. The determined angle of incidence is compared with the expected angle of incidence for this frequency bin. The expected angle of incidence is stored in a data memory, for example. The misalignment of the radar sensor is identified based on the deviation between the measured angle of incidence and the expected angle of incidence. In addition, this deviation can be used as a measure of the misalignment.

[0011] In one embodiment, an angle of incidence is determined for a plurality of frequency bins of the Doppler spectrum respectively. The determined angle of incidence is compared with the expected angle of incidence of the frequency bin. The deviation between the determined angle of incidence and the expected angle of incidence is used to identify the misalignment of the radar sensor, in particular to determine a measure of the misalignment of the radar sensor.

[0012] In one embodiment, the radar sensor has a plurality of transmit and / or receive antennas, wherein the angle of incidence of the at least one frequency bin is determined by means of the transmit and / or receive antennas. In particular, the propagation time difference of the received signals of the individual receive antennas can be used to determine the angle of incidence of the frequency bin. The propagation time difference corresponds to a phase difference that can be determined separately for each frequency bin in a frequency range.

[0013] In one embodiment, the partial interval has a spectrum that includes the following Doppler frequencies: the Doppler frequencies correspond to a speed range between 0 and the vehicle speed.

[0014] In one embodiment, the partial interval is determined based on the direction of movement of the vehicle, the nominal adjustment, and the opening angle of the radar sensor.

[0015] In one embodiment, the direction of movement of the vehicle is determined by means of at least one sensor at the vehicle, and / or the deviation during straight-line driving is compensated by time averaging.

[0016] In one embodiment, the determined angle of incidence for the frequency bin is dominated by the power reflected on at least one stationary object, wherein the frequency bin is determined by the relative speed between the radar sensor and the stationary object.

[0017] In one embodiment, for the average deviation between the determined angle of incidence and the expected angle of incidence, not only the quality of the determined angle of incidence is considered, but also the value of the determined angle of incidence.

[0018] The quality of the determined angle of incidence can be determined by estimating the probability of whether the determined angle of incidence actually exists. The existence probability can be directly used for weighting.

[0019] For each frequency segment in a pre-given frequency interval for the Doppler frequency, the deviation between the determined angle of incidence and the expected angle of incidence can be analyzed, where the frequency segment corresponds to all possible measurable relative velocities caused by the vehicle speed of a stationary object. The accuracy of determining the deviation can be improved by averaging the deviations of multiple frequency segments.

[0020] In one embodiment, in the averaged deviation, compared to the deviation of the determined angle of incidence from the expected angle of incidence of the frequency segment with a smaller angle of incidence, the deviation of the determined angle of incidence from the expected angle of incidence of the frequency segment with a larger angle of incidence relative to the driving direction is weighted more strongly. For this purpose, a corresponding graph, characteristic curve or formula can be set, which determines the factor of the deviation with respect to different expected angles of incidence. The graph, characteristic curve or formula is stored, for example, in the memory of the control unit.

[0021] To take into account the value of the angle of incidence, the frequency segment with a large angle relative to the driving direction F can be weighted more strongly, because at a large angle a small change in the angle already results in a relatively large change in the relative velocities vr1, vr2 and the Doppler frequency. For example, it can be used as a weighting function Here, it concerns the angle relative to the driving direction

[0022] In one embodiment, deviations higher than a pre-given limit value between the determined angle of incidence and the expected angle of incidence of the frequency segment are not considered in the averaged deviation. Thereby, deviations in the frequency segments generated by moving objects can be excluded. The limit value is determined in advance by experiments and stored, for example, in the memory of the control unit.

[0023] For example, the own range (Spanne) of the expected angle of incidence can also be calculated from the cosine effect for each frequency segment in the Doppler spectrum. If the measured value is outside this range, it is not considered. Alternatively, a constant upper limit (e.g., 5° or 10°) can simply be used for the allowed deviation.

[0024] According to another aspect, a control unit is provided, which is configured to implement this method.

[0025] According to another aspect of the present invention, there is provided a computer program. The computer program includes instructions which, when the computer program is implemented by a computer or a control unit, cause the computer or the control unit to implement the method.

[0026] Furthermore, according to one aspect of the present invention, there is provided a machine-readable storage medium on which the computer program is stored.

[0027] By analyzing the Doppler spectrum, the own speed of the vehicle can be determined. This can be done independently of the exact installation situation of the radar sensor (i.e., adjustment and possible misalignment). For this purpose, the fact is fully utilized that no stationary object can move faster than the vehicle's speed as seen by the radar sensor. Accordingly, a sudden power drop is searched for in the Doppler spectrum at the (negative) own speed of the vehicle.

[0028] If, in addition to the power, the measured angle of incidence is also considered, the direction of movement can be estimated in addition to the speed of movement. Here, the direction of movement is related to the emission direction of the radar sensor. To determine the misalignment, the direction of movement also needs to be compared with the current direction of movement of the vehicle accordingly and corrected with the nominal position of the radar sensor.

[0029] The direction of movement of the vehicle can be determined by means of additional vehicle-side sensors. However, alternatively, it can also be considered that the vehicle travels approximately in a straight line most of the time. Therefore, the change in the direction of the vehicle can also be compensated by observation over time (i.e., by averaging over time).

[0030] By comparing the direction of movement obtained by the radar sensor with the direction of movement provided by the vehicle or obtained by another sensor, it is possible to determine the misalignment of the radar sensor. Thereby, automated misalignment recognition can be provided, which can avoid pre-adjustment or calibration of the radar sensor. Therefore, vehicle systems based on radar sensors (such as assistance systems) can be configured more economically.

[0031] In the case where the transmitted signal generated by the radar sensor has a constant frequency, only the Doppler effect causes a difference between the transmitted frequency and the received frequency. Here, the distance of the object is irrelevant. Therefore, each relative speed corresponds to a Doppler frequency. Therefore, based on the speed of movement, a defined frequency range in the Doppler spectrum can be determined, and for this purpose, a more accurate analysis of the angle of incidence is worthwhile.

[0032] For a plausibility check, in addition to power analysis, a simple angular observation can be performed. The stationary object with the highest relative speed (i.e., the vehicle speed) is directly in the driving direction. Thus, for a front radar sensor, regardless of a possible slight misalignment, it can be assumed that the reception angle should be approximately zero at (negative) vehicle speed. For an object "directly in front", the horizontal and vertical offsets must ultimately both be zero.

[0033] However, for the remaining speed ranges the following applies: All points on a cone oriented rotationally symmetrically around the driving direction have the same relative speed with respect to the radar sensor. In principle, thus the power of stationary objects with an infinite number of horizontal and vertical offset combinations can overlap in the so-called frequency bands of the Doppler spectrum.

[0034] As long as one of the two opening angles of the radar sensor is only a few degrees, the cosine effect in this direction can be neglected. For example, the cosine of 8° corresponds to the value 0.99. Thus, for example, the cosine effect with respect to the elevation angle of the radar sensor can be neglected. Thereby, the three-dimensional cone can be reduced to two intersecting lines with the "horizontal" sensor plane.

[0035] According to one embodiment of the method, the radar sensor arranged on the vehicle has a plurality of transmit antennas and / or receive antennas in order to determine the angle of incidence.

[0036] According to another embodiment of the method, the direction of movement of the vehicle is determined by at least one sensor on the vehicle. Thus, most vehicles have a rotation and yaw rate sensor - for example for an electronic stability program. These vehicle sensors can be coupled conductively with the control unit or connected indirectly via an existing control unit. Thereby, the control unit can directly read and analyze the vehicle sensors or use the measurement data provided by other control units.

[0037] Alternatively or additionally, the deviation during straight-line driving can be compensated for by time averaging. It can be assumed that the vehicle drives straight for most of the time. The angular offset can be calculated by comparing the direction of movement determined by the radar sensor with the direction of movement determined by the above-mentioned vehicle sensors. This angular offset is compared with the angular offset expected for the nominal position. If the difference exceeds, for example, a defined limit value, the misalignment is determined by the control unit. Thus, the misalignment of the radar sensor can be defined as the deviation from the nominal installation.

[0038] According to another embodiment of the method, the radar sensor arranged on the vehicle is arranged at the front of the vehicle, at the rear of the vehicle or at the side of the vehicle. In addition, a radar sensor configured as an angle sensor (Ecksensor) can also be used.

[0039] Regarding the offset estimation, radar sensors installed at the sides and in the corners have the following advantages: Objects located on opposite sides of the vehicle are usually not within the field of view of the vehicle. Correspondingly, for angle observation, only one intersection line with the "horizontal" sensor plane remains, and the observation angle can be clearly predicted for each relative velocity. This simplifies the angle estimation and improves the accuracy of the offset estimation.

[0040] Furthermore, due to the rotationally mounted installation of such sensors with respect to the driving direction, even in the case of a large angle with respect to the driving direction, a lot of power can be received from stationary objects for such sensors. This is caused by the fact that the corresponding observation angle lies within the main lobe (Hauptkeule) of the transmitting antenna of the angle sensor or the object is detected with this main lobe during passing. This additionally improves the accuracy of the offset estimation. Therefore, the frequency band in the Doppler spectrum does not correspond to exactly one single relative velocity, but to a small velocity interval. However, due to the course of the cosine function with a large slope for angles far from zero, this slight ambiguity in velocity is less pronounced compared to the corresponding ambiguity in the received angle.

[0041] Finally, for such radar sensors, even objects located at the rear side can still be within the field of view. Therefore, the frequency / velocity range to be analyzed extends in the direction of positive relative velocity. Thus, in addition to the magnitude of the own velocity, the interval of the Doppler frequencies that can be analyzed is also determined by the opening angle and the mounting angle of the corresponding radar sensor.

[0042] Similar to the front sensors, radar sensors configured as rear sensors can be considered, where the reverse sign of the own velocity is taken into account.

[0043] If the initial assumption is violated and the assumed own velocity is not exactly correct, an additional offset will result, which changes from one frequency band to the next. Although this requires a more accurate analysis, in this way, the assumed own velocity can also be corrected retrospectively.

[0044] According to another embodiment of the method, the radiation emitted by at least one radar antenna of the radar sensor has a small vertical opening angle and / or a vertical emission angle with respect to the vehicle chassis. The antenna can thus emit the transmitted power starting from the vehicle chassis slightly tilted upwards. This can additionally minimize the influence of the roadway.

[0045] The reflected power of all stationary objects at an angle to the direction of travel (or in a small range around the direction of travel) falls within the same frequency band in the Doppler spectrum because the stationary objects have the same relative velocity with respect to the sensor. All objects within the range of the radar sensor can be located on the surface of a cone that opens along the direction of travel. Thus, an infinite number of combinations of, for example, azimuth and elevation angles result. If the radar sensor only scans the horizontal plane, the cone reduces to two intersecting lines with that plane. Thus, the remaining (azimuth) angles are unique except for the sign. The objects are only located at the same angle to the left or right of the direction of travel. Finally, since the opposite sides are not within the field of view of the radar sensor, this ambiguity typically cancels itself out for an angular radar.

[0046] The more different incident directions overlap in a frequency band, the more difficult it is to estimate the angle. However, two angles can be determined reliably simultaneously. Additionally, it is advantageous if the expected angles are already known beforehand. This can be achieved, for example, by restricting the search range.

[0047] According to another embodiment, the deviation averaged by the control unit over all analyzed frequency bands between the determined angle of incidence and the expected angle of incidence takes into account the quality and absolute value of the angle of incidence determined in the frequency band. Additionally, large deviations in individual frequency bands indicate that the power in that frequency band is dominated by moving objects, and thus these powers should not be considered when averaging.

[0048] According to one configuration, a method for detecting a misalignment of a radar sensor on a vehicle side is provided. In one step, the direction of movement of a radar sensor arranged on a vehicle is determined by the radar sensor. Subsequently or in parallel, the direction of movement of the vehicle is determined by at least one vehicle-side acceleration sensor. A comparison can be performed by the control unit based on the independently determined directions of movement of the vehicle. In the case of a determined deviation between the two determined directions of movement, the control unit identifies a misalignment. Description of the Drawings

[0049] The preferred embodiments of the present invention are further explained below based on highly simplified schematic diagrams. The drawings show:

[0050] Figure 1 A schematic diagram of a vehicle with a sensor arrangement in the case of straight-line vehicle travel;

[0051] Figure 2 A schematic diagram of a vehicle with a sensor arrangement in the case of a rotatably arranged radar sensor;

[0052] Figure 3 A schematic diagram of an angular sensor of a sensor arrangement in the case of straight-line travel;

[0053] Figure 4 Schematic view of a vehicle with a sensor arrangement in the case of straight - line driving with a symmetric object arrangement;

[0054] Figure 5 Schematic view of an angular sensor of a sensor arrangement in the case of a rotationally arranged radar sensor with a symmetric object arrangement;

[0055] Figure 6 Schematic view of a radar sensor;

[0056] Figure 7 Showing the received power of the reflected and received radar signals plotted against the relative speed v α / v.

[0057] In the figures, identical structural elements have the same reference numerals respectively. Detailed description

[0058] Figure 1 Schematic view of a vehicle 1 with a sensor arrangement 2 in the case of straight - line driving of the vehicle 1. According to this embodiment, the sensor arrangement 2 has a first radar sensor 4, which is arranged at the front of the vehicle. A second radar sensor 6 is configured as an angular sensor and is positioned at the transition between the front of the vehicle and the right side of the vehicle 1 in the driving direction F.

[0059] Furthermore, the sensor arrangement 2 has an acceleration sensor 8 on the vehicle side. The sensors 4, 6, 8 of the sensor arrangement 2 are coupled in a data - conductive manner to a control unit 10. Thereby, the control unit can read the sensors 4, 6, 8, and can analyze the measured values of the sensors, and can execute a method for determining the misalignment of at least one of the radar sensors 4, 6.

[0060] For clarity, only the scanning area A of the first radar sensor 4 is shown. In the scanning area A, two objects SO1, SO2 are arranged asymmetrically with respect to the vehicle 1.

[0061] From the perspective of the vehicle 1, the stationary objects SO1, SO2 move towards the vehicle 1 with a speed v ego For the objects SO1, SO2 with a lateral (and vertical) offset, the relative speeds v r1 , v r2 decrease with the cosine of the angle with respect to the driving direction F. Thus, overall, the relative speeds v r1 , v r2 are distributed over the interval [-v ego ; 0].

[0062] For a radar sensor 4 with an exact forward orientation, when driving straight, the angle with respect to the driving direction or the direction of movement of the vehicle 1 directly corresponds to the observation / reception angle:

[0063] Or

[0064] In this case, for two azimuth angles that belong to a determined relative speed or Doppler frequency due to the cosine effect Apply:

[0065]

[0066] Figure 2 Schematic illustration of a vehicle 1 with a sensor arrangement 2 in the case of a rotationally arranged radar sensor 4.

[0067] The horizontal misalignment of the radar sensor 4 mathematically results in an angular deviation

[0068]

[0069] However, this angular deviation also occurs for the deviation during straight driving.

[0070] Figure 3 Schematic illustration of the angle sensor 6 of the sensor arrangement 2 in the case of the straight driving of the vehicle 1. Additionally, the scanning area A of the angle sensor 6 is shown.

[0071] Due to the rotationally mounted position of the angle sensor 6 with respect to the driving direction F, even at a large angle with respect to the driving direction High power of stationary objects SO1, SO2 can be received, because the corresponding observation angle Is located in the main lobe of the antenna.

[0072] Another object SO1' is arranged behind the object SO1, where the reflected power overlaps with the reflected power of the object SO1. Therefore, the two powers obtained fall into the same frequency band in the Doppler spectrum, because the two powers have the same relative speed with respect to the sensor 6.

[0073] Figure 4 Schematic illustration of a vehicle 1 with a sensor arrangement 2 in the case of straight driving with a symmetric object arrangement. Thus, the two objects SO1, SO2 are arranged at symmetric angles with respect to the vehicle 1. From this, the following relationship can be obtained:

[0074]

[0075] Figure 5Schematic illustration of the front sensor 4 of the sensor arrangement 2 in the case of a rotationally arranged radar sensor 4 with a symmetric object arrangement.

[0076] For each analyzed frequency band, the angle relative to the driving direction is known in advance and this angle is stored in the data memory of the control device. Thus, the known angles of the frequency bands can be used to identify misalignment of the radar sensor. For this purpose, the known angles of the frequency bands are compared with the measured angles of the same frequency bands. From this comparison, the angle deviation can be determined Angle deviation can be the same for multiple or each analyzed frequency band.

[0077] For the combination of values for the individual frequency bands, it is proposed here to minimize the square of the error. When weighting these errors, the following factors can be considered, for example:

[0078] - The quality of the estimation of the observed angle in each frequency band;

[0079] - Frequency bands with a large angle relative to the driving direction F are weighted more strongly because in the case of a large angle a small angle change already results in a relatively large change in the relative speed v r1 、v r2 and the Doppler frequency (see );

[0080] - In the case of large deviations, the angle of incidence can originate from a moving object and is therefore ignored by the control unit.

[0081] For example, the quality value of the estimation of the observed angle for each frequency band can be stored in the data memory of the control device. In addition, the following angle values can be stored in the data memory: the angle value indicating since when the angle relative to the driving direction F is classified as large and weighted more strongly. In addition, the following angle values regarding large deviations of the angle of incidence can be stored: the angle value indicating that the received signal comes from a moving object and is therefore ignored by the control unit.

[0082] In addition, in all these methods, temporal averaging of the measured values for the individual frequency bands can also be performed.

[0083] If two angles of incidence of the stationary objects SO1, SO2 can be determined in the frequency band then prior knowledge of the own speed v ego is not absolutely necessary. Not only the own speed v of the vehicle 1 ego, and the misalignment or the driving direction F can be directly determined by the differential or the average value of two observation angles .

[0084] If then the following applies:

[0085]

[0086]

[0087] The movement of the vehicle 1 is usually defined not with respect to the sensors 4 and 6, but with respect to the midpoint of the rear axle. However, from this, the movement of any other point on the rigid vehicle 1 can be derived. Additionally, during straight-line driving, the movement vectors of all points on the vehicle 1 are the same.

[0088] If the used speed v of the vehicle ego differs from the actual speed, it is also possible to wrongly identify the misalignment of the radar sensor. The deviation of the ideal angle with respect to the direction of movement is greatest near the negative own speed v ego . Therefore, the frequency band in the Doppler spectrum contains most of the information about the error in the own speed estimation.

[0089] Figure 6 The block diagram of the radar sensor is shown schematically. The high-frequency oscillator 18 is set as follows: the frequency of the high-frequency oscillator is controllable. The high-frequency oscillator 18 generates a transmit signal, which reaches the antenna 22 through the mixer 20 and is then transmitted by the antenna as a radar lobe. The radar echo generated by an object in the vehicle's surroundings is received by the antenna 22 and mixed in the mixer 20 with a part of the transmit signal generated by the high-frequency oscillator 18 at the reception moment. Thereby, an intermediate-frequency signal 24 is obtained by beat (Schwebung) for further analysis in the analysis processing unit 26. The frequency of the transmit signal generated by the high-frequency oscillator 18 is modulated and forms a series of CW signal ramps with a constant frequency. Therefore, it relates to a CW radar transmit signal. If an extended object located at a certain distance in front of the vehicle is located, the time spacing of the ramps determines the frequency difference of the following signals: the signals are mixed with each other in the mixer 20 and the frequency of the intermediate-frequency signal 24 is obtained. When the vehicle moving in front moves relative to the radar sensor, the frequency difference depends on the Doppler shift, which in turn depends on the relative speed.

[0090] First, the intermediate frequency signal 24 is sampled and digitized into a time signal, and then it is converted into a Fourier spectrum, for example, by means of a fast Fourier transform. In this spectrum, each located object is presented in the form of a peak at a determined frequency, which depends on the spacing and relative speed of the object. If the same object is located once on the rising ramp and later again on the falling ramp, the frequencies of the two peaks can be added. Since the ramps have opposite slopes, the distance-related parts cancel each other out here. Thus, only the Doppler component depending on the relative speed remains. Conversely, if the frequencies of the two peaks are now subtracted, the speed-related components cancel each other out, thus obtaining a pure distance component, which allows the determination of the distance of the object. Usually, more than two modulation sequences or at least two ramps with different slopes are used. If there are two or more objects, this makes it easier to assign the peaks obtained in the spectrum to the relevant objects.

[0091] The analysis and processing unit 26 is used to analyze the spectrum of the intermediate frequency signal 24. The discrete spectrum of the intermediate frequency signal is used and analyzed for digital analysis. Correspondingly, the frequency axis is divided into a finite number of frequency segments, and the spectrum is a discrete function that describes the power assigned to each frequency segment.

[0092] Figure 7 The received power P of the reflected and received radar signal plotted with respect to the relative speed v α / v is shown in decibels (dB). Here, the angle α determines the angular deviation between the driving direction of the vehicle and the connecting direction to the object. The own speed of the vehicle is denoted by v. The relative speed of the vehicle with respect to the object is obtained from the following formula: v α = -v·cos(α). The measured relative speed v of the object is illustrated on the x-axis α = -v·cos(α) and the quotient of the own speed v of the vehicle. Thus, the power reaches its maximum at v α / v = -1. For smaller (numerically larger) relative speeds, the signal suddenly drops to 0.

[0093] When the angle α approaches the limit value 0, the slope of the cosine function becomes smaller and smaller. As a result, the relative speeds obtained for different angles α become more and more similar and thus the radar echoes of more and more scatter centers fall into the same frequency segment. This leads to the following result: As the limit value -1 is approached, the power of the frequency segment rises significantly. Thus, in the case of the relative speed v α / v = -1 with respect to α = 0, the sudden drop in power is so significant that it can be well detected in the spectrum.

Claims

1. A method for detecting a misalignment of radar sensors (4, 6) arranged at a vehicle (1), wherein, - determining a Doppler spectrum for the radiation transmitted and received by the radar sensors; - For at least one frequency band of the Doppler spectrum, determine the angle of incidence at least in a partial interval - Compare the determined angle of incidence with the expected angle of incidence for the frequency band; - identifying the misalignment of the radar sensors based on a deviation between a measured angle of incidence and an expected angle of incidence, wherein, for the determined angle of incidence for the frequency band is dominated by the power reflected at at least one stationary object (SO1, SO2), said at least one stationary object having a relative velocity corresponding to the frequency band with respect to the radar sensor (4, 6).

2. The method according to claim 1, wherein Determine the angle of incidence for a plurality of frequency segments of the Doppler spectrum respectively wherein the determined angle of incidence is compared with the expected angle of incidence of the frequency segment, wherein the deviation between the determined angle of incidence and the expected angle of incidence is determined as a measure of the misalignment of the radar sensors (4, 6).

3. The method according to claim 1 or 2, wherein The radar sensors (4, 6) have a plurality of transmit antennas and / or receive antennas, wherein the angle of incidence of the frequency band is determined by means of the transmit antennas and / or receive antennas 4. The method according to claim 1 or 2, wherein The partial interval has a spectrum containing the following Doppler frequencies: the Doppler frequencies are generated by a speed range between 0 and the speed of the vehicle.

5. The method according to claim 1 or 2, wherein Determine the partial interval based on the direction of movement of the vehicle (1), a nominal adjustment, and the opening angle of the radar sensors (4, 6).

6. The method according to claim 5, wherein, Determine the direction of movement of the vehicle (1) by means of at least one sensor (4, 6) at the vehicle (1), and / or compensate for deviations during straight-line driving by time averaging.

7. The method according to claim 2, wherein, In order to determine the averaged deviation between the determined angle of incidence and the expected angle of incidence, not only the quality of the determined angle of incidence is considered, but also the absolute value of the determined angle of incidence is taken into account.

8. The method according to claim 7, wherein, In the averaged deviation, the deviation between the determined angle of incidence and the expected angle of incidence of a frequency segment with a larger angle of incidence relative to the driving direction is weighted more strongly than the deviation between the determined angle of incidence and the expected angle of incidence of a frequency segment with a smaller angle of incidence.

9. The method according to claim 7 or 8, wherein In the averaged deviation, deviations above a pre-given limit value between the determined angle of incidence and the expected angle of incidence of the frequency segment are not considered.

10. The method according to claim 1 or 2, wherein The radar sensors emit radiation at a constant transmission frequency.

11. A control unit (10) configured to carry out the method according to any one of the preceding claims.

12. A computer program product comprising instructions which, when executed by a computer or a control unit, cause the computer or the control unit to carry out the method according to any one of claims 1 to 10.

13. A machine-readable storage medium having stored thereon instructions which, when executed by a computer or a control unit, cause the computer or the control unit to carry out the method according to any one of claims 1 to 10.

Images