METHOD FOR DETERMINING AN EGO-VELOCITY ESTIMATED VALUE AND AN ANGLE ESTIMATED VALUE OF TARGETS

Abstract

A method for determining an ego-velocity estimated value and an angle estimated value of targets using a synthetic aperture radar sensor. A distance is measured between the synthetic aperture radar sensor and each respective target. A relative velocity of the respective target is measured using the Doppler effect. An angle estimation of an angle estimated value takes place, which characterizes the angle between the direction of the ego-velocity of the synthetic aperture radar and the respective target. An individual ego-velocity estimated value of the synthetic aperture radar sensor is ascertained using the relative velocity and the angle estimated value for each target. A classification and distribution of the individual ego-velocity estimated values relating to stationary targets takes place, whose individual ego-velocity estimated values are situated within a predefinable range relative to one another, and relating to moving targets, whose individual ego-velocity estimated values are situated outside the range.

Claims

1. A method for determining an ego-velocity estimated value and an angle estimated value of targets using a synthetic aperture radar sensor, comprising the following steps: measuring, using the synthetic aperture radar sensor, a respective distance between the synthetic aperture radar sensor and each respective target; measuring, using the synthetic aperture radar sensor, a respective relative velocity of each respective target using the Doppler effect; performing an angle estimation of one respective angle estimated value for each respective target, which characterizes an angle between a direction of an ego-velocity of the synthetic aperture radar sensor and each respective target; ascertaining an individual ego-velocity estimated value of the synthetic aperture radar sensor using the respective relative velocity and the respective angle estimated value for each respective target; classifying and distributing the individual ego-velocity estimated values relating to stationary targets, whose individual ego-velocity estimated values are situated within a predefinable range relative to one another, and relating to moving targets, whose individual ego-velocity estimated values are situated outside the predefinable range; ascertaining a combined ego-velocity estimated value from the individual ego-velocity estimated values of the stationary targets; and ascertaining a corrected angle estimated value for the stationary targets using the combined ego-velocity estimated value and the respective measured relative velocity.

2. The method as recited in claim 1, wherein the predefinable range is an error tolerance range, which is ascertained from an error for the measurement of the relative velocity and from an error for the angle estimation.

3. The method as recited in claim 1, wherein the angle estimation for each respective target takes place using multiple receiving and/or transmission channels at different positions.

4. The method as recited in claim 1, wherein an averaged velocity value for the stationary targets is determined as the combined ego-velocity estimated value by weighted and or unweighted averaging.

5. The method as recited in claim 1, wherein for each moving target, the respective angle estimated value resulting from the angle estimation is adopted as the angle estimated value for the moving target.

6. The method as recited in claim 1, wherein a velocity estimated value for each moving target is ascertained from the respective relative velocity of the moving target, which is measured using the Doppler effect.

7. The method as recited in claim 1, wherein an elevation angle is taken into consideration when ascertaining the individual ego-velocity estimated value of the synthetic aperture radar sensor using the respective relative velocity and the respective angle estimated value for each respective target.

8. The method as recited in claim 1, wherein the radar sensor is a chirp sequence radar.

9. The method as recited in claim 1, wherein an ascertainment of each respective relative velocity takes place using the Doppler effect with the aid of a Keystone processing.

10. A non-transitory machine-readable memory medium on which is stored a computer program for determining an ego-velocity estimated value and an angle estimated value of targets using a synthetic aperture radar sensor, the computer program, when executed by a computer, causing the computer to perform the following steps: measuring, using the synthetic aperture radar sensor, a respective distance between the synthetic aperture radar sensor and each respective target; measuring, using the synthetic aperture radar sensor, a respective relative velocity of each respective target using the Doppler effect; performing an angle estimation of one respective angle estimated value for each respective target, which characterizes an angle between a direction of an ego-velocity of the synthetic aperture radar sensor and each respective target; ascertaining an individual ego-velocity estimated value of the synthetic aperture radar sensor using the respective relative velocity and the respective angle estimated value for each respective target; classifying and distributing the individual ego-velocity estimated values relating to stationary targets, whose individual ego-velocity estimated values are situated within a predefinable range relative to one another, and relating to moving targets, whose individual ego-velocity estimated values are situated outside the predefinable range; ascertaining a combined ego-velocity estimated value from the individual ego-velocity estimated values of the stationary targets; and ascertaining a corrected angle estimated value for the stationary targets using the combined ego-velocity estimated value and the respective measured relative velocity.

11. A synthetic aperture radar sensor, comprising: a sensor array; wherein the synthetic aperture radar sensor is configured to determine an ego-velocity estimated value and an angle estimated value of targets using the sensor array, the synthetic aperture radar sensor being configured to: measure a respective distance between the synthetic aperture radar sensor and each respective target; measure a respective relative velocity of each respective target using the Doppler effect; perform an angle estimation of one respective angle estimated value for each respective target, which characterizes an angle between a direction of an ego-velocity of the synthetic aperture radar sensor and each respective target; ascertain an individual ego-velocity estimated value of the synthetic aperture radar sensor using the respective relative velocity and the respective angle estimated value for each respective target; classify and distribute the individual ego-velocity estimated values relating to stationary targets, whose individual ego-velocity estimated values are situated within a predefinable range relative to one another, and relating to moving targets, whose individual ego-velocity estimated values are situated outside the predefinable range; ascertain a combined ego-velocity estimated value from the individual ego-velocity estimated values of the stationary targets; and ascertain a corrected angle estimated value for the stationary targets using the combined ego-velocity estimated value and the respective measured relative velocity.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Exemplary embodiments of the present invention are represented in the figures and explained in greater detail in the following description.

[0025] FIG. 1 schematically shows a representation of a traffic situation, in which various targets and associated angles and relative velocities are represented, and in which the synthetic aperture radar sensor according to the present invention is used.

[0026] FIG. 2 shows a flowchart of one exemplary embodiment of the method according to the present invention.

[0027] FIG. 3A shows a diagram of the distribution of individual ego-velocity estimated values of the radar sensor for different targets.

[0028] FIG. 3B shows a histogram for the distribution from FIG. 3A.

[0029] FIG. 4 shows a location diagram of a trajectory, which has been generated based on one specific embodiment of the method according to the present invention, and of a trajectory, which has been generated with the aid of an odometry sensor.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

[0030] FIG. 1 schematically shows a representation of a traffic situation, including a vehicle F, which includes a synthetic aperture radar sensor S according to the present invention, and multiple other vehicles, which are identified as targets Z1 through Z3. Further targets not represented here are typically present in the surroundings such as, for example, buildings, the infrastructure of the road, i.e., traffic signs, guard rail and the like, or the road itself. Vehicle F, and thus also radar sensor S, move at an ego-velocity v.sub.ego along a straight line. Starting from radar sensor S, azimuth angle α.sub.1, α.sub.2, α.sub.3 for each represented target Z1, Z2, Z3 is represented in each case between the direction of ego-velocity v.sub.ego and the direction of respective target Z1, Z2, Z3. In addition, relative velocity v.sub.rel,1, v.sub.rel,2, v.sub.rel,3 of each target Z1, Z2, Z3 is represented relative to radar sensor S. If one of the targets, for example, target Z1, is a stationary target, i.e., it does not move, associated relative velocity v.sub.rel,1 is given as a projection of ego-velocity v.sub.ego of radar sensor S at associated azimuth angle α.sub.1. The projections are shown in FIG. 1 for all three targets Z1, Z2, Z3. In the case of a moving target, for example, target Z2, which moves at an unknown velocity, the velocity of target Z2 is part of relative velocity v.sub.rel,2 and measured relative velocity v.sub.rel,2 deviates from the projection.

[0031] FIG. 2 shows a flowchart of one exemplary embodiment of the method according to the present invention. In this method, a multitude of targets, identified here generally with i, are examined. Measurements 1 are carried out while vehicle F and radar sensor S are moving. Radar sensor S is designed as a sensor array and includes multiple transmission channels and, if necessary, multiple receiving channels. Measurements 1 are carried out using a frequency-modulated continuous wave radar modulation (FMCW), in which chirp signals having rapidly rising linear frequency ramps of the same slope are output at predefined temporal intervals. The reflected signals are received and processed as received signals. A mixture of the instantaneous transmit signal with the received signal results in a low-frequency beat signal, whose frequency is proportional to the distance of target i. Measurements 1 are carried out in such a way that the Doppler effect and the Doppler shift in the beat frequency are insignificant or are taken into consideration in the evaluation.

[0032] A Keystone processing 2 is subsequently carried out. In this case, an estimation of the Doppler shift or the Doppler frequency is carried out by determining the temporal development of the phase of the complex measured signals across the frequency ramps and by compensating for the corresponding linear distance change (migration) for each estimated value. In this way, relative velocities v.sub.rel,i are ascertained for each target i. A distance estimation from the time range into the frequency range subsequently takes place with the aid of a conventional Fourier processing 3, in particular, a Fast Fourier Transform (FFT). The generated two-dimensional spectra (distance and relative velocity) of the individual transceiver channel combinations are non-coherently averaged 4. For this purpose, the amount of each individual one of these spectra is formed and these amounts or their square values are then added together. A detection at a constant false alarm rate (CFAR) 5 is carried out in order to identify the targets in the measurements.

[0033] In addition, an angle estimation 6 is carried out, in which azimuth angle estimated values {circumflex over (θ)}.sub.i for the targets are ascertained. Azimuth angle estimated value {circumflex over (θ)}.sub.i represents the azimuth angle between a measuring axis of radar sensor S and target i, and thus also reflects the installation situation of radar sensor S. Since the installation situation is known, it is possible to convert azimuth angle estimated value {circumflex over (θ)}.sub.i via coordinate transformation into an estimated value for azimuth angle α.sub.i between the direction of ego-velocity v.sub.ego and the direction of target i. For the case shown in FIG. 1, the measuring axis is perpendicular to the direction of ego-velocity v.sub.ego. This results in the following correlation: θ.sub.i=90−α.sub.i. Digital beam forming is used for angle estimation 6. In this case, measurements are carried out simultaneously over the multiple receiving and transmission channels at different positions on the sensor array and a phase difference is calculated, from which azimuth angle estimated values {circumflex over (θ)}.sub.i may then be ascertained. The influence of ego-velocity v.sub.ego of radar sensor S for this type of angle estimation is insignificant, so that azimuth angle estimated values {circumflex over (θ)}.sub.i are ascertained separately from ego-velocity v.sub.ego of radar sensor S. In angle estimation 6, an elevation angle ϕ.sub.i between the plane in which the vehicle moves and the height at which target i is detected, is also ascertained.

[0034] Thus, relative velocity {circumflex over (v)}.sub.rel,i, azimuth angle estimated value {circumflex over (θ)}.sub.i and, optionally, elevation angle ϕ.sub.i are known for each target i. Thus, an individual ego-velocity estimated value {circumflex over (v)}.sub.ego,i according to formula 1 is calculated 7 separately for each target i:

[00001]$\begin{array}{cc}{\hat{v}}_{\mathrm{ego},i}=\frac{{\hat{v}}_{\mathrm{rel},i}}{\sin {\hat{\theta }}_i.Math.\cos {\varphi }_i}& \left(\mathrm{Formula}1\right)\end{array}$

[0035] Individual ego-velocity estimated values {circumflex over (v)}.sub.ego,i calculated in this way for several targets i are shown in a diagram in FIG. 3a. FIG. 3b shows a histogram, in which the averaged number n ascertained for each of a multitude of different individual ego-velocity estimated values {circumflex over (v)}.sub.ego,i is plotted. In both figures, it is apparent that individual ego-velocity estimated values {circumflex over (v)}.sub.ego,i accumulate in a range B. In a typical traffic situation, significantly more stationary objects than moving objects having the same relative velocity radially to radar sensor S are present.

[0036] Referring to FIG. 2, a clustering 8 takes place, in which individual ego-velocity estimated values {circumflex over (v)}.sub.ego,i situated within range B, are assigned to stationary targets and individual ego-velocity estimated values {circumflex over (v)}.sub.ego,i situated outside range B are assigned to moving targets. The moving targets are thus identified (MTI—moving target indication) and separated from the stationary targets. Range B is defined via the errors in measurement 1 and in angle estimation 6 and represents an error tolerance range.

[0037] Individual ego-velocity estimated values {circumflex over (v)}.sub.ego,i assigned to the stationary targets, i.e., which are situated within range B are averaged 9, in order to obtain a combined ego-velocity estimated value {circumflex over (v)}.sub.ego,k. Various types of averaging processes may be carried out, for example, a classical averaging such as, for example, an arithmetic average, a weighted averaging, for example, with weights as a function of the signal-to-noise ratio, a determination of the maximum in the histogram, a formation of a median, etc. Since the combined ego-velocity estimated value {circumflex over (v)}.sub.ego,comb has been calculated, in principle, without the moving targets, it may be regarded as actual ego-velocity {circumflex over (v)}.sub.ego of radar sensor S. In this way, an autofocus is achieved. For each stationary target, an angle calculation 10 also takes place from relative velocity {circumflex over (v)}.sub.rel,i for the stationary target ascertained by Keystone processing 2 with the aid of the Doppler effect and from calculated individual ego-velocity estimated value {circumflex over (v)}.sub.ego,i for the stationary target with the aid of formula 2:

[00002]$\begin{array}{cc}{\hat{\theta }}_{\mathrm{corr},i}=\arcsin \left(\frac{{\hat{v}}_{\mathrm{rel},i}}{{\hat{v}}_{\mathrm{ego},i}}\right)& \left(\mathrm{Formula}2\right)\end{array}$

[0038] As a result, a corrected angle estimated value {circumflex over (θ)}.sub.corr,i is calculated, which may be regarded as the actual azimuth angle of the target with respect to radar sensor S.

[0039] For the moving targets, however, above-described angle calculation 10 would result in an erroneous angle estimation, since the velocity component of the moving target is unknown and may therefore not be taken into consideration. For a moving target, therefore, azimuth angle estimated value {circumflex over (θ)}.sub.i ascertained in angle estimation 6 is adopted 11. No improved angle estimation is achieved as a result, however, the erroneous angle estimation is avoided. Finally, a radial velocity estimated value of the moving target may be calculated 12 by subtracting the combined ego-velocity estimated value {circumflex over (v)}.sub.ego,comb ascertained for the stationary targets by averaging 9 from relative velocity {circumflex over (v)}.sub.rel,i ascertained by Keystone processing 2 with the aid of the Doppler shift.

[0040] FIG. 4 shows a comparison between an odometry trajectory T.sub.o, which has been ascertained in a conventional manner with the aid of an odometry sensor, and a trajectory T.sub.v, which has been generated based on one specific embodiment of the method according to the present invention. It is apparent that the two trajectories very closely match one another, and thus the autofocus provides precise results with the aid of the method according to the present invention.