Radar system

Abstract

A radar system with an antenna array (18) having a plurality of transmitting antennas (16) and / or receiving antennas (18) and with a control and evaluation device (26) which is configured to transmit radar signals and to receive radar echoes from radar targets and to detect the radar targets in a localization space having at least two spatial dimensions and at least two velocity dimensions, characterized in that the control and evaluation device (26) is configured to calculate, for a selected point in the localization space, for a plurality of combinations of transmitting antennas and receiving antennas, each expected measurement result on the assumption that there is a radar target at the point on the basis of a model and to determine localization data of a detected target by comparing the calculated measurement results with measurement results derived from actually received signals.

Description

Technical Field

[0001] The present invention relates to a radar system having an antenna array and a control and evaluation device, the antenna array having a plurality of transmitting antennas and / or a plurality of receiving antennas, the control and evaluation device being configured to transmit radar signals and receive radar echoes from a radar target, and to detect the radar target in a positioning space having at least two spatial dimensions and at least two velocity dimensions based on the received signals.

[0002] In particular, this invention relates to a radar system for environmental monitoring in a motor vehicle. The radar system may have one or more radar sensors. The antenna array consists of all the transmitting and receiving antennas of all the radar sensors belonging to the system. Background Technology

[0003] In motor vehicles, radar sensors are used to enable comfort features such as adaptive cruise control and safety features such as emergency braking assist. The main advantage of these sensors is that they directly measure physical quantities, rather than interpreting, for example, camera images. Radar sensors transmit high-frequency radar beams via an antenna array and receive the beams reflected from objects. The detected object can be stationary or moving. Using the received radar beams, the distance and direction (angle) to the object can be calculated. Furthermore, the velocity of the object relative to the radar sensor can be calculated. Radar sensors typically operate in the frequency range of 76 to 81 GHz.

[0004] One known method for radar modulation is the so-called chirp-sequence method. Here, many rapid frequency ramps—the so-called chirps—are transmitted. This increases the range resolution (ΔΔ). d (Depends on the bandwidth used) B The resolution in terms of relative velocity (Δ v (Depends on the measurement duration) T : .

[0005] Here, c represents the speed of light. f 0 represents the center frequency of the frequency ramp.

[0006] Examples of this type of radar system are described in DE 10 2018 202 289 A1, DE 10 2018 202 290 A1 and DE 10 2018 202 293 A1.

[0007] By using a collaborative radar sensor network, the performance of environmental detection can be significantly improved. Here, joint evaluation of data from multiple radar sensors achieves both increased sensitivity and improved environmental detection accuracy because a larger antenna array and more high-frequency (HF) channels are available compared to a single sensor. In particular, the antenna array can also include dummy antennas, which are generated by receiving signals from spatially offset transmitting antennas in time-division, frequency-division, or code-division multiplexing. However, with very large dummy antenna arrays and long simultaneous bandwidths and measurement durations, individual antennas measure targets at extremely different distances, relative velocities, and angles (especially in the near field), making joint evaluation of the entire dummy antenna array difficult.

[0008] In cooperative radar sensor networks, a distinction is made between systems with coherent high-frequency oscillator signal allocation and those without. Systems with high-frequency oscillator signal allocation exhibit significantly better phase noise characteristics; however, oscillator signal allocation comes with high costs. Therefore, depending on the application, solutions without oscillator signal allocation may be advantageous for practical implementation. Summary of the Invention

[0009] The objective of this invention is to realize a radar system with improved performance.

[0010] According to the present invention, the task is accomplished by the control and evaluation device being configured to: for a selected point in the positioning space, assuming the presence of a radar target at that point, calculate each expected measurement result based on a model for multiple combinations of transmitting and receiving antennas, and determine the positioning data of the detected target by comparing the calculated measurement results with measurement results derived from the actually received signals.

[0011] This invention offers the following advantages: by using a common model to compensate for the effects on the received signal caused by the different range, relative velocity, and angular orientations of different transmitting and receiving antenna or array pairs (more precisely, their phase centers) relative to the radar target, the positioning data (the radar target's position coordinates and relative velocity vector) can be given directly in a coordinate system fixed relative to the antenna array. By dividing the positioning space into a (not necessarily uniform) grid of cells, the model can be used to predict, for each set of positioning data describing a given target, how the signals received by different virtual antennas, or the measurements (range, relative velocity, angle) derived from them, can be distinguished from each other, utilizing the resolution determined by the grid's fineness. Then, for each assumption about the located radar target, the actual received signal can be checked to see if it matches the predicted value. In this way, the reliability of object positioning can be significantly improved. This type of signal evaluation can be combined with traditional evaluation methods in various ways, thus achieving a reasonable trade-off between computational complexity and measurement accuracy for each task.

[0012] Advantageous configurations of the invention are given in the dependent claims.

[0013] In one implementation, the expected received signal is calculated for each cell in the positioning space, and the correlation between the measured received signal and the calculated received signal is evaluated. For example, the values ​​calculated for the expected received signal can form the parameters of a matched filter (correlation filter), which is then applied to the actually received signal to improve the signal-to-noise ratio. Alternatively, the necessary computational load can be reduced, for example, by using a coarser cell grid in less correlated regions of the positioning space, and / or by limiting the method to a low-dimensional subspace of up to 6-dimensional positioning space.

[0014] In another implementation, the antenna array is divided into multiple subarrays (where each subarray may be, for example, an antenna array of a single radar sensor), and the signals received in each subarray are first subjected to conventional FFT processing to determine preliminary values ​​of positioning data (range, radial velocity, angle). Using the results obtained in this way, differences (e.g., phase differences) between the signals received in different subarrays can then be compensated, followed by coherent or incoherent summation of the compensated signals.

[0015] The spectra obtained from different subarrays exhibit frequency shifts. At points corresponding to the center of the positioning spatial cell, this shift is caused by differences between expected measurement parameters for distance, relative velocity, and angle. Therefore, the frequency shift can be compensated for at these points. At the remaining points of the spectrum, at least approximate compensation can be achieved through interpolation between grid points. The compensated signal can then be coherently or incoherently summed again.

[0016] In another implementation, conventional FFT processing is performed on each subarray, followed by conventional target detection based on the resulting multidimensional spectrum, employing a reduced detection threshold if necessary. For different subarrays, slightly offset localization data is thus obtained, where the offset corresponds to predictions provided by the model. The model therefore provides correction factors that achieve a fine estimate of the coordinates in the localization space. Attached Figure Description

[0017] The embodiments will now be explained in more detail with reference to the accompanying drawings. The drawings show: Figure 1 A block diagram of the radar system according to the present invention; Figure 2 A sketch used to illustrate the model used in a radar system; Figures 3 to 5 Charts used to display the predicted values ​​derived from the model; and Figures 6 to 8 Flowcharts of different evaluation methods that can be implemented in the radar system according to the present invention. Detailed Implementation

[0018] exist Figure 1 The image schematically illustrates a radar system with three radar sensors 10, 12, and 14, which can be installed at different locations within a motor vehicle.

[0019] Each radar sensor has at least one transmitting antenna 16 and at least one receiving antenna 18. In practice, the number of transmitting and receiving antennas for each radar sensor can be significantly greater than the number shown in the example here. The transmitting and receiving antennas 16 and 18 of all radar sensors together form an antenna array 18, in which the antennas of each individual radar sensor 10, 12, and 14 form subarrays 20, 22, and 24, respectively.

[0020] The central control and evaluation unit 26 controls the functions of the three radar sensors 10, 12, and 14 and evaluates the signals received by these radar sensors to determine the location data of the located radar target, namely the position coordinates x, y and velocity component v. x v y .

[0021] Due to central control and data evaluation, radar sensors 10, 12, and 14 constitute a cooperative radar system in which results obtained from one radar sensor can be compared with results from other radar sensors to obtain more accurate and reliable positioning data.

[0022] In the example described here, radar sensors 10, 12, and 14 are FMCW radar sensors, in which the frequency of the signal transmitted by transmitting antenna 16 is ramp-modulated. By mixing the signal received by receiving antenna 18 with a portion of the simultaneously transmitted signal, a beat frequency signal (baseband signal) is obtained, the frequency of which depends on the propagation time of the radar signal from the sensor to the radar target and back. A spectrum is formed from this signal using a Fast Fourier Transform (FFT), in which each located target is represented by a peak at a specific frequency. In this way, distance information about the located objects is obtained.

[0023] Due to the Doppler effect, the frequency of the baseband signal also depends on the relative motion of the corresponding object. However, when the frequency ramp is sufficiently steep, this Doppler frequency shift within the ramp is negligible. However, if multiple ramps (chirps) are transmitted and received consecutively, a measurable Doppler phase shift occurs between the ramps. Therefore, by performing a Fourier transform on the ramps, the radial relative velocity of each located target can also be determined in the two-dimensional (range-Doppler) spectrum.

[0024] The angular information of the located target is obtained by comparing the amplitude and / or phase of signals received by receiving antennas 18 that are spatially offset from each other, or transmitted by transmitting antennas 16 that are spatially offset from each other. In the example shown, the receiving antennas 18 of each sensor are offset from each other only in the horizontal direction, so that only the azimuth of the target can be determined. In this case, the measurable position coordinates and velocity components of the radar target constitute a four-dimensional positioning space with two spatial dimensions (horizontal plane) and two velocity dimensions. If the antennas of a subarray are also offset from each other in the vertical direction, the elevation angle of the target can also be determined, and the positioning space has three spatial dimensions and three velocity dimensions. For simplicity, only the four-dimensional positioning space is considered in this specification.

[0025] If a subarray (e.g., subarrays 22 and 24) has multiple transmit antennas 16, the difference between signals received by the receive antenna depends not only on the positioning angle and the spatial distance between the receive antennas, but also on the spatial distance between the receive antenna and the transmit antenna 16 that transmitted the signal. The larger the aperture, i.e., the spatial extension of the array in the horizontal direction, the higher the azimuth resolution of the radar sensor. If only one transmit antenna is used for transmission, the aperture is determined solely by the receive antenna 18. In a MIMO (Multiple-Input Multiple-Output) radar with multiple transmit antennas 16, time-division, frequency-division, or code-division multiplexing can ensure that it is possible to determine which transmit antenna transmitted each received signal. In this case, the aperture of the antenna array is virtually increased by the multiple transmit antennas 16, because switching from one transmit antenna to another is equivalent to adding an additional (virtual) receive antenna. Therefore, by applying the MIMO principle, angular resolution can be significantly improved.

[0026] In the example shown, the bistatic antenna concept is implemented in radar sensors 10, 12, and 14, meaning that the transmitting antenna 16 and the receiving antenna 18 are different. However, the radar system may also include radar sensors with a monostatic antenna concept, wherein at least a portion of the antenna can function as both a transmitting and receiving antenna.

[0027] In the example shown, radar sensor 10 has a local oscillator 28 that provides the signal transmitted by transmitting antenna 16. Another oscillator 30 generates a high-frequency signal, which is coherently distributed to the other two radar sensors 12 and 14, so that the transmitting antennas 16 of these radar sensors transmit coherent signals, i.e., signals with a fixed and known phase relationship. This simplifies the evaluation of signals, such as those transmitted by one transmitting antenna 16 of radar sensor 12 and received by one receiving antenna 18 of radar sensor 14. By evaluating this cross-sensor signal, the effective aperture of antenna array 18 is significantly increased. Correspondingly, signals transmitted by one transmitting antenna 16 of radar sensor 12 and received by one receiving antenna 18 of radar sensor 10 can also be evaluated. However, in order to down-convert the received signal to baseband and then use the signal from local oscillator 28, which is incoherent with the transmitted signal provided by oscillator 30, significant phase noise may appear in the signals transmitted by radar sensors 12 or 14.

[0028] Therefore, conventional assessments of radar signals described to date primarily provide only a range, a radial velocity, and a positioning angle for each radar target. In the control and assessment device 26, these data are converted to a Cartesian coordinate system fixed relative to the antenna array 18, which unfolds to have dimensions x, y, v.x v y The positioning space is limited. However, a difficulty arises because subarrays 20, 22, and 24 are relatively far apart from each other, and therefore the distances, radial velocities, and positioning angles measured by different radar sensors are different from each other. This makes the evaluation of cross-sensor signals particularly difficult, as the evaluation depends on the significantly different distances, radial velocities, and positioning angles from the radar target to the different sensors.

[0029] To better manage these complexities, the control and evaluation device 26 is configured to model the differences between measurement parameters obtained from different radar sensors 10, 12, and 14 at different points in the positioning space. The following will be based on... Figure 2 Explain the model used for this purpose.

[0030] exist Figure 2 In the Cartesian coordinate system with x and y axes, the positions of radar target 32 ​​and the two antennas of antenna array 18, P0 and P1, are marked. i For simplicity, assume the antenna is as follows: Figure 1 The antennas are arranged on a horizontal straight line, forming the y-axis, so that their positions can be consecutively numbered using index i. It is further assumed that the optical axes of the radar sensors, and therefore the directions of the main lobes of the antennas, are parallel to each other and parallel to the x-axis. These assumptions do not always need to be satisfied in practice. However, generalizing to cases where the antenna positions are also offset from each other in the x-direction and / or the antennas have different orientations relative to the x-axis does not present difficulties. The origin of the coordinate system is set to coincide with the antenna position P0. This antenna position therefore has coordinates (0, y0) = (0, 0). Antenna position P i It has coordinates (0, y) i Furthermore, radar target 32 ​​has coordinates (x, y). It is understandable that... Figure 2 The xy plane shown is a subspace of the overall four-dimensional positioning space. The two velocity dimensions are not shown here.

[0031] At positions P0 and P i The antenna at point P0 can be either a receiving antenna 18 or a transmitting antenna 16. It should be initially assumed that they are monolithic transmitting and receiving antennas. That is, the signal transmitted by the antenna at P0 is reflected at radar target 32 ​​and re-received by the same antenna. The distance d0 is measured here. Accordingly, using the antenna located at P... i The antenna at point 32 measures the distance d to radar target 32. i .

[0032] Using the antenna located at P0 and adjacent antennas belonging to the same subarray (not shown), the angle θ0 of the radar target 32 ​​observed from position P0 can be determined. Correspondingly, the angle θ0 of the radar target 32 ​​observed from position P0 can also be determined. i The same radar target was observed at an angle θ of 32. i .

[0033] For any antenna position P i d measured at the location i and θ i The value of is then determined by the following relation: .

[0034] If v x and v y If the velocity component (not shown) of radar target 32 ​​relative to antenna array 18 is given, then at antenna position P... i radial velocity v relative to the radar target r,i Measured values: Furthermore, for the tangential velocity v, which cannot be directly measured... t,i Applicable to: .

[0035] Now, if we assume that radar target 32 ​​is a stationary target, then it has a component v x and v y The relative velocity vector is equal in magnitude and opposite in direction to the velocity vector of the vehicle equipped with the radar system. Based on the model given by equations (1) to (4), it is in principle possible to calculate for each point in the four-dimensional positioning space: at antenna position P i How much do the distance, radial velocity, and angle values ​​measured at point P0 differ from those measured at antenna position P0?

[0036] Figure 3 This is a graph of relevant segments in the xy-plane, showing the difference in distance *d* calculated based on the model for each point in the plane. Here, the distance difference, represented by different shades, is given in bins, where the bin size corresponds to the spatial range resolution, which is inversely proportional to the bandwidth *B* of the frequency-modulated radar signal. Different units are used for the x and y axes. Units in the x-direction are given in 100 m, while units in the y-direction are given in meters.

[0037] To compute this graph, the location space ( Figure 3The model is divided into units 34 (xy subspace), and the x and y coordinates of the center point of each unit 34 are substituted into equations (1) to (4) of the model. Figure 3 This is shown in a very rough manner. It can be seen that the velocity difference increases with the distance from the line y=0.

[0038] Figure 4 A corresponding plot of the calculated difference in radial velocity vr is shown, again presented in segments. Here, the segment size corresponds to the velocity resolution, which is inversely proportional to the measurement duration T.

[0039] at last, Figure 5 A corresponding graph showing the calculated differences for the azimuth angle θ is shown. Different shades here indicate different angular deviations in degrees.

[0040] Figures 3 to 5 The graph shown is based on the following parameters: Bandwidth B: 3 GHz, corresponding to a distance segment size of 0.05 m. Measurement duration T: 20 ms, corresponding to a velocity range of 0.1 m / s. Offset between the two antennas: 1 m Antenna angular orientation relative to the driving axis: 0° Vehicle speed relative to the ground: 20 m / s.

[0041] As needed, the corresponding graph can be calculated for any pair of antenna positions.

[0042] In MIMO systems, or typically in bistatic systems, the position of the transmitting antenna must also be considered when comparing measurements taken at two locations of the receiving antenna. The distance measured at the receiving antenna location, for example d... i , is the arithmetic mean of the distance from the transmitting antenna to the target and the distance from the target to the receiving antenna, where these two distances are calculated according to equation (1). Similarly, for the receiving antenna position P i The radial velocity vr,i at the location is obtained by substituting the azimuth angle at the location of the transmitting antenna and the azimuth angle at the location of the receiving antenna into equation (3) once, and then averaging the obtained velocity values.

[0043] However, when estimating the angle, the launch angle θ must be considered separately. tx and the receiving angle θ rx If function a tx,i (θ) is the complex unidirectional antenna characteristic of transmitting antenna i, and the function a rx,jLet (θ) be the complex unidirectional antenna characteristic of the receiving antenna j. Then, according to the following formula, for a receiving antenna j with a transmission angle θ... tx The transmitting antenna i to the receiving angle θ rx The path of the receiving antenna j is used to obtain the complex amplitude of the received signal: .

[0044] The required unidirectional antenna characteristics can be determined using known methods for calibrating antenna patterns.

[0045] In environmental detection within driver assistance systems for motor vehicles, not all areas of the positioning region exhibit the same correlation, resulting in varying accuracy requirements. Therefore, to reduce computational load, a coarser cell grid can be used in specific regions of the positioning space, requiring only the calculation of measurement differences for a smaller number of grid points. Similarly, it is possible to restrict the method to a subspace of the positioning space, i.e., a space with lower dimensions. For example, in some cases, performing the method only on the xy-plane but not on the velocity space may be sufficient.

[0046] In some cases, it may be sufficient to combine multiple transmit antennas and / or multiple receive antennas into a subarray, so that calculations only need to be performed on the antenna positions corresponding to the phase center of the subarray.

[0047] The main steps of the evaluation method that can be implemented in the control and evaluation device 26 are as follows: Figure 6 The flowchart is presented below. In step S1, a set of received signals is calculated for each cell 34 in the positioning space in the manner described above. This set of received signals is expected when a radar target is present at the center point of the relevant cell. The expected measurement result BINex (Zelle) can here be represented by segments in the range, radial velocity, and angular dimensions, or by the segment offsets between the expected signals at different antenna positions. These expected segment offsets are used in step S2 to create a matched filter, by which the signal a actually measured at the relevant antenna position is filtered to improve the signal-to-noise ratio.

[0048] In optional step S3, the number of units considered and thus the amount of data can be reduced, for example, by mathematical operations such as finding the maximum or average value. For example, for each position unit in the xy plane, there are "pillars" of units in both velocity dimensions. However, many of these velocity units can typically be discarded, leaving only a few velocity units with the strongest power in the position units. Similarly, it is conceivable to perform the evaluation according to the method described herein only on stationary targets, thereby reducing the dimensionality of the positioning space to two spatial dimensions.

[0049] Then, in step S4, object detection is performed by determining the occupancy probability of each remaining cell in the location space. In the simplest case, a binary occupancy map is created by threshold comparison, where each cell either contains a value of one (object present) or a value of zero (empty). In an optional step S5, the occupancy map is predicted based on the detected velocity until the next measurement time point, and / or the occupancy map is aggregated over time.

[0050] Figure 5 This illustrates another conceivable evaluation method requiring less computation. In step S10, for each of the multiple subarrays, for example... Figure 1 Subarrays 20, 22, and 24 in the model are processed using 3D-FFT in a classical manner in the distance, Doppler, and angular dimensions. Based on the model given by equations (1) to (4), the expected differences between the 3D spectra calculated in step 10 are then calculated in step S11. In the subsequent S12, the segment offsets obtained from the model are then used to adjust the spectrum to eliminate the differences between the measurement parameters, after which the spectrum is incoherently summed.

[0051] Alternatively, in step 13, for each cell, the expected measurement parameter d can be used. j θ j v r,j Interpolate the three 3D spectra, then perform coherent or incoherent summation on the spectrum. That is, for the dimension d of the spectrum... j θ j v r,j For each segment location, check whether the relevant point in the positioning space is located on a grid point, i.e., whether it is the center point of the cell. If the point is located between two grid points, interpolate between the deviations of the measurement parameters calculated for these two grid points to correct the spectrum based on the deviations.

[0052] Then, following step S12 or S13 are optional reduction step S14, detection step S15, and optional prediction step S16, which correspond to Figure 6 Steps S3 to S5 in the process.

[0053] If only incoherent summation of the spectrum is performed in step S12 or S13, the evaluation may also include radar sensors that do not acquire the assigned high-frequency oscillator signal, but rather... Figure 1 The radar sensor 10 generates a local oscillator signal, similar to that in radar. However, due to increased phase noise, it may be meaningful to combine only those signals obtained from a single-base phase.

[0054] Another variation of the evaluation method is in Figure 8 Presented in the middle. Similar to Figure 7 In step S10, in Figure 8 In step S20, the signal 'a' obtained from each subarray undergoes 3D-FFT processing. This is followed by a detection step S21, in which object detection is performed classically for each spectrum obtained in step S20, employing a reduced detection threshold where necessary so that even weaker signals can be evaluated. Then, in the subsequent correction step S22, the model given by equations (1) to (4) is used to: aggregate the complex spectral data of the subarrays for each detected target, and to perform a fine estimation of the radar target's coordinates in the positioning space using the correction data from this model. Then, in step S23, the fine estimate of the detected target is output, and / or in step S24, it is used to update the occupancy map.

Claims

1. A radar system having an antenna array (18) and a control and evaluation device (26), the antenna array having a plurality of transmitting antennas (16) and / or receiving antennas (18), the control and evaluation device being configured to transmit radar signals and receive radar echoes from a radar target (32) and detect the radar target in a positioning space based on the received signals, the positioning space having at least two spatial dimensions and at least two velocity dimensions, characterized in that, The control and evaluation device (26) is configured to: for a selected point in the positioning space, assuming the presence of a radar target at the point, calculate each expected measurement result based on a model for multiple combinations of transmitting and receiving antennas, and determine the positioning data of the detected target by comparing the calculated measurement results with measurement results derived from the actually received signal.

2. The radar system according to claim 1, wherein, The positioning space or at least one of its subspaces is divided into a grid of cells, and the selected point is the center point of the cell.

3. The radar system according to claim 2, wherein, The size of the unit varies for different regions within the positioning space.

4. The radar system according to claim 2 or 3, wherein, To compare the measurement results, the correlation between the actual received signal and the signal derived from the model was studied.

5. The radar system according to claim 4, wherein, The received signal is evaluated using a matched filter method.

6. The radar system according to any one of claims 3 to 5, wherein, Calculate the occupancy probability for each unit.

7. The method according to any one of the preceding claims, wherein, For the multiple subarrays (20, 22, 24) of the antenna array (18), at least three-dimensional spectra with distance, relative velocity and angle dimensions are calculated respectively. The spectral values ​​in the spectrum of each subarray are corrected point by point based on the model, and the corrected spectra are summed.

8. The radar system according to claim 7, wherein, The expected measurement results are calculated for each grid point in the positioning space, and interpolation is performed between the grid points when correcting the spectrum.

9. The radar system according to claim 7 or 8, wherein, Perform coherent summation on the spectrum.

10. The radar system according to claim 1 or 2, wherein, First, target detection is performed in the spectrum of the received signal. Then, groups of associated targets are identified based on the model, and the spectrum data is fused.

11. The radar system according to any one of the preceding claims, having a plurality of radar sensors (10, 12, 14) constituting a cooperative radar system.

12. The radar system according to claim 11, wherein, The high-frequency signal of the oscillator (30) is coherently distributed to multiple radar sensors (12, 14).

Images