Radar system and method for operating a radar system

Abstract

This invention relates to a radar system for a motor vehicle, comprising a transmitting and receiving device and an analysis and processing device. The transmitting and receiving device has a true aperture and is configured to emit radar radiation, receive reflected radar radiation, and generate a radar measurement signal. The analysis and processing device is configured to, using the generated radar measurement signal, calculate at least one first angle estimate for an object reflecting radar radiation based on the true aperture. Furthermore, the analysis and processing device is configured to, using motion information about the movement of the radar system and the generated radar measurement signal, calculate at least one second angle estimate for the object using a synthetic aperture radar algorithm. Finally, the analysis and processing device is configured to fuse the at least one first angle estimate with the at least one second angle estimate. The invention also relates to a method for operating the radar system.

Description

Technical Field

[0001] The present invention relates to a radar system for a motor vehicle and a method for operating the radar system for a motor vehicle. Background Technology

[0002] In the field of motor vehicles, radar systems used to measure the distance, relative speed, and angle of objects (e.g., vehicles and obstacles) are employed for safety and comfort functions. An exemplary radar device is known in this regard from DE102015218538A1.

[0003] Therefore, the application of synthetic aperture radar (SAR) in the automotive field is studied in particular. A "SAR sensor" refers to a radar sensor that obtains angular information through Doppler measurements. SAR systems are described in Harrer et al.'s "Synthetic aperture radar algorithm for a global amplitude map" (14th Workshop on Positioning, Navigation and Communication (WPNC), Bremen, 2017, pp. 1-6). Other applications are known from Gisder et al.'s "Application of AStream-Based SAR-Backprojection Approach For Automotive Environment Perception" (19th International Radar Conference, 2018).

[0004] The synthetic aperture principle allows for exceptionally accurate angle measurements during the self-motion of radar sensors, where radar measurements at different local locations are used as a synthetic antenna aperture or antenna surface. Synthetic aperture is achieved by placing the transmitting and receiving antennas at different local locations at each measurement moment due to the radar's self-motion, and thus it can be computationally treated as if a large antenna aperture existed along the driving trajectory. Therefore, angle measurement resolution that is impossible with existing real antenna apertures can be achieved using only a single transmitting and receiving antenna. This is especially true because a large synthetic aperture can be achieved through the radar's self-motion, an aperture that would be impractical or impossible with real antenna apertures due to the large number of antenna elements required and the limited structural space in a vehicle.

[0005] With a fixed measurement duration, the size of the synthetic aperture depends on the speed of motion. Therefore, significantly improved angular resolution can be achieved during rapid movement, while only low angular resolution is achievable during very slow movement. In contrast, in the case of a real antenna array, the angular resolution is fixedly predetermined by the array geometry.

[0006] To analyze and process the measured radar signals as synthetic aperture radar (SAR), it is typically assumed that the radar environment is stationary. Additionally, the self-motion of the radar sensor should be known to determine the location where individual measurements are taken. For this purpose, the radar's own trajectory is used as input to the SAR analysis and processing algorithm and represents the basis for calculating the SAR image.

[0007] Based on the analysis and processing algorithm, self-velocity estimation is sufficient for radar image calculations, instead of relying on more accurate radar trajectories. Here, the trajectory is assumed to be linear, and mapping to more complex trajectories is not possible.

[0008] The synthetic aperture principle is independent of the modulation method used. Nowadays, typical transmission frequencies are at 24 GHz or 77 GHz, with a maximum available bandwidth of less than about 4 GHz, but usually significantly lower, for example, 0.5 GHz.

[0009] Radar systems in the automotive field typically use FMCW modulation (frequency modulated continuous wave) with a fast ramp (fast linear frequency modulation), where multiple linear frequency ramps of the same slope pass successively. The instantaneous mixing of the transmitted and received signals produces a low-frequency signal with a so-called beat frequency, which is proportional to the range. Radar systems are usually designed so that the beat frequency component caused by the Doppler frequency becomes negligible. The obtained range information is largely unique and definite, and the Doppler frequency shift can then be determined by observing the phase evolution of the complex range signal over the ramps. Range determination and velocity determination are typically performed independently using a two-dimensional Fourier transform.

[0010] In SAR analysis, the same fast linear frequency modulation (FLFM) measurement principle can be used. Range analysis is largely the same. Doppler analysis across slopes is replaced by SAR analysis. In the results, this does not provide Doppler measurements, but rather angle measurements under the assumption of a stationary target and under the knowledge of its own motion.

[0011] For SAR analysis, two categories of algorithms can be distinguished. First, there are algorithms that can handle arbitrary synthetic apertures, such as back projections, at the cost of greater computational overhead. Second, there are algorithms that are limited to certain aperture types (e.g., linear apertures), but are more computationally efficient for this purpose. An example is the Keystone algorithm, as described in Perry et al.'s "Coherent Integration With Range Migration Using Keystone Formatting" (IEEE Radar Conference 2007).

[0012] In the field of motor vehicles, efficient computation of SAR images is crucial because real-time processing is required.

[0013] The maximum effective aperture of a real antenna array that can be used for angle estimation using digital beamforming (DBF) is realized in the direction perpendicular to the aperture plane. The synthetic aperture is constructed along the driving direction in SAR analysis processing. Therefore, the maximum angular resolution is realized perpendicular to the driving direction (boresight). Typically, the effective aperture decreases from the maximum angular resolution, as a cosine function, from the boresight to the broadside (in the driving direction). In automotive applications, SAR analysis processing is therefore particularly suitable for stationary targets on the side of the driving trajectory, as it provides high angular resolution for these stationary targets.

[0014] Conversely, SAR analysis and processing are not suitable for moving targets or targets in the direction of travel. If such targets are to be detected and processed in an application, alternative or additional measurement methods are required. For this purpose, conventional radar systems can typically be used. Angle estimation is then performed using the true aperture. Summary of the Invention

[0015] The present invention provides a radar system for a motor vehicle and a method for operating the radar system for a motor vehicle.

[0016] Preferred embodiments are described below.

[0017] According to a first aspect, the present invention therefore relates to a radar system for a motor vehicle, the radar system having a transmitting and receiving device and an analysis and processing device. The transmitting and receiving device has a true aperture and is configured to emit radar radiation, receive reflected radar radiation, and generate radar measurement signals. The analysis and processing device is configured to, using the generated radar measurement signals, calculate at least one first angle estimate for an object reflecting radar radiation, based on the true aperture. Furthermore, the analysis and processing device is configured to, using motion information about the movement of the radar system and using the generated radar measurement signals, calculate at least one second angle estimate for the object using a synthetic aperture radar (SAR) algorithm. Finally, the analysis and processing device is configured to fuse the at least one first angle estimate with the at least one second angle estimate.

[0018] According to a second aspect, the present invention relates to a method for a radar system for operating a motor vehicle. Radar radiation is emitted through a transmitting and receiving device of the radar system having a true aperture. The transmitting and receiving device receives the reflected radar radiation and generates a radar measurement signal. At least one first angle estimate is calculated for an object at which radar radiation is reflected. The generated radar measurement signal and the true aperture are taken into consideration. Furthermore, using a synthetic aperture radar (SAR) algorithm, at least one second angle estimate is calculated for the object, taking into account the generated radar measurement signal and motion information regarding the movement of the radar system. The at least one first angle estimate is fused with the at least one second angle estimate.

[0019] Advantages of the present invention

[0020] The radar system includes a transmitter-receiver unit (i.e., a radar sensor) that simultaneously possesses a real aperture for conventional angle estimation (e.g., based on digital beamforming, DBF) and a synthetic aperture for SAR-based angle estimation. Conventional angle estimation has the maximum resolution perpendicular to the real aperture, while in the case of SAR analysis, the maximum angular resolution is perpendicular to the direction of motion. These two directions may differ significantly from each other depending on the direction of motion and the sensor's mounting angle. This allows for a large aperture across the entire line-of-sight and thus good angular resolution. The one-dimensional or two-dimensional target angle of a stationary target estimated based on the real (especially multiple-input multiple-output, MIMO) aperture is improved by fusing this target angle with a second angle estimate obtained through SAR angle estimation to form a new, more accurate angle estimate.

[0021] Depending on the angle, the real aperture and the synthetic aperture have different large effective components. Since the effective aperture determines the accuracy of the angle estimate under otherwise identical measurement parameters (e.g., signal-to-noise ratio), the two angle estimates can be fused in an angle-dependent manner, taking into account the size of the respective aperture. Therefore, the accuracy of the angle estimate is improved not only compared to a dedicated SAR sensor but also to a dedicated sensor with a real antenna array—provided the vehicle is moving at least slowly.

[0022] When the vehicle is stationary, the method according to the invention is implicitly simplified to a conventional angle estimation based on a real antenna array, with the resolution and accuracy achieved therein.

[0023] Therefore, according to the present invention, angle estimates are fused based on both the real aperture and the synthetic aperture. These angle estimates can be combined in such a way that improved angular resolution and accuracy can be achieved across the entire angular range and for each speed of motion. Optimal angular resolution can be achieved across the entire line-of-sight by designing the radar system and combining the two apertures (i.e., the real aperture and the synthetic aperture). A fixed angular resolution can also be maintained during stationary operation (i.e., when the radar system's speed is negligible), while during high-speed operation, the angular resolution is significantly improved by using a larger synthetic aperture.

[0024] This invention combines the advantages of sensors that obtain angular resolution from Doppler measurements based on SAR principles with the advantages of sensors that perform angle estimation based on real antenna arrays. This eliminates the need for separate sensors for each angle estimation method. It also effectively avoids the dual-mode combination of switching between the two measurement methods, thereby saving costs and reducing the complexity of the sensor system.

[0025] Another advantage of the radar system according to the invention is the multi-valued resolution of SAR angle estimation. In itself, SAR angle estimates are not uniquely defined with respect to which side the target is on. Here, the multivaluedness is mirror-symmetric with respect to the axis given by the direction of travel. This multivaluedness can be resolved by fusing the multivalued angle estimated using SAR algorithms with the uniquely defined estimated angle of the true (especially MIMO) aperture.

[0026] According to another embodiment of a radar system for motor vehicles, the analysis processing device is configured to weight at least one first angle estimate and at least one second angle estimate during fusion.

[0027] According to another embodiment of a radar system for motor vehicles, the analysis and processing apparatus is configured to, during fusion, weight the first angle estimate based on a first angle estimate and / or a second angle estimate. Therefore, the weighting depends not only on external parameters, such as the orientation of the radar system, but also on the corresponding value of the angle itself. The weighting can particularly result in, within a defined angle range, either only the first angle estimate or only the second angle estimate being further passed for analysis and processing.

[0028] According to another embodiment of a radar system for motor vehicles, the analysis and processing apparatus is configured to weight a first angle estimate based on the actual aperture, and to weight a second angle estimate based on the synthetic aperture of the radar system. The weighting can also be performed in an angle-dependent manner.

[0029] According to another embodiment of a radar system for motor vehicles, the analysis processing device is configured to fuse at least one first angle estimate with at least one second angle estimate based on an effective aperture related to the true aperture and an effective aperture related to the synthetic aperture of the radar system.

[0030] According to another embodiment of a radar system for motor vehicles, at least one first angle estimate and at least one second angle estimate respectively include azimuth and elevation angle.

[0031] According to another embodiment of a radar system for motor vehicles, the analysis and processing apparatus is further configured to perform moving target identification using the generated radar measurement signals, in order to identify whether the object is stationary. The analysis and processing apparatus is configured to calculate at least one second angle estimate for a stationary object only.

[0032] According to another embodiment of a radar system for motor vehicles, the analysis processing device is configured to obtain a first angle estimate from the radar aperture (i.e., from the phase curves on the transmit and receive channels). For example, a digital beamforming (DBF) algorithm can be used for this purpose.

[0033] According to another embodiment, the radar system for a motor vehicle also includes an interface configured to receive motion information about the movement of the radar system from external sensors. Thus, motion information is received from additional sensors (e.g., an odometer). Alternatively, motion information can be estimated using an algorithm itself, employing an autofocus method.

[0034] According to another embodiment, the radar system is a side-looking radar system. That is, the radar system is oriented perpendicular to the direction of travel. Angle fusion can improve the performance of the side-looking radar system. Specifically, in this case, the curves for the effective true aperture and the effective synthetic aperture have similar angular correlations. However, improvements can be derived based on velocity. At slow speeds, the true aperture guarantees better estimation accuracy because it is larger than the synthetic aperture. At high speeds, the opposite is true: better estimation accuracy is achieved through the synthetic aperture compared to through the true aperture.

[0035] According to another embodiment of a radar system for motor vehicles, the analysis and processing device is configured to consider not only the azimuth angle but also the elevation angle when fusing at least one first angle estimate with at least one second angle estimate.

[0036] According to another embodiment of the method for operating a radar system for motor vehicles, at least one first angle estimate is fused with at least one second angle estimate based on the effective aperture related to the true aperture and the effective aperture related to the synthetic aperture of the radar system. Attached Figure Description

[0037] The attached diagram shows:

[0038] Figure 1 A schematic block diagram of a radar system for a motor vehicle according to an embodiment of the present invention is shown.

[0039] Figure 2 The aperture length is shown as a function of the self-velocity;

[0040] Figure 3 A schematic top view of a motor vehicle having a radar system according to an embodiment of the present invention is shown;

[0041] Figure 4 The effective aperture of an azimuth-dependent angle radar system according to one embodiment of the present invention is shown.

[0042] Figure 5 The effective aperture, depending on the azimuth angle, of a radar system oriented in the driving direction according to an embodiment of the present invention is shown.

[0043] Figure 6 The effective aperture of a side-radar system according to an embodiment of the present invention is shown, depending on the azimuth angle; and

[0044] Figure 7 A flowchart illustrating a method for operating a radar device according to an embodiment of the present invention is shown.

[0045] In all the accompanying drawings, identical or functionally identical elements and devices are equipped with the same reference numerals. The numbering of the method steps is for clarity and should generally not imply a definite chronological order. In particular, multiple method steps may be performed simultaneously. Detailed Implementation

[0046] Figure 1 A schematic block diagram of a radar system 1 for a motor vehicle is shown. The radar system includes a transmit / receive unit 2. This transmit / receive unit can also be configured as a MIMO system. The transmit / receive unit 2 has a true aperture. This should be understood as the transmit / receive unit 2 generating radar measurement signals or sensor data, which can be analyzed and processed using conventional (i.e., not SAR-based) methods to obtain angle estimates, in particular.

[0047] The sensor data is transmitted to the analysis and processing unit 4. The analysis and processing unit 4 includes, for example, a microprocessor, an integrated circuit, etc., to analyze and process the radar measurement signal. A first angle estimate is generated using conventional methods (e.g., FMCW method, fast linear frequency modulation method, MIMO method, etc.). This first angle estimate describes at least one angle at which radar radiation is reflected. Here, the azimuth and (optionally additionally) elevation angle can be estimated.

[0048] Furthermore, the analysis and processing unit 4 is coupled to the interface 3. The interface 3 can detect motion information about the movement of the radar system 1 from an external sensor (e.g., an odometer). The motion information may include, for example, the speed of the radar system 1. Using the motion information about the movement of the radar system 1 and the generated radar measurement signal, the analysis and processing unit 4 calculates at least one second angle estimate for the object using a SAR algorithm.

[0049] Finally, the analysis and processing device 4 fuses at least one first angle estimate with at least one second angle estimate.

[0050] Radar system 1 therefore also functions as a SAR sensor, which should be understood as obtaining angular information from Doppler measurements. This will be explained below based on a linear frequency modulation sequence method. However, the invention is not limited thereto. Thus, other modulation types can also be applied, which use sequences of transmitted waveforms for range and velocity determination or SAR estimation.

[0051] In the case of a SAR sensor employing a linear frequency modulated sequence modulation method, a time series of FMCW ramps is transmitted while the radar system 1 moves in space. Thus, each ramp is transmitted and received at a different location, allowing the time series of frequency ramps to be interpreted as a synthetic aperture, which is then used to perform radar measurements at a given moment. Therefore, a large aperture can be synthesized from a small real aperture through motion.

[0052] Even in the case of SAR sensors, multiple transmit and / or receive channels can be used in some designs. These transmit and / or receive channels can be used, for example, to identify moving targets. Besides SAR, multiple transmit and / or receive channels (real aperture) can also be used for conventional angle estimation. Therefore, both real and synthetic apertures can be implemented in a single sensor.

[0053] The true aperture is preferably used primarily for SAR functionality, but can also be used in parallel for conventional angle estimation. Here, it is advantageous to use the same waveform for both analysis processes.

[0054] Length L of the synthesized aperture SAR With the vehicle's own speed v Ego And with the measured duration T Mess Proportional.

[0055] L SAR =v Ego T Mess

[0056] Figure 2 The length l (in meters) of the composite (virtual) aperture is shown as a function of the self-velocity v (in meters per second). SAR1, SAR2, and SAR3 correspond to virtual apertures for different measurement times of 0.03 seconds, 0.05 seconds, and 0.1 seconds. DBF corresponds to a real aperture of 5 centimeters, independent of velocity.

[0057] The effective aperture L can be obtained from two antenna arrays (synthetic and real). eff A monotonic function is used to generate weights for combining two angular measurements, the effective apertures being obtained in the given directions. Effective aperture L eff It is also a function of the length L of the corresponding aperture and the orientation δ of the antenna array.

[0058] L eff =L cos(θ-δ)

[0059] In this formula, θ is the azimuth angle defined in the vehicle coordinate system. Here, θ = 0° corresponds to an object in the direction of travel. The SAR orientation δ is -90°.

[0060] Figure 3 A schematic top view of a motor vehicle equipped with radar system 1 is shown to illustrate angle definitions and marking conventions. The observation direction A2 of the synthetic aperture is always orthogonal to the travel direction A1. The observation direction A3 of the real (especially MIMO) array depends on the orientation δ relative to the travel direction A1.

[0061] In the following appendix Figures 4 to 6 In this system, all angles are converted to the vehicle coordinate system, where 0° corresponds to the direction of travel.

[0062] Figure 4 For an angle radar system (i.e., with a directional δ = 45°), the effective aperture L depends on the azimuth angle θ. eff Effective aperture L eff Corresponding to the aperture of the target as seen at a given angle. Curves 301 and 302 correspond to a composite aperture of 21 cm or 10.5 cm, and curve 303 corresponds to a true aperture of 5 cm. Line 304 illustrates sensor orientation.

[0063] The synthesized aperture and the real (or MIMO) aperture have their maximum and minimum values ​​at different locations. By fusing, it is possible to achieve angular resolution and accuracy that are proportional to the maximum value of the effective aperture (real and synthesized) at the corresponding angle.

[0064] Figure 5 For a radar system 1 oriented in the direction of travel (i.e., with orientation δ = 0°), the effective aperture L, depending on the azimuth angle θ, is shown. eff Curves 401 and 402 again correspond to the composite apertures of 21 cm and 10.5 cm, and curve 403 corresponds to the actual aperture of 5 cm. Line 404 illustrates the sensor orientation.

[0065] Figure 6 For side radar system 1 (i.e., with a directional δ = 90°), the effective aperture L depends on the azimuth angle θ. eff Curves 501 and 502 again correspond to synthetic apertures of 21 cm and 10.5 cm, respectively, and curve 503 corresponds to a true aperture of 5 cm. Line 504 illustrates sensor orientation. Here, the effective aperture size is upgraded in the same way for both the true and synthetic apertures across the angular range. For angular fusion, the true aperture (e.g., with the aid of DBF) provides the underlying performance, which is improved with larger SAR apertures in the context of velocity.

[0066] Figure 7A flowchart is shown for a method of operating radar device 1, which is particularly applicable to the radar device 1 described above. Conversely, the radar device 1 can be constructed to implement the following method.

[0067] In the first method step S1, radar radiation is emitted by the transmitting and receiving device 2 with a real aperture of the radar system 1. The transmitting and receiving device 2 receives the reflected radar radiation and generates a radar measurement signal. The radar measurement signal is then preprocessed.

[0068] In step S2, range Doppler processing is performed using a SAR algorithm. Here, motion information regarding the motion of radar system 1, particularly the velocity and / or trajectory of radar system 1, is considered. This consideration is performed, for example, using a linear frequency modulated Z-transform.

[0069] In step S3, peak detection is performed. The distance and velocity are described for potential targets (objects).

[0070] In step S4, the DBF algorithm is used for analysis, taking into account the actual aperture. This yields at least one first angle estimate, specifically the azimuth and (optionally) elevation angle.

[0071] In step S5, moving target identification is performed to determine whether the corresponding object is stationary. This can be achieved by considering the angle obtained in step S4, the distance obtained in step S3, the velocity obtained in step S3, and the self-velocity of radar system 1.

[0072] If no moving target is involved, i.e., if the object is stationary, then in step S7, at least one second angle estimate is calculated based on Doppler information using a SAR algorithm, taking into account the self-velocity of radar system 1. In particular, the azimuth and (optionally) elevation angles can be calculated again.

[0073] In step S8, at least one second angle estimate is fused with at least one first angle estimate calculated in step S4. During angle fusion, if the transmitting / receiving device 2 has a line-of-sight extending to both sides of the driving direction, multi-value resolution can also be performed on at least one second angle estimate of the object calculated using the SAR algorithm. During angle fusion, not only the azimuth angle but also the elevation angle is considered.

[0074] If the target is identified as not stationary in step S6, the target's velocity is calculated based on Doppler information in step S10. The azimuth is determined using DBF information.

[0075] Finally, in step S9, parameters such as distance (range), elevation angle, azimuth angle, and fused azimuth and elevation angles are estimated or output.

Claims

1. A radar system (1) for a motor vehicle, said radar system having: A transmitting and receiving device (2) with a real aperture, the transmitting and receiving device being configured to emit radar radiation, receive reflected radar radiation and generate radar measurement signals; Analysis and processing device (4), the analysis and processing device is configured to, a. Using the generated radar measurement signal, based on the true aperture, calculate at least one first angle estimate for an object that reflects the radar radiation at that object; b. Using motion information about the motion of the radar system (1) and the generated radar measurement signals, at least one second angle estimate is calculated for the object using a synthetic aperture radar (SAR) algorithm, wherein, The synthetic aperture of the radar system (1) can be synthesized by means of the real aperture through the motion; c. Fusing the at least one first angle estimate with the at least one second angle estimate, wherein the analysis processing device (4) is configured to weight the at least one first angle estimate and the at least one second angle estimate during the fusion.

2. The radar system (1) according to claim 1, wherein The analysis and processing device (4) is configured to weight the first angle estimate based on the first angle estimate and / or the second angle estimate.

3. The radar system (1) according to claim 1 or 2, wherein The analysis and processing device (4) is configured to weight the first angle estimate based on the true aperture and to weight the second angle estimate based on the synthetic aperture of the radar system (1).

4. The radar system (1) according to any one of the preceding claims, wherein, The analysis and processing device (4) is configured to fuse the at least one first angle estimate with the at least one second angle estimate based on the effective aperture related to the true aperture and the effective aperture related to the synthetic aperture of the radar system (1).

5. The radar system (1) according to any one of the preceding claims, wherein, The at least one first angle estimate and the at least one second angle estimate respectively include azimuth and elevation.

6. The radar system (1) according to any one of the preceding claims, wherein, The analysis and processing device (4) is further configured to perform moving target identification using the generated radar measurement signal in order to identify whether the object is stationary, wherein the analysis and processing device (4) is configured to calculate the at least one second angle estimate for the object only for stationary objects.

7. The radar system (1) according to any one of the preceding claims, wherein, The analysis and processing device (4) is configured to consider not only the azimuth angle but also the elevation angle when fusing the at least one first angle estimate with the at least one second angle estimate.

8. A method for using a radar system (1) for operating a motor vehicle, the method comprising the following steps: With the aid of the transmitting and receiving device (2) with a real aperture of the radar system (1), radar radiation is emitted, the reflected radar radiation is received, and radar measurement signals are generated. Using the generated radar measurement signal, based on the true aperture, at least one first angle estimate is calculated for an object that reflects the radar radiation at that object; Using a synthetic aperture radar (SAR) algorithm, at least one second angle estimate is calculated for the object, taking into account the generated radar measurement signals and motion information about the movement of the radar system. The synthetic aperture of the radar system (1) can be synthesized by means of the real aperture through the motion; The at least one first angle estimate and the at least one second angle estimate are fused together, wherein, during the fusion, the at least one first angle estimate and the at least one second angle estimate are weighted.

Images