Computing device for evaluating radar data from at least one MIMO radar sensor with a non-uniform antenna array

Abstract

A computing device for evaluating radar data from at least one MIMO radar sensor (2) with a non-uniform antenna array (3) is proposed, comprising at least one interface (4) for receiving the radar data, and at least one computing module (5) which is provided for dividing the radar data into first radar data of a uniform portion (6) of a virtual antenna array (7) and second radar data of a non-uniform portion (8) of the virtual antenna array (7), further processing the first and second radar data separately from one another in order to obtain a uniform beam pattern (9) and a non-uniform beam pattern (10), combining the uniform and non-uniform beam patterns (9, 10) in order to obtain an overall beam pattern (11), and evaluating the overall beam pattern (11) by means of a trained artificial neural network (12) in order to determine a unique angular position of at least one radar target (13). receive.

Description

[0001] The invention relates to a computing device for evaluating radar data from at least one MIMO radar sensor with a non-uniform antenna array. Furthermore, the invention also relates to a control unit for a vehicle, a radar system, and a corresponding vehicle. Furthermore, the invention relates to a computer-implemented method for evaluating radar data from at least one MIMO radar sensor with a non-uniform antenna array and to a corresponding computer program product. Furthermore, the invention also relates to a method for training an artificial neural network.

[0002] MIMO (Multiple Input Multiple Output) radar (Radio Detection and Ranging) sensors with non-uniform antenna arrays are known from the prior art. These MIMO radar sensors typically provide ambiguous results regarding the angular positions of detected radar targets. These ambiguities cannot be reliably resolved using known methods. The present invention overcomes this problem and enables the unambiguous determination of the angular positions of radar targets.

[0003] DE 10 2023 204 002 A1 describes an antenna array for a radar sensor. DE 10 2023 201 585 A1 describes a method for radar angle estimation. US 2022 / 0283265 A1 describes a high-resolution, unambiguous radar. Fuchs J. [et al.]: A Machine Learning Perspective on Automotive Radar Direction of Arrival Estimation. In: IEEE Access Vol. 10, pp. 6775-6797, 2022. - ISSN 2169-3536 shows machine learning techniques for determining object angles for automotive radar sensors.

[0004] A computing device for evaluating radar data from at least one MIMO radar sensor with a non-uniform antenna array is proposed. The computing device comprises at least one interface for receiving the radar data. The computing device comprises at least one computing module designed to divide the radar data into first radar data from a uniform portion of a virtual antenna array and second radar data from a non-uniform portion of the virtual antenna array. The computing module is designed to further process the first and second radar data separately to obtain a uniform beam pattern and a non-uniform beam pattern. The computing module is designed to combine the uniform and non-uniform beam patterns to obtain an overall beam pattern.The computing module is designed to evaluate the entire beam pattern using a trained artificial neural network in order to obtain a unique angular position of at least one radar target.

[0005] The computing device is preferably intended for use in a vehicle. In particular, the radar target can be embodied as an object located in the surroundings of the vehicle, such as another vehicle, a pedestrian, a lane marking, a traffic light, a building, a street sign, or the like. Preferably, a control unit, in particular an electronic control unit (ECU), or a MIMO radar sensor of the vehicle can comprise the computing device. “Intended” should be understood in particular to mean specially programmed, specially equipped, and / or specially designed. The fact that an object is intended for a function should be understood in particular to mean that the object performs the function in at least one operating state.

[0006] The computing device, in particular the computing module, can be designed at least partially in particular as a microprocessor, as a microcontroller, as an integrated circuit, in particular as an FPGA (Field Programmable Gate Array), as an accelerator for artificial neural networks, as an application-specific integrated circuit (ASIC), or the like. The interface is preferably designed as a wired interface. Alternatively, it is conceivable for the interface to be wireless. The interface is provided in particular to connect the MIMO radar sensor and the computing module for data transmission purposes. In particular, the interface is provided to provide the radar data received by the MIMO radar sensor to the computing module.

[0007] The radar data is preferably at least digital raw radar data. In particular, analog radar signals received by the MIMO radar sensor have undergone at least one analog-to-digital conversion in the MIMO radar sensor to obtain digital raw radar data. The MIMO radar sensor is preferably designed as an FMCW (frequency modulated continuous wave) radar sensor. The MIMO radar sensor, in particular the transmitter of the MIMO radar sensor, is / are preferably designed to emit radiation with a ramp-like modulated frequency, so-called frequency ramps. A frequency ramp represents, in particular, a linear change in the frequency of the emitted radiation over a period of time. Other waveforms of the radar radiation are also possible, for example waveforms based on "phase-modulated continuous wave (PMCW) waveforms" or "orthogonal frequency division multiplexing (OFDM) waveforms."

[0008] The MIMO radar sensor preferably has at least one physical antenna array. The antenna array is formed in particular by a plurality of antennas, which are arranged spaced apart from one another, in particular in a common plane. Some of the antennas serve in particular as transmitters, which are intended to transmit radar signals. Another part of the antennas serves in particular as receivers, which are intended to receive the radar signals transmitted by the transmitters and reflected in the vicinity of the MIMO radar sensor, in particular from radar targets. The radar data, in particular, is generated from the received radar signals. The antennas can be designed in particular as planar (patch) antennas, as waveguide antennas, or the like. The antenna array is preferably designed as a non-uniform antenna array."Uneven" means, in particular, that the phase centers of adjacent transmitters and / or adjacent receivers, viewed along at least one direction, are partially non-equidistant from one another. Preferably, the phase centers of adjacent transmitters and / or receivers of the antenna array, viewed along at least one direction, are partially equidistant from one another.

[0009] A virtual antenna array of the MIMO radar sensor results, in particular, from the antenna array when several, preferably all, receivers receive radar signals originally transmitted by several, preferably all, transmitters, in particular at the same time. The virtual antenna array comprises, in particular, a plurality of virtual antennas. A virtual antenna specifies, in particular, which combination of transmitter and receiver is responsible for a radar signal detected at the respective position in the virtual antenna array.

[0010] The non-uniform antenna array results in, in particular, a virtual antenna array that has a uniform portion and a non-uniform portion. The uniform portion includes, in particular, virtual antennas resulting from combinations of antennas arranged equidistantly within the antenna array. The non-uniform portion includes, in particular, virtual antennas resulting from combinations of antennas arranged non-equidistantly within the antenna array.

[0011] The computing module is particularly intended to further process the first and second radar data separately from one another, preferably using different methods. For further processing, generally known methods such as digital beamforming (DBF), fast Fourier transformation (FFT), or the like can be used. The further processing of the radar data results in particular in a beam pattern. The beam pattern preferably indicates several hypotheses for the angular position of a radar target, of which in particular a single hypothesis corresponds to the correct angular position. The uniform beam pattern results in particular from the further processing of the first radar data. The non-uniform beam pattern results in particular from the further processing of the second radar data. An angular position preferably comprises at least one angular position in one azimuth dimension.An angular position can, in particular, consist of an angular position in an azimuth dimension and an angular position in an elevation dimension. In a simplified antenna array without elevation measurement capability, the elevation angle is assumed to be zero.

[0012] In particular, the ambiguity cannot be resolved from the uniform beam pattern or from the non-uniform beam pattern alone. Preferably, the computing module is designed to combine, in particular superimpose, the uniform and non-uniform beam patterns in order to obtain the entire beam pattern. In particular, the computing module is designed to use the entire beam pattern as input data for the artificial neural network. Preferably, the computing module is designed to divide the entire beam pattern, in particular a complex one, into a real part and an imaginary part and to use the real part and the imaginary part together as input data for the artificial neural network.

[0013] The artificial neural network can be designed, in particular, as a convolutional neural network (CNN), as a perceptron, as a recurrent neural network (RNN), as a spiking neural network (SNN), or as another artificial neural network deemed appropriate by a person skilled in the art. Preferably, the artificial neural network is designed as a two-dimensional artificial neural network. In particular, the computing module is designed to execute the artificial neural network. The artificial neural network is trained, in particular by means of supervised training, to select the correct hypothesis for an angular position of a radar target from the hypotheses contained in the entire beam pattern and to output this angular position as the unique angular position of the radar target.Preferably, the artificial neural network is designed to output a unique angular position of each radar target depending on a plurality of radar targets contained in the radar data. In particular, the computing device can comprise a plurality of computing modules. In particular, various of the data processing steps described above can be performed by different computing modules.

[0014] The inventive design of the computing device advantageously allows for the determination of unambiguous angular positions of radar targets using MIMO radar sensors with non-uniform antenna arrays. A computationally efficient computing device for angular position determination can advantageously be provided.

[0015] Furthermore, it is proposed that the artificial neural network be designed as a two-dimensional convolutional neural network. In particular, the input data for the artificial neural network are designed as two-dimensional matrices. In particular, the output data of the artificial neural network are designed as a two-dimensional matrix. Preferably, one dimension of the matrices is proportional to the azimuth dimension and another dimension of the matrices is proportional to the elevation dimension, for example in a UV representation. The beam patterns are in particular two-dimensional power distributions, wherein the hypotheses for an angular position correspond to power peaks in the power distribution. Preferably, the artificial neural network is provided to output an output beam pattern that comprises only a single power peak per radar target, corresponding to the unique angular position.Advantageously, an efficient and low-memory analysis of beam patterns can be enabled.

[0016] It is further proposed that the artificial neural network have a plurality of intermediate layers (hidden layers) which differ from one another at least with regard to their parameters. In particular, the artificial neural network has at least three, preferably at least five, and particularly preferably at least seven intermediate layers. Most preferably, the artificial neural network has between three and seven intermediate layers. Each intermediate layer preferably has a plurality of, for example five, filter kernels. Each intermediate layer preferably has an activation function, for example a Rectified Linear Unit (ReLU). The parameters by which the intermediate layers differ from one another are, in particular, weights of neurons in the intermediate layers. In particular, the parameters are set by training the artificial neural network.Advantageously, clear angular positions can be determined particularly reliably.

[0017] It is further proposed that the artificial neural network be trained using a plurality of radar data containing known angular positions of at least one radar target to detect a correct angular position of the at least one radar target in a plurality of ambiguous angular positions. The artificial neural network is preferably trained using supervised training. In particular, the radar data contain a plurality of different angular positions of at least one radar target, in particular different combinations of angular positions in azimuth and elevation dimensions. In particular, the radar data can contain multiple radar targets, in particular with different angular positions.In particular, the radar data can contain a plurality of different angular positions in combination with different and / or identical speeds and / or distances of the at least one radar target from the MIMO radar sensor. The MIMO radar sensor is preferably designed as a 4D MIMO radar sensor, which can detect, in particular, an angular position in the azimuth dimension, an angular position in the elevation dimension, a speed, and a distance from radar targets. The radar data is preferably simulated radar data. In particular, the radar data is simulated depending on a configuration of the MIMO radar sensor, in particular the antenna array. Alternatively or additionally, it is conceivable for the artificial neural network to be trained using real radar data.Preferably, the weights of the artificial neural network are adjusted during training such that the trained artificial neural network can detect a correct angular position of at least one radar target in a plurality of ambiguous angular positions. Advantageously, a particularly reliable artificial neural network can be provided.

[0018] It is also proposed that the computing module be provided to further process the first radar data using a two-dimensional fast Fourier transform and / or the second radar data using digital beamforming. In particular, the fast Fourier transform requires less computing time than digital beamforming. In particular, the fast Fourier transform can only be meaningfully applied to radar data from a uniform virtual antenna array or at least a uniform portion of the virtual antenna array. In particular, the further processing of the first radar data using the two-dimensional fast Fourier transform only provides a partial area of ​​the uniform beam pattern, in particular corresponding to a partial area of ​​a field of view of the MIMO radar sensor. The uniform beam pattern is preferably constructed periodically due to the uniformity of the portion of the virtual antenna array.Preferably, the computing module is designed to multiply the partial area of ​​the uniform beam pattern obtained by the two-dimensional fast Fourier transform and to combine it into a complete uniform beam pattern with the size of the non-uniform beam pattern. Preferably, the computing module is designed to combine the uniform beam pattern thus obtained with the non-uniform beam pattern to obtain the entire beam pattern. Preferably, the computing module is designed to further process the second radar data using conventional digital beamforming (CBF). This advantageously enables a computationally time-saving evaluation of the radar data.

[0019] Furthermore, a control unit for a vehicle is proposed. The control unit comprises at least one computing device according to the invention. The control unit can be designed, in particular, as a high-performance vehicle computer. In particular, the control unit can be provided to process, in particular evaluate, signals from additional sensors of the vehicle, such as cameras or LIDAR (Light Detection and Ranging) sensors. In particular, the control unit can be provided to generate control commands for at least partially autonomous control of the vehicle, in particular depending on the results of the radar data analysis. Advantageously, a control unit with high-performance radar data processing can be provided.

[0020] Furthermore, a radar system is proposed. The radar system comprises at least one MIMO radar sensor with a non-uniform antenna array. The radar system comprises at least one computing device according to the invention or at least one control unit according to the invention. In particular, the computing device can be integrated into the MIMO radar sensor. In particular, the computing device can be arranged in a housing of the MIMO radar sensor. In particular, the computing device can be designed as a signal processor of the MIMO radar sensor. Alternatively, the control unit can comprise the computing device and be arranged separately from the MIMO radar sensor. In particular, the radar system can comprise a plurality of MIMO radar sensors and / or computing devices or control units. In particular, a computing device can be provided for evaluating radar data from multiple MIMO radar sensors.Advantageously, a radar system can be provided with powerful radar data processing.

[0021] Furthermore, an automated vehicle is proposed. The automated vehicle comprises at least one radar system according to the invention. An "automated vehicle" is to be understood in particular as a vehicle with one of the automation levels 1 to 5 of the SAE J3016 standard. In particular, the automated vehicle has technical equipment required for these automation levels. The technical equipment includes, in particular, environmental detection sensors, such as at least one MIMO radar sensor, lidar sensors, cameras and / or acoustic sensors, control units, or the like. The vehicle is preferably designed as a land vehicle, in particular as a road vehicle.The automated vehicle can be designed, in particular, as a passenger car, preferably as a passenger transport vehicle, as a truck, as a construction vehicle, as an agricultural vehicle, or as any other vehicle deemed appropriate by a person skilled in the art. Alternatively, the automated vehicle can also be designed as an aircraft, for example, a drone, an airplane, a helicopter, a vertical takeoff and landing aircraft, or the like, as a rail vehicle, for example, a locomotive, a wagon, or the like, or as a watercraft, in particular, a ship, a boat, or the like. Advantageously, a particularly roadworthy automated vehicle can be provided.

[0022] Furthermore, a computer-implemented method for evaluating radar data from at least one MIMO radar sensor with a non-uniform antenna array is proposed. The received radar data is divided into first radar data from a uniform portion of a virtual antenna array and second radar data from a non-uniform portion of the virtual antenna array. The first and second radar data are further processed separately to obtain a uniform beam pattern and a non-uniform beam pattern. The uniform and non-uniform beam patterns are combined to obtain an overall beam pattern. The overall beam pattern is evaluated using a trained artificial neural network to obtain a unique angular position of at least one radar target. Advantageously, a computationally efficient method for precisely determining the angular position of radar targets can be provided.

[0023] Furthermore, a computer program product for evaluating radar data from at least one MIMO radar sensor with a non-uniform antenna array is proposed. The computer program product comprises execution instructions that, when executed by a computing device according to the invention, cause the device to execute a method according to the invention. Advantageously, a computer program product can be provided that enables computationally efficient and precise angular position determination of radar targets.

[0024] Furthermore, a method for training an artificial neural network, which is used by a computing device according to the invention to obtain an unambiguous angular position of at least one radar target, is proposed. A plurality of radar data containing known angular positions of at least one radar target is input into the artificial neural network. Parameters of the artificial neural network are adjusted as a function of an output of the artificial neural network. In particular, the weights of the artificial neural network are adjusted, in particular such that the trained artificial neural network can recognize a correct angular position of at least one radar target in a plurality of ambiguous angular positions. The artificial neural network can advantageously be trained specifically for angular position recognition in radar data.

[0025] The invention is illustrated by an exemplary embodiment in the following figures. They show: Fig. 1 shows a schematic representation of an automated vehicle according to the invention, Fig. 2 a computing device according to the invention of the automated vehicle according to the invention from Fig. 1 in a schematic representation, Fig. 3 an antenna array of a MIMO radar sensor of the automated vehicle according to the invention from Fig. 1 in a schematic representation, Fig. 4 a virtual antenna array of the MIMO radar sensor Fig. 3 in a schematic representation, Fig. 5 beam patterns in a schematic representation, Fig. 6 an artificial neural network of the computing device according to the invention Fig. 2 in a schematic representation, Fig. 7 Output data of the artificial neural network from Fig. 6 in a schematic representation, Fig. 8 a flow diagram of a computer-implemented method according to the invention in a schematic representation and Fig. 9 a flow diagram of a method according to the invention in a schematic representation.

[0026] Fig. 1 shows a schematic representation of an automated vehicle 16. In the present exemplary embodiment, the automated vehicle 16 is embodied, for example, as a land vehicle, in particular as a passenger car. The automated vehicle 16 comprises at least one radar system 17. The radar system 17 comprises at least one MIMO radar sensor 2 with a non-uniform antenna array 3 (cf. Fig. 3). The radar system 17 comprises at least one control unit 15. The control unit 15 comprises at least one computing device 1. In an alternative embodiment, it is conceivable that the MIMO radar sensor 2 comprises the computing device 1. Furthermore, in Fig. 1 shows a radar target 13. For example, the radar target 13 is embodied as a foreign vehicle.

[0027] Fig. 2 shows the computing device 1 of the automated vehicle 16 from Fig. 1 in a schematic representation. The computing device 1 is provided for evaluating radar data from the MIMO radar sensor 2. The computing device 1 comprises at least one interface 4 for receiving the radar data. The computing device 1 comprises at least one computing module 5.

[0028] Fig. 3 shows the, in particular physical, antenna array 3 of the MIMO radar sensor 2 of the automated vehicle 16 from Fig. 1 in a schematic representation. The antenna array 3 is formed by a plurality of antennas 18, 19, which are arranged spaced apart from one another in a common plane. Some of the antennas 18, 19 serve as transmitters 18, which are intended to transmit radar signals. Another part of the antennas 18, 19 serves as receivers 19, which are intended to receive the radar signals transmitted by the transmitters 18 and reflected in the vicinity of the MIMO radar sensor 2, in particular at radar targets 13. The radar data is generated from the received radar signals. Fig. 3 shows the positions of phase centers 20, 21 of the antennas 18, 19. The phase centers 20 of the transmitters 18 are shown as crosses. The phase centers 21 of the receivers 19 are shown as plus signs. The antenna array 3 is designed as a non-uniform antenna array. In the present exemplary embodiment, for example, the phase centers 20 of adjacent transmitters 18, viewed along a direction parallel to an abscissa axis 22 and along a direction parallel to an ordinate axis 23, are not arranged equidistant from one another, in particular in the center of the antenna array 3. The phase centers 21 of the receivers 19 are arranged equidistant from one another along the direction parallel to the abscissa axis 22 and along the direction parallel to the ordinate axis 23.

[0029] Fig. 4 shows a virtual antenna array 7 of the MIMO radar sensor 2 from Fig. 3 in a schematic representation. The virtual antenna array 7 comprises a plurality of virtual antennas 24. The virtual antenna array 7 has a uniform portion 6 and a non-uniform portion 8. The uniform portion 6 comprises virtual antennas 24 that result from combinations of antennas 18, 19 arranged equidistantly in the antenna array 3. The non-uniform portion 8 comprises virtual antennas 24 that result from combinations with antennas 18, 19 that are not arranged equidistantly in the antenna array 3. The uniform portion 6 and the non-uniform portion 8 are shown on the right side of the Fig. 4 is shown separately. The computing module 5 is intended to divide the radar data into first radar data of the uniform portion 6 of the virtual antenna array 7 and second radar data of the non-uniform portion 8 of the virtual antenna array 7.

[0030] Fig. 5 shows beam patterns 9-11 in a schematic representation. Shown are a uniform beam pattern 9, a non-uniform beam pattern 10, and an entire beam pattern 11. The computing module 5 is designed to further process the first and second radar data separately to obtain the uniform beam pattern 9 and the non-uniform beam pattern 10. The beam patterns 9-11 indicate several hypotheses 25, 26 for the angular position of the radar target 13, of which a single hypothesis 25 corresponds to the correct angular position. The uniform beam pattern 9 results from the further processing of the first radar data. The non-uniform beam pattern 10 results from the further processing of the second radar data.

[0031] The computing module 5 is provided to further process the first radar data using a two-dimensional fast Fourier transform and / or the second radar data using digital beamforming. The further processing of the first radar data using the two-dimensional fast Fourier transform yields only a partial area 27 of the uniform beam pattern 9, in particular corresponding to a partial area of ​​a field of view of the MIMO radar sensor 2. The uniform beam pattern 9 is periodically structured due to the uniformity of the portion 6 of the virtual antenna array 7. The computing module 5 is provided to multiply the partial area 27 of the uniform beam pattern 9 obtained by the two-dimensional fast Fourier transform and to combine it into a complete uniform beam pattern 9 with the size of the non-uniform beam pattern 10.The computing module 5 is intended to further process the second radar data using conventional digital beamforming.

[0032] Beam patterns 9-11 are two-dimensional power distributions, with hypotheses 25, 26 for an angular position corresponding to power peaks in the power distribution. One dimension 28 of beam patterns 9-11 is proportional to an azimuth dimension, and another dimension 29 of beam patterns 9-11 is proportional to an elevation dimension, shown here as an example in a UV representation. The calculation module 5 is designed to combine the uniform and non-uniform beam patterns 9, 10, in particular to superimpose them, in order to obtain the entire beam pattern 11.

[0033] Fig. 6 shows an artificial neural network 12 of the computing device 1 from Fig. 2 in a schematic representation. The computing module 5 is intended to evaluate the entire beam pattern 11 using the trained artificial neural network 12 in order to obtain a unique angular position of the at least one radar target 13. The computing module 5 is intended to use the entire beam pattern 11 as input data for the artificial neural network 12. The computing module 5 is intended to divide the, in particular complex, entire beam pattern 11 into a real part 30 and an imaginary part 31 and to use the real part 30 and the imaginary part 31 together as input data for the artificial neural network 12.

[0034] The computing module 5 is designed to execute the artificial neural network 12. The artificial neural network 12 is trained, in particular by means of supervised training, to select the correct hypothesis 25 from the hypotheses 25, 26 contained in the entire beam pattern 11 for an angular position of the radar target 13 and to output this angular position as the unique angular position of the radar target 13. The artificial neural network 12 is designed as a two-dimensional convolutional neural network.

[0035] The artificial neural network 12 has a plurality of intermediate layers 14 that differ from one another at least with respect to their parameters. The parameters by which the intermediate layers 14 differ from one another are weights of neurons in the intermediate layers 14. The parameters are adjusted by training the artificial neural network 12.

[0036] Fig. 7 shows output data of the artificial neural network 12 from Fig. 6 in a schematic representation. The artificial neural network 12 is trained by means of a large number of radar data containing known angular positions of at least one radar target 13 to recognize a correct angular position of the at least one radar target 13 in a plurality of ambiguous angular positions. The artificial neural network 12 is trained by means of supervised training. The radar data contain a large number of different angular positions of at least one radar target 13, in particular different combinations of angular positions in azimuth and elevation dimensions. The artificial neural network 12 is provided to output an output beam pattern 32 as output data, which comprises only a single power peak for each radar target 13 corresponding to the unique angular position.

[0037] Fig. Figure 8 shows a flowchart of a computer-implemented method for evaluating radar data from the MIMO radar sensor 2 with a non-uniform antenna array 3 in a schematic representation. In a first method step 33, the received radar data is divided into first radar data from the uniform portion 6 of the virtual antenna array 7 and second radar data from the non-uniform portion 8 of the virtual antenna array 7. In a second method step 34, the first and second radar data are further processed separately to obtain a uniform beam pattern 9 and a non-uniform beam pattern 10. In a third method step 35, the uniform and non-uniform beam patterns 9, 10 are combined to obtain an overall beam pattern 11.In a fourth method step 36, the entire beam pattern 11 is evaluated by means of the trained artificial neural network 12 in order to obtain a unique angular position of the at least one radar target 13.

[0038] A computer program product for evaluating radar data of the at least one MIMO radar sensor 2 with the non-uniform antenna array 3 comprises execution instructions which, when the program is executed by the computing device 1, cause the computing device 1 to execute the computer-implemented method.

[0039] Fig.9 shows a flowchart of a method for training the artificial neural network 12, which is used by the computing device 1 to obtain a unique angular position of the at least one radar target 13, in a schematic representation. In a first method step 37, radar data containing a known angular position of the at least one radar target 13 is input into the artificial neural network 12. The radar data can contain a plurality of radar targets 13, in particular with different angular positions. In a second method step 38, parameters of the artificial neural network 12 are adjusted depending on an output of the artificial neural network 12.Through training, the weights of the artificial neural network 12 are adjusted, in particular such that the trained artificial neural network 12 can recognize a correct angular position of the at least one radar target 13 in a plurality of ambiguous angular positions.

[0040] Method steps 37, 38 can be repeated multiple times, so that a total of a plurality of radar data containing known, in particular different, angular positions of the at least one radar target 13 is input into the artificial neural network 12. The radar data can contain a plurality of different angular positions in combination with different and / or identical speeds and / or distances of the at least one radar target 13 from the MIMO radar sensor 2. The MIMO radar sensor 2 is embodied, for example, as a 4D MIMO radar sensor that can detect an angular position in the azimuth dimension, an angular position in the elevation dimension, a speed, and a distance from radar targets 13. The radar data can be simulated radar data. The radar data can be simulated depending on a configuration of the MIMO radar sensor 2, in particular of the antenna array 3.Alternatively or additionally, it is conceivable that the artificial neural network 12 is trained using real radar data. Reference symbol 1 computing device 2 MIMO radar sensors 3 antenna array 4 Interface 5 Calculation module 6 equal share 7 virtual antenna array 8 uneven proportion 9 uniform beam pattern 10 uneven beam pattern 11 total beam pattern 12 artificial neural network 13 radar target 14 Intermediate layer 15 Control unit 16 vehicles 17 Radar system 18 transmitters 19 recipients 20 Phase center 21 Phase center 22 Abscissa axis 23 Ordinate axis 24 virtual antenna 25 Hypothesis 26 Hypothesis 27 sub-area 28 dimensions 29 dimensions 30 real part 31 Imaginary part 32 output beam patterns 33 Process step 34 process steps 35 process steps 36 process steps 37 Process step 38 process steps

Claims

[1] Computing device for evaluating radar data from at least one MIMO radar sensor (2) with a non-uniform antenna array (3), comprising at least one interface (4) for receiving the radar data, and at least one computing module (5) which is provided for dividing the radar data into first radar data of a uniform portion (6) of a virtual antenna array (7) and second radar data of a non-uniform portion (8) of the virtual antenna array (7), further processing the first and second radar data separately from one another in order to obtain a uniform beam pattern (9) and a non-uniform beam pattern (10), combining the uniform and non-uniform beam patterns (9, 10) in order to obtain an overall beam pattern (11), and evaluating the overall beam pattern (11) by means of a trained artificial neural network (12) in order to obtain an unambiguous angular position of at least one radar target (13). [2] Computing device according to claim 1, wherein the artificial neural network (12) is designed as a two-dimensional convolutional neural network. [3] Computing device according to claim 1 or 2, wherein the artificial neural network (12) has a plurality of intermediate layers (14) which differ from one another at least with regard to their parameters. [4] Computing device according to one of the preceding claims, wherein the artificial neural network (12) is trained by means of a plurality of radar data containing known angular positions of at least one radar target (13) to recognize a correct angular position of the at least one radar target (13) in a plurality of ambiguous angular positions. [5] Computing device according to one of the preceding claims, wherein the computing module (5) is provided to further process the first radar data by means of a two-dimensional fast Fourier transformation and / or the second radar data by means of digital beamforming. [6] Control unit for a vehicle (16), comprising at least one computing device (1) according to one of the preceding claims. [7] Radar system comprising at least one MIMO radar sensor (2) with a non-uniform antenna array (3) and at least one computing device (1) according to one of claims 1 to 5 or at least one control device (15) according to claim 6. [8] Automated vehicle comprising at least one radar system (17) according to claim 7. [9] Computer-implemented method for evaluating radar data from at least one MIMO radar sensor (2) with a non-uniform antenna array (3), wherein the received radar data are divided into first radar data from a uniform portion (6) of a virtual antenna array (7) and second radar data from a non-uniform portion (8) of the virtual antenna array (7), wherein the first and second radar data are further processed separately from one another to obtain a uniform beam pattern (9) and a non-uniform beam pattern (10), wherein the uniform and non-uniform beam patterns (9, 10) are combined to obtain an overall beam pattern (11), and wherein the overall beam pattern (11) is evaluated by means of a trained artificial neural network (12) to obtain a unique angular position of at least one radar target (13). [10] Computer program product for evaluating radar data from at least one MIMO radar sensor (2) with a non-uniform antenna array (3), comprising execution instructions which, when the program is executed by a computing device (1) according to one of claims 1 to 5, cause the computing device (1) to execute a method according to claim 9. [11] Method for training an artificial neural network (12) used by a computing device (1) according to one of claims 1 to 5 in order to obtain a unique angular position of at least one radar target (13), wherein a plurality of radar data containing known angular positions of at least one radar target (13) is input into the artificial neural network (12) and parameters of the artificial neural network (12) are set as a function of an output of the artificial neural network (12).

Images