DE-102019111248-B4 - Vehicle radar system and methods for calibrating it

DE102019111248B4DE 102019111248 B4DE102019111248 B4DE 102019111248B4DE-102019111248-B4

Abstract

Method for calibrating a vehicle radar system, wherein the vehicle radar system comprises a transmitting antenna arrangement with a plurality of transmitting antennas and a receiving antenna arrangement with a plurality of receiving antennas, wherein the method comprises the following steps: Transmitting a large number of signals using the transmitting antenna array; Receive a multitude of signals with the receiving antenna arrangement; Receiving a multitude of antenna responses based on the multitude of received signals, each of the antenna responses containing positional information regarding a target object; Applying a multitude of hypothetical calibration matrices to each of the multitude of receiving antenna responses to obtain a multitude of calibrated array responses, each of the multitude of hypothetical calibration matrices incorporating calibration information relating to the vehicle radar system; Applying at least one beam-shaping matrix to each of the multitude of calibrated array reactions to obtain a multitude of beam-shaping images; Deriving at least one uncertainty metric for each of the plurality of beamforming images, wherein each of the plurality of uncertainty metrics is representative of a beamforming image resolution; Selecting at least one of the multitude of hypothetical calibration matrices based on the multitude of uncertainty metrics, wherein the selected hypothetical calibration matrix is associated with the uncertainty metric of the one with the best beamforming image resolution; and Using the selected hypothetical calibration matrix to calibrate the vehicle radar system.

Inventors

Dani Raphaeli
Oded Bialer
IGAL BILIK

Assignees

GM Global Technology Operations LLC
SILANTRIX LTD.

Dates

Publication Date: 20260513
Application Date: 20190501
Priority Date: 20180515

Claims (12)

A method for calibrating a vehicle radar system, wherein the vehicle radar system comprises a transmit antenna array with a plurality of transmit antennas and a receive antenna array with a plurality of receive antennas, the method comprising the following steps: Transmitting a plurality of transmit signals with the transmit antenna array; Receiving a plurality of receive signals with the receive antenna array; Obtaining a plurality of antenna responses based on the plurality of receive signals, each of the antenna responses containing position information with respect to a target object; Applying a plurality of hypothetical calibration matrices to each of the plurality of receive antenna responses to obtain a plurality of calibrated array responses, each of the plurality of hypothetical calibration matrices containing calibration information with respect to the vehicle radar system; Applying at least one beamshaping matrix to each of the plurality of calibrated array responses to obtain a plurality of beamshaping images; Deriving at least one uncertainty metric for each of the plurality of beamforming images, wherein each of the plurality of uncertainty metrics is representative of a beamforming image resolution; Selecting at least one of the plurality of hypothetical calibration matrices based on the plurality of uncertainty metrics, wherein the selected hypothetical calibration matrix is associated with the uncertainty metric of the one with the best beamforming image resolution; and Using the selected hypothetical calibration matrix to calibrate the vehicle radar system.
Procedure according to Claim 1 , further comprising the following steps: generating a modulated signal US 1 with a transmitter; mixing the modulated signal MS 1 with a code sequence C 1 to C N to produce the plurality of transmit signals Tx 1 to Tx N , wherein the code sequence C 1 to C N includes at least one separate code for each of the plurality of transmit antennas in the transmit antenna arrangement, the mixing step being performed before the transmit step; and decoupling the plurality of receive signals Rx 1 to Rx M with the code sequence C 1 to C N to produce a plurality of separate receive signals S 1,1 to S M,N , wherein the separate receive signals S 1,1 to S M,N include at least N number of separate signals for each of the plurality of receive antennas in the receive antenna arrangement, the decoupling step being performed after the receive step; wherein the receiving step further comprises receiving a plurality of receiving antenna responses X 1,1 to X K,J based on the plurality of separate receiving signals S 1,1 to S M,N , which in turn are based on the plurality of receiving signals Rx 1 to Rx M.
Procedure according to Claim 1 , wherein the step of obtaining further includes carrying out a pre-beam forming process which involves separating the multiple received signals Rx 1 to Rx M are divided into a multiple of separate received signals S 1,1 to S M,N , and the step of separation is performed by a receiver that is part of the vehicle radar system.
Procedure according to Claim 3 , wherein the receiving step further includes performing a pre-beam shaping process which involves filtering the separate received signals S 1,1 to S M,N based on a Doppler frequency shift f and/or a range r and is assigned to the filtered signals one or more Doppler range bins.
Procedure according to Claim 4 , wherein the receiving step further comprises performing a pre-beam shaping process which includes determining whether one or more Doppler range bins adequately indicate the presence of a target object and, if so, using the Doppler range bin to obtain the plurality of receiving antenna responses X 1,1 to X K,J .
Procedure according to Claim 1 , wherein the first application step further comprises retrieving a first set of hypothetical calibration matrices C 1,1 and C 2,1 from the electronic memory in the vehicle radar system, using the first set of hypothetical calibration matrices C 1,1 and C 2,1 to generate one or more subsequent sets of hypothetical calibration matrices C 1,p and C 2,p , and using the subsequent sets of hypothetical calibration matrices C 1,p and C 2,p to generate the plurality of hypothetical calibration matrices C 1,1 to C 1,p and C 2,1 to C 2,p that are applied to the plurality of receiving antenna responses X 1,1 to X K,J .
Procedure according to Claim 1 , wherein the first application step further comprises applying each of the plurality of receiving antenna responses X 1,1 to X K,J to each of the plurality of hypothetical calibration matrices C 1,1 to C 1,p and C 2,1 to C 2,p to obtain a plurality of calibrated array responses Z 1,1,1 to Z k,j,p , wherein the first application step is performed by the vehicle radar system according to the following equation: Z k , j , p = C 1, p X k , j C 2, p where the hypothetical calibration matrices C 1,1 to C 1,p contain altitude calibration information, the hypothetical calibration matrices C 2,1 to C 2,p contain azimuth calibration information, k represents a Doppler range bin, j represents a receive radar frame, and p represents a calibration iteration.
Procedure according to Claim 1 , wherein the second application step further comprises applying the first beam-shaping matrix F 1 and the second beam-shaping matrix F 2 to each of the plurality of calibrated array reactions Z 1,1,1 to Z K,J,P to obtain a plurality of beam-shaping images Y 1,1,1 to Y K,J,P , and the second application step is performed by the vehicle radar system according to the following equation: Y k , j , p = F 1 Z k , j , p F 2 where the first beamformer matrix F 1 contains height information regarding a target object, the second beamformer matrix F 2 contains azimuth information regarding the target object, k represents a Doppler range bin, j represents a receive radar frame, and p represents a calibration iteration.
Procedure according to Claim 1 , wherein the vehicle radar system continues to derive a plurality of uncertainty metrics until at least one holding condition is satisfied, the at least one holding condition being selected from a plurality of holding conditions which includes: a condition if a gradient associated with an uncertainty metric is flat, a condition if a gradient associated with an uncertainty metric is at a local minimum, or a condition if a previous iteration has resulted in an uncertainty metric with better beamforming image resolution.
Procedure according to Claim 1 , wherein the derivation step further comprises calculating an overall uncertainty metric g p for a given set of hypothetical calibration matrices C 1,p , C 2,p , wherein the overall uncertainty metric g p is representative of an overall value for the given set of hypothetical calibration matrices C 1,p , C 2,p with respect to the beamforming image resolution, and wherein the selection step further comprises selecting the given set of hypothetical calibration matrices C 1,p , C 2,p which is associated with the overall uncertainty metric g p with the best beamforming image resolution.
Procedure according to Claim 1 , wherein the application step further comprises obtaining a first new calibration matrix Ĉ 1 and a second new calibration matrix Ĉ 2 based on the selected hypothetical calibration matrix, calculating a new beamforming image Ŷ using the first new calibration matrix Ĉ 1 and the second new calibration matrix Ĉ 2 , and determining one or more target parameters for a target object by evaluating the new beamforming image Ŷ, wherein the application step is performed by the vehicle radar system.
A vehicle radar system, wherein the vehicle radar system is mounted on a carrier vehicle and comprises: a transmitter; a transmit antenna arrangement with a plurality of transmit antennas coupled to the transmitter, wherein the transmit antenna arrangement transmits a plurality of transmit signals; a receive antenna arrangement with a plurality of receive antennas, wherein the receive antenna arrangement receives a plurality of receive signals; and a receiver coupled to the receive antenna arrangement, wherein the receiver is configured to: receive a plurality of antenna responses based on the plurality of receive signals, each of the antenna responses containing position information with respect to a target object; apply a plurality of hypothetical calibration matrices to each of the plurality of receive antenna responses to obtain a plurality of calibrated array responses, each of the plurality of hypothetical calibration matrices containing calibration information with respect to the vehicle radar system; Applying at least one beam-shaping matrix to each of the plurality of calibrated array responses to obtain a plurality of beam-shaping images; deriving at least one uncertainty metric for each of the plurality of beam-shaping images, wherein each of the plurality of uncertainty metrics is representative of a beam-shaping image resolution; selecting at least one of the plurality of hypothetical calibration matrices based on the plurality of uncertainty metrics, wherein the selected hypothetical calibration matrix is associated with the uncertainty metric of the one with the best beam-shaping image resolution; and using the selected hypothetical calibration matrix to calibrate the vehicle radar system.

Description

Technical field The present disclosure relates generally to radar systems, and in particular to systems and methods for calibrating them. background Many modern vehicles are equipped with advanced safety and driver assistance systems that require robust and precise object detection and tracking systems to control the vehicle's maneuvers. These systems utilize periodic or continuous object detection and control algorithms to estimate various object parameters, such as relative object range, speed, direction of travel, and size. For example, radar devices detect and locate objects (i.e., the reflected signal is returned as an echo to the radar and processed there to determine various pieces of information, such as the travel time of the transmitted/received signals). Advanced radar systems in use today can employ a multiple-input multiple-output (MIMO) concept, which uses multiple antennas at the transmitter to send independent waveforms and multiple antennas at the receiver to receive the radar echoes. In a "co-localized" MIMO radar configuration, the antennas in both the transmitter and receiver are positioned close enough together that each antenna observes the same aspect of an object, thus creating the assumption of a point target. A matched filter bank is used in the MIMO receiver to extract the waveform components. When signals are transmitted from different antennas, the echoes of each signal carry independent information about detected objects and their different propagation paths. Phase differences caused by different transmitting antennas, together with phase differences caused by different receiving antennas, mathematically form a virtual antenna array, resulting in a larger virtual aperture with fewer antenna elements. Conceptually, the virtual array is created by interleaving each of the antenna elements of the transmitter tx and the receiver rx such that the elements in the virtual array represent tx-rx pairs for the respective transmitter tx and receiver rx antennas in the MIMO array. For co-localized MIMO antennas, a transmit array with n transmit antennas and a receive array with m receive antennas generates a virtual array with mxn virtual receiver elements. In other words, the waveforms are extracted by the matched filters at the receiver such that a total of mxn extracted signals are present in the virtual array. The mxn virtual receiver elements can be used to create a beamforming image. Over time, certain aspects of the radar configuration, such as predefined values used in generating the beamforming image, may need to be adjusted or calibrated to maintain the desired level of accuracy in a given radar configuration. The DE 10 2014 208 899 A1 This document describes a method for calibrating the antenna pattern of a MIMO radar sensor with Ntx transmit antenna elements and Nrx receive antenna elements, comprising the following steps: before commissioning the radar sensor: - storing an antenna pattern that assigns a control vector to each of several angles, composed of a transmit control vector and a receive control vector; after commissioning: - performing a radar measurement to locate an object, - checking whether the located object is a single target or multiple targets, - if it is a single target: - performing a SIMO measurement with each of the transmit antenna elements, - estimating the angle of the object based on the measurement results, - calculating a first reference value for each transmit antenna element, dependent on the components of the transmit control vector, - calculating a second reference value for each transmit antenna element, dependent on the results of the SIMO measurements, and - Correcting the transmit control vector based on a known relationship between the first and second comparison values for each transmit antenna element. The US 2012 / 0 081 247 A1 This describes a radar system. The radar system includes a transmitting antenna and a receiving antenna, formed by a group of radiation elements. Antenna beams are calculated in P directions using a BFC function. Target detection via sidelobes of the beams is processed by an algorithm that compares the levels received in a range-velocity resolution cell, where a single detection for each range-velocity resolution cell is at most not possible. Processing means assume that for a given resolution cell of the radar, either in velocity mode or range mode, or alternatively depending on the implemented Processing may yield more than one echo with a sufficient signal-to-noise ratio to be detectable; and if for each resolution cell there is more than one detectable echo from the multitude of beams formed by BFC, only the echo and BFC that obtain the maximum power or maximum signal-to-noise ratio will be considered valid. The US 2017 / 0 301 988 A1 This describes an antenna system and a method configured to compute calibration element voltage gain patterns as functions of a digital antenna model and a variety of complex beam