阅读说明：本技术 借助雷达传感器估计点目标的横向速度或笛卡尔速度 (Estimating the lateral or Cartesian velocity of a point target by means of a radar sensor ) 是由 M·朔尔 于 2018-12-14 设计创作，主要内容包括：一种用于借助雷达传感器(尤其用于机动车的雷达传感器)根据在各个分析处理通道中获得的信号来估计雷达目标的速度((v<Sub>x</Sub>,v<Sub>y</Sub>))的方法,所述分析处理通道(i)相应于在一个方向上的相关的发送和接收天线(20,10,12)的不同中间天线位置(y<Sub>i</Sub>),该方法具有以下步骤：针对不同的分析处理通道(i)确定雷达目标的各个个体径向速度(v<Sub>r,i</Sub>),所述各个个体径向速度被分配给各个分析处理通道；基于雷达目标的所确定的个体径向速度(v<Sub>r,i</Sub>)来估计雷达目标的速度((v<Sub>x</Sub>,v<Sub>y</Sub>)),其中,该速度((v<Sub>x</Sub>,v<Sub>y</Sub>))包括关于切向速度(v<Sub>y</Sub>)的信息；还涉及一种用于实施该方法的雷达传感器。(Method for estimating the speed ((v) of a radar target from signals obtained in individual evaluation channels by means of a radar sensor, in particular for a motor vehicle x ,v y ) In the direction of the respective transmitting and receiving antenna (20, 10, 12), the evaluation channel (i) corresponding to different intermediate antenna positions (y) of the respective transmitting and receiving antenna (20, 10, 12) in one direction i ) The method comprises the following steps: determining individual radial velocities (v) of the radar target for different analysis processing channels (i) r,i ) The individual radial velocities are assigned to individual analysis processing channels; determining individual radial velocities (v) based on radar targets r,i ) To estimate the velocity ((v) of the radar target x ,v y ) Wherein the velocity ((v)) x ,v y ) Involving a tangential velocity (v) y ) The information of (a); it also relates to a radar sensor for implementing the method.)

1. A method for estimating the velocity ((v) of a radar target from signals obtained in individual evaluation channels (i) by means of a radar sensor, in particular for a motor vehicle_x,v_y) In the direction of the respective transmitting and receiving antenna (22, 10, 12), the evaluation channel corresponding to different intermediate antenna positions (y) of the respective transmitting and receiving antenna (22, 10, 12) in one direction_i) The method has the following steps:

determining individual radial velocities (v) of the radar target for different analysis processing channels (i)_r,i) Said respective individual radial velocities being assigned toEach analysis processing channel;

determining individual radial velocities (v) based on the radar target_r,i) To estimate the velocity ((v) of the radar target_x,v_y) Wherein the velocity ((v))_x,v_y) Involving a tangential velocity (v)_y) The information of (1).

2. Method according to claim 1, in which method the velocity ((v) of the radar target is estimated_x,v_y) In the step of (d), the speed ((v))_x,v_y) Comprising a speed (v) in relation to the direction of advance relative to the radar sensor_x) And tangential velocity (v)_y) The information of (1).

3. Method according to claim 1 or 2, in which method the velocity ((v) of the radar target is estimated_x,v_y) In the step of (a), estimating a Cartesian velocity ((v) of the radar target)_x,v_y))。

4. The method according to any of the preceding claims, having the steps of:

determining individual longitudinal and transverse angles (theta) of the radar target for each analysis processing channel (i)_i) The individual longitudinal and transverse angles are distributed to the analysis processing channels;

wherein the velocity ((v) of the radar target is estimated_x,v_y) Based on the determined individual radial velocity (v) of the radar target_r,i) And based on the determined individual longitudinal and transverse angles (theta) of the radar target_i) To estimate the velocity ((v) of the radar target_x,v_y))。

5. Method according to any one of the preceding claims, in which method the respective individual radial velocity (v) of the radar target is determined for the respective analysis processing channel (i)_r,i) Based on the analysis processing channel (i)To determine said respective individual radial velocities (v) from respective frequency positions (fa (i) of said signals (a) and (b)_r,i)。

6. Method according to any one of the preceding claims, in which method the respective individual radial velocity (v) of the radar target is determined for the respective analysis processing channel (i)_r,i) In the step (b), the respective individual radial velocities (v) are determined from the remaining frequency positions (f) taking into account the respective frequency position shifts (Δ fa (i))_r,i) Wherein each frequency position offset (Δ fa (i)) for each analysis processing channel (i) corresponds to the radar target to each corresponding intermediate antenna position ((0, y)_i) Distance difference of).

7. The method according to claim 6, having the steps of:

estimating the longitudinal and lateral angle (theta) of the radar target from the amplitude and/or phase relationship between the signals of the individual analysis processing channels corresponding to different intermediate antenna positions (y) of the associated transmitting and receiving antennas (22, 10, 12) in the mentioned direction_i)；

Determining, from the estimated longitudinal and transverse angles (theta), respective frequency position shifts (Δ fa (i)) of the signals in the respective evaluation channels (i), which correspond to the radar target to the respective intermediate antenna positions ((0, y)_i) Distance difference of);

wherein respective individual radial velocities (v) of the radar target are determined for the respective analysis processing channels (i)_r,i) In the step (a), the respective individual radial velocity (v) is determined from the remaining frequency position of the signal in the evaluation channel (i) taking into account the respective determined frequency position offset (Δ fa (i))_r,i)。

8. Method according to claim 4, in which method the respective individual longitudinal and transverse angles (θ) of the radar target are determined_i) Is/are as followsIn step (ii), the relative intermediate antenna position (y) of each analysis processing channel (i) is taken into account_i) Determining respective individual longitudinal and transverse angles (theta) of the radar target based on the estimated longitudinal and transverse angles (theta) of the radar target and the estimated distance (d) of the radar target_i)。

9. The method according to any of the preceding claims, having the steps of:

estimating the longitudinal and lateral angle (theta) of the radar target from the amplitude and/or phase relationship between the signals of the individual evaluation channels corresponding to different intermediate antenna positions ((0, y) of the associated transmitting and receiving antennas (22, 10, 12) in the direction in question_i) Wherein the signals of the individual evaluation channels (i) are evaluated at the individual frequency positions taking into account the individual first frequency position shifts (Δ fa (i)) of the signals in the individual evaluation channels (i) which correspond to the radar target to the individual respective intermediate antenna positions ((0, y)) and/or taking into account the individual second frequency position shifts (Δ fb (i)) of the signals in the individual evaluation channels (i)_i) A second frequency position shift (Δ fb (i)) corresponding to an individual radial velocity (v) of the radar target for each analysis processing channel (i)_r,i) The difference of (a).

10. A radar sensor, in particular for a motor vehicle, having an antenna arrangement with a plurality of antennas (10, 12) arranged at different positions in one direction, and having a control and evaluation device (30) which is designed to carry out the method according to one of claims 1 to 9.

Technical Field

The invention relates to a method for a radar sensor, in particular for a motor vehicle, and to a radar sensor, in particular for a motor vehicle.

Background

Conventional radar sensors detect radar targets in polar coordinates. For example, radial distance, radial relative velocity, and angle in azimuth and/or elevation are measured. The determination of the lateral (or tangential) or angular velocity can only be carried out by the change in the observed angle over time or using complex object models for extended radar objects.

Radar sensors are used in motor vehicles, for example, to measure the distance, relative speed and azimuth angle to a vehicle or other radar target located in front of the own vehicle. The plurality of antennas are then arranged, for example, at a distance from one another in a horizontal line, so that different azimuth angles of the located radar target result in differences in the propagation lengths of the radar signals, which the radar signals travel from the radar target until the respective antenna. Such propagation length differences result in corresponding differences in the amplitude and phase of the signals received by the antennas and analytically processed in the corresponding analysis processing channels. The following is fully utilized for angle estimation: the amplitude and phase relationship of the signals obtained by the different receiving antennas depends in a characteristic way on the angle of the radar target. The angle of incidence of the radar signal and thus the azimuth angle of the radar target can then be determined by comparing the (complex) amplitudes received in the different channels with the respective amplitudes in the antenna diagram. In a corresponding manner, the elevation angle of the radar target can also be estimated by means of antennas arranged vertically above one another.

Disclosure of Invention

In the process of further improving the performance of the radar sensor, d and v estimation can be carried out with higher resolution. An increase in the available sensor size (i.e. the size or aperture of the antenna arrangement) will also enable an increase in the accuracy of the angle estimation and an improvement in the angle separation. In the case of FMCW (frequency modulated continuous wave) measurement methods with linear frequency ramps and of the evaluation of the received signal by means of discrete fourier transforms, in particular FFT (fast fourier transform), the width of the fourier transformed distance window (entrenugsbin) corresponds to the distance difference Δ r, where Δ r is c/(2F), where c is the speed of light and F is the frequency offset of the linear frequency ramp of the FMCW transmitted signal. This distance difference is also referred to as distance resolution.

Thus, "distance resolution" is understood to mean the following minimum distance difference: in this minimum distance difference (for the same relative speed), the two measured values of the distance to the radar sensor can still be mapped onto separate windows for a given operating mode of the radar sensor. In performing the FFT, the range resolution corresponds to the spacing of two range windows in the FFT, i.e., the width of the range window. Here and in the following, the terms "distance resolution" and "width of the distance window" may be used synonymously. In contrast, the term "distance separability" is understood to mean twice the width of the distance window. If the bandwidth of the radar sensor is increased, for example, when the frequency offset of the transmission signal is F — 2GHz, a range resolution of Δ r — 7.5cm can be achieved. If the aperture or (in the case of a MIMO (Multiple Input Multiple Output) radar sensor) virtual aperture is increased to a value of similar order of magnitude at the same time, takeDepending on the angle of the radar target, the propagation length difference between the received signals of the individual antennas or evaluation channels can already be so great that, in the fourier spectrum of the evaluation channels, (depending on the evaluation channel) information about the amplitude and/or phase of the received signals is contained not only in the frequency window determined by the d, v estimate of the detected radar target but also in one or more adjacent frequency windows. If the distance of the radar target directly in front is 5m and the offset between the intermediate antenna position and the outer antenna position is, for example, 40mm, approximately 0.5 results^°The angle difference of (a). If the lateral velocity of the radar target is, for example, 2.7m/s (10km/h), no relative velocity is seen at the middle antenna position, whereas a radial relative velocity of 0.025m/s is seen at the outer antenna position or-0.025 m/s is seen on the other side. The window size of the FFT is 0.1m/s (measured velocity resolution) corresponding to the frequency location offset of the-1/4 window, 0 window, or +1/4 window in the respective three analysis processing channels.

The object of the present invention is to provide a method and a radar sensor, by means of which a fast estimation of the tangential velocity of a radar target can be achieved.

According to the invention, this object is achieved by a method for estimating the velocity of a radar target by means of a radar sensor (in particular for a motor vehicle) from signals obtained in individual evaluation channels which correspond to different intermediate antenna positions of the associated transmitting and receiving antenna in one direction, comprising the following steps: determining individual radial velocities of the radar target for different analysis processing channels, the individual radial velocities being assigned to the respective analysis processing channels; and estimating a velocity of the radar target based on the determined individual radial velocity of the radar target, wherein the velocity comprises information on the tangential velocity. I.e. the velocity comprises information describing the tangential velocity. The speed may be, inter alia, tangential speed, lateral speed or cartesian speed. Cartesian velocities are understood to be velocities which are illustrated in an orthogonal coordinate system. The cartesian speed here preferably includes a speed in the forward direction and a tangential speed with respect to the radar sensor. For a radar sensor oriented directly in the forward direction, the tangential velocity corresponds to the lateral velocity.

This enables the tangential velocity of the radar target to be estimated based on a single measurement with a single modulation period of the transmitted signal, in particular for a transmitted signal with a frequency modulation pattern in the form of a frequency ramp or a fast chirp sequence, for example. In particular, the tangential velocity can be estimated from the respective frequency spectrum on the basis of a single measurement of the radial velocity in the respective analysis processing channel.

Thus, the measurement of tangential or cartesian velocity can be performed very quickly. In particular, the measurement can be carried out directly for a single radar target and as a function of the detection of the radar target within a single modulation period or within a frequency ramp of the transmission signal. Thereby, a measurement of the tangential or cartesian speed can be achieved even in case of objects that are visible only within one or a few periods and are therefore difficult to handle by complex object models, such as fast laterally moving objects or buildings at the edge of a road.

This enables a significantly faster reaction of the driver assistance system or of the system for automated driving, in particular in the case of radar targets at relatively short distances. This is particularly advantageous for protecting vulnerable traffic participants (VRU), such as pedestrians.

A radar target is understood to be a reflection center assigned only a single position. The term "radar target" is used synonymously with "point target".

In the method, a transmission signal of the radar sensor is transmitted by at least one antenna and a signal is received by at least one antenna. In an advantageous embodiment, the transmitted signal is frequency-modulated in the form of a ramp. In an advantageous embodiment, the radar sensor is an FMCW radar sensor.

Here and in the following, the estimated or determined variables which relate to the radar sensor as a whole are also referred to as "global" variables, while the variables which relate to the individual evaluation channels and the associated intermediate antenna positions of the transmitting and receiving antennas are referred to as "individual" variables. Thus, for example, the tangential velocity to be estimated may be referred to as global tangential velocity.

In addition, the object is achieved by a radar sensor, in particular for a motor vehicle, having an antenna arrangement with a plurality of antennas arranged at different positions in one direction, and having a control and evaluation device which is designed to carry out the method.

Advantageous embodiments and embodiments of the invention are specified in the dependent claims.

In an advantageous embodiment, in the step of estimating the tangential velocity or cartesian velocity of the radar target, the tangential velocity or cartesian velocity of the radar target is estimated based on the following relationship:

wherein i ═ 1, …, l denotes the respective analytical processing channel, where v_r,iIs the determined individual radial velocity, θ, of the radar target_iIs the longitudinal and transverse angle (Aspekwinkel) of the radar target assigned to each analysis processing channel, and (v)_x,v_y) Is the Cartesian velocity of the radar target, where v_yIs the tangential velocity, and v_xIs the speed of the radar target in the forward direction relative to the radar sensor.

In an advantageous embodiment, in the step of determining individual radial velocities of the radar target for the individual evaluation channels, the individual radial velocities are determined on the basis of the individual frequency positions of the signals in the evaluation channels.

In an advantageous embodiment, a discrete spectrum is calculated in each evaluation channel by means of a fourier transformation, and the respective frequency position of the signal in the relevant evaluation channel is determined with a resolution which is finer than the frequency spacing of the support points of the discrete spectrum. The resolution of the frequency position is understood here as the following variable: the parameter describes the fineness of possible values of the frequency positions that can be obtained in the determination. The support points of the discrete spectrum may also be referred to as frequency points or frequency bins.

The respective frequency positions may be determined, for example, by interpolation of the spectrum, search of the peak maxima of the signal or by matching (fitting) the frequency parameters of the model function to the signal in the spectrum.

Drawings

Embodiments are further explained below with reference to the drawings. The figures show:

fig. 1 shows a block diagram of a radar sensor for a motor vehicle according to the invention;

FIG. 2 shows a schematic diagram of frequency bins of Fourier spectra of various analysis processing channels;

FIG. 3 illustrates the relationship between antenna position and radar target;

fig. 4 shows a block diagram for illustrating the method according to the invention.

Detailed Description

The radar sensor shown in fig. 1 has a plurality of receiving antennas or antenna elements 10, 12 on a common substrate 18. The radar sensor is installed in the motor vehicle in such a way that several of the antennas 10, 12 are located at the same height next to one another at a horizontal position yi, i being 0, …, k. Fig. 1 symbolically shows a radar beam, which is received by an antenna at a corresponding azimuth angle θ i (longitudinal angle and transverse angle).

The high-frequency part 20 for steering the transmitting antenna 22 comprises a local oscillator 24, which generates a radar signal to be transmitted. The radar echoes received by the antennas 10, 12 are each supplied to a mixer 28 where they are mixed with the transmit signal supplied by the oscillator 24. In this way, for each of the antennas 10, 12, a baseband or intermediate frequency signal Z0, Z1.., Zi.,. is obtained, which is supplied to the electronic control and evaluation unit 30.

The control and evaluation unit 30 contains a control section 32 which controls the function of the oscillator 24. In the example shown, the radar sensor relates to an FMCW radar, that is to say the frequency of the transmitted signal provided by the oscillator 24 is modulated periodically in the form of a sequence of rising and/or falling frequency ramps.

Furthermore, the control and evaluation device 30 comprises an evaluation part having an analog/digital converter 34 with k channels, which digitizes the intermediate frequency signals Z0 to Zk obtained from the k antennas 10, 12 and records them separately over the duration of the individual frequency ramps. The time signal thus obtained is then converted channel by channel into the corresponding frequency spectrum by means of a Fast Fourier Transform (FFT) in a transform stage 36. In these frequency spectra, each radar target is present in the form of a peak whose frequency position depends on the signal propagation time from the radar sensor to the radar target and back to the radar sensor and (due to the doppler effect) on the relative velocity of the radar target. Then, from the frequency positions of the two peaks (which have been obtained for the same radar target but have different slopes in terms of frequency ramp, for example a rising ramp and a falling ramp), the distance d and the relative speed v of the relevant radar target can be calculated in a known manner. The estimated distance d may be referred to as the global distance of the radar target, unlike the individual distances di assigned to the individual analysis processing channels.

As schematically shown in fig. 1 in terms of radar beams, the different positions of the antennas 10, 12 result in: radar beams transmitted by the same antenna, reflected at a radar target and then received by different antennas travel different lengths of propagation and therefore have a phase difference that depends on the azimuth angle θ of the radar target. The corresponding intermediate frequency signals Z0 to Zk also have corresponding phase differences. The amplitude (magnitude) of the received signal also varies from antenna to antenna and is also dependent on the azimuth angle θ.

The velocity estimation means 38 estimates from the signals in the frequency spectrum of the analysis processing channel the cartesian velocity of the radar target, which has a component vx (relative to the velocity of the radar sensor in the forward direction) and a tangential velocity vy. As will be further explained later. In addition, the angle estimation device 40 estimates the azimuth of the radar target from these signals.

In the case of high bandwidth (large frequency offset corresponding to FMCW modulation) and large size of the antenna arrangement, the complex amplitudes in the individual receive channels are contained at different frequency positions fa (i) in the spectrum of the received signal, depending on the azimuth angle θ of the radar target and the range d of the radar target. Fig. 2 schematically shows the frequency positions fa (i) of the signals (peaks of the frequency spectrum) obtained for the radar target in the evaluation channel i, wherein successive frequency windows of the fourier spectrum are shown in the direction of increasing frequency f.

Fig. 3 shows the relationship with a point target at coordinates (x, y) as a radar target, which has a cartesian velocity (vx, vy), for the antenna position of the evaluation channel in top view, which is denoted by the index i at coordinates (0, yi). The distance of the point target is denoted by di, and the vertical and horizontal angles of the received radar signal are denoted by θ i. To simplify the illustration, it is assumed that the origin (0,0) represents the middle point of the antenna array and corresponds to the middle antenna position. In this example, vx is 0, which corresponds to the following case: in this case, the point target moves in the lateral direction just in front of the radar sensor. For simplicity, the figure shows the following: in this case, the relative radial velocity and the absolute radial velocity are zero with respect to the origin. The radial velocity vr is measured at the antenna position (0, y0) at the origin, 0 ═ 0. The radial velocity vr, i is measured at the antenna position (0, yi). This corresponds to the projection of the cartesian velocities (vx, vy) in the radial direction of the antenna position and is therefore dependent on the longitudinal and transverse angles θ i of the radar target at the antenna position.

The relationship between the aspect angle θ i of the analysis processing channel i, the cartesian velocities (vx, vy) of the point target, and the individual radial velocities, which are derived from the spectral estimation in each analysis processing channel i, is illustrated by equation (1). With the corresponding definition of the matrix M, the equation can be rewritten as:

v _r＝Mv _xy

then, the data in each channel can be processed by each analysis according to the least squares method (KQ estimation)Individual radial velocity v_r,iEstimating Cartesian velocityv _xyThe vector of (a) is:

the larger the angular difference and the more accurately the relative velocity can be determined, the better the corresponding estimate. The least squares estimate may be computed numerically, for example, by means of pseudo-inverse, Singular Value Decomposition (SVD) or QR Decomposition.

The distance di of the analysis processing channel "seen" by the radar sensor due to the propagation length difference depends on the antenna configuration. Therefore, in a bistatic system or a MIMO system, the effects of paths (distance or propagation time) from a transmission antenna to a target and from the target to a reception antenna are added and averaged. Thus, the distance di is the average distance of the outbound and inbound paths over the average propagation time of the signal. Accordingly, the intermediate antenna positions of the transmit and receive antennas are considered.

The control and evaluation device 30 is designed to carry out a method for estimating the cartesian velocity of a radar target, which is illustrated by way of example in fig. 4 and is implemented in the velocity estimation device 38.

In step S10, the frequency position fa (i) (peak position) of the signal of the channel i is determined at high resolution by interpolating the frequency spectrum.

In the first angle estimation in step S12, for each located object, i.e., each radar target (each peak in the frequency spectrum), the complex amplitudes obtained in the I reception channels are compared with the antenna diagram to thereby estimate the azimuth angle θ of the radar target. In this case, the vectors of the complex amplitudes are evaluated at the same frequency positions fref in the frequency spectra of the channels. The estimated longitudinal-lateral angle θ may be referred to as a "global" longitudinal-lateral angle, as opposed to the individual longitudinal-lateral angles of the individual analysis processing channels.

In step S14, a distance-dependent window shift (shift Δ fa (i) in frequency position) of the channel is determined from the azimuth angle θ and the antenna position yi. The frequency location offset may be referred to as a distance-dependent frequency location offset. The frequency position offset may also be referred to as a frequency position correction. As mentioned above, the frequency location offset is determined by the high distance resolution of the measurement. Frequency position shifts occur with both vertical and horizontal angles. The frequency position offset may also be zero depending on the longitudinal and lateral angles.

In step S16, the offset Δ fa (i) is subtracted from the frequency position fa (i), and in step S18 the remaining frequency position fa (i) - Δ fa (i) is analytically processed in order to determine from this the individual radial velocity vr, i of the channel i. This is according to the FMCW equationThe process is carried out. Here, k is the window position corresponding to the remaining frequency position, c is the speed of light, d is the distance, F is the frequency offset of the ramp, F0 is the intermediate frequency, vr is the radial velocity, and T is the duration of the ramp. The frequency positions analyzed are the remaining frequency positions in consideration of the frequency position shift.

In step S20, the individual longitudinal and transverse angles θ i are determined from the global distance d and the estimated azimuth angle θ (e.g. from the geometrical relationship of these quantities), for example taking into account the antenna position yi. The individual longitudinal and transverse angles can also be calculated from the cartesian coordinates of the radar target and the intermediate antenna position, for example.

In step S22, the cartesian velocity of the radar target is estimated based on equations (1) and (2). These equations can be converted into radial and tangential velocities (or angular velocities) with respect to the origin, for example, for the output of the radar sensor.

In an optional step S24, the azimuth angle θ is estimated by the angle estimation means 40 in an improved second angle estimation, wherein the vectors of complex amplitudes are processed analytically at respective frequency positions in the respective frequency spectrum of the channel taking into account the frequency offset Δ fa (i) and/or the respective velocity-dependent frequency offset Δ fb (i) (derived from the azimuth angle θ and the antenna position yi). This particularly relates to the step of second estimation of the longitudinal and lateral angles of the radar target, which is also performed in addition to the above-described (first) estimation step S12 of the longitudinal and lateral angles.

The method may in particular be an iterative method in which the longitudinal-transverse angle-dependent estimation steps S14, S16, S18, S20, S22 are carried out again on the basis of the second estimation of the longitudinal-transverse angle.

As can be seen in the figures, these steps may be performed in parallel with each other or in other orders.

In an advantageous embodiment, the difference in the individual radial velocities is taken into account if the distance of the radar target is below a distance threshold. Since at very large distances this effect is too small to distinguish individual radial velocities. The distance threshold may be selected based on a test measurement or a theoretical speed resolution of the radar sensor.

In the embodiments described herein, operation is by way of a bistatic antenna scheme. Alternatively, however, a monostatic antenna scheme may be used in which the same antenna(s) are used for transmission and reception.

The method can be used particularly advantageously in FMCW radar sensors which operate with the aid of a so-called fast chirp sequence. In this case, a plurality of frequency ramps (chirp) are traversed in a rapid sequence, which have a large slope and only a relatively short duration. The two-dimensional spectrum on each analysis processing channel i is derived by a 2D-FFT performed on a single ramp and on a sequence of ramps. Thus, the frequency position fa (i) and the frequency offsets Δ fa (i) and Δ fb (i) are typically two-dimensional vectors.