METHOD FOR DETERMINING A LONGITUDINAL SPEED OF A VEHICLE USING A RADAR SENSOR AND AN INSTALLATION ORIENTATION OF THE RADAR SENSOR WHEN DRIVING IN A CURVE

Abstract

A method for determining a longitudinal velocity of a vehicle using at least one radar sensor and an installation orientation of the at least one radar sensor during cornering, wherein the method comprises: determining at least one velocity vector of the at least one radar sensor during cornering of the vehicle, wherein the at least one velocity vector contains a longitudinal velocity component and a lateral velocity component of the at least one radar sensor, transmitting the at least one velocity vector to a module for estimating the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor, and estimating the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor at least on the basis of the at least one velocity vector transmitted to the module and via the module.

Claims

1. A method to determine a longitudinal velocity of a vehicle using at least one radar sensor and an installation orientation of the at least one radar sensor during cornering, the method comprising: determining at least one velocity vector of the at least one radar sensor during cornering of the vehicle, the at least one velocity vector having a longitudinal velocity component and a lateral velocity component of the at least one radar sensor; transmitting the at least one velocity vector to a module to estimate the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor; and estimating the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor at least on the basis of the at least one velocity vector transmitted to the module and via the module.

2. The method according to claim 1, wherein at least one velocity vector of each of at least two radar sensors is determined during cornering, the velocity vectors of the at least two radar sensors are transmitted to the module, and the estimation of the longitudinal velocity of the vehicle and each installation orientation of the at least two radar sensors is performed on the basis of the transmitted velocity vectors of the at least two radar sensors.

3. The method according to claim 1, wherein the estimation of the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor is performed substantially simultaneously.

4. The method according to claim 1, wherein the module is executed on the at least one radar sensor or a central processing unit of the vehicle.

5. The method according to claim 1, wherein the at least one velocity vector of the at least one radar sensor is determined on the basis of a yaw rate of the at least one radar sensor, a sideslip angle of the vehicle, a horizontal incidence angle of the at least one radar sensor, and/or a vertical incidence angle of the at least one radar sensor.

6. The method according to claim 1, wherein at least one inaccurate longitudinal velocity of the vehicle is determined by an odometry sensor of the vehicle and is transmitted to the module, wherein at least one scaling factor for the transmitted inaccurate longitudinal velocity of the vehicle is estimated substantially simultaneously with the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor.

7. The method according to claim 1, wherein at least one yaw rate sensor of the vehicle determines at least one inaccurate yaw rate of the vehicle, which is transmitted to the module, and wherein at least one scaling factor for the inaccurate yaw rate of the vehicle is estimated substantially simultaneously with the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor.

8. The method according to claim 1, wherein a yaw rate of the vehicle is estimated substantially simultaneously with the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor.

9. The method according to claim 1, wherein the installation orientation of the at least one radar sensor is determined by estimating a difference between a parameterized installation angle and a true installation angle.

10. The method according to claim 1, wherein the module is a Kalman filter.

11. The method according to claim 1, wherein a measurement vector with the longitudinal velocity component and the lateral velocity component of the at least one radar sensor is combined with a state vector to be estimated with the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor in the module to form a state-to-measurement equation.

12. The method according to claim 11, wherein, in the module, a state-to-measurement matrix is formed from the state-to-measurement equation, and the longitudinal velocity of the vehicle and the installation orientation of the at least one radar sensor are estimated by the module by the state-to-measurement matrix.

13. A radar system for a vehicle comprising: at least one radar sensor; and at least one module, wherein the radar system is set up to perform the method according to claim 1.

14. The radar system according to claim 13, wherein the at least one radar sensor is connected to the module by a proprietary or open data channel, or a CAN bus.

15. A vehicle comprising a radar system according to claim 13.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The present invention will become more fully understood from the detailed description given hereinbelow and the accompanying drawings which are given by way of illustration only, and thus, are not limitive of the present invention, and wherein:

[0029] FIG. 1 is a vehicle according to an exemplary embodiment of the invention during cornering;

[0030] FIG. 2 is a vehicle with a difference between a parameterized installation angle and a true installation angle;

[0031] FIG. 3 is a radar system according to an exemplary embodiment for the vehicles in FIGS. 1 and 2;

[0032] FIG. 4 is a measurement of a radar sensor of the radar system in FIG. 3;

[0033] FIG. 5 is a state-to-measurement matrix;

[0034] FIGS. 6a-6d are diagrams with parameters estimated using the radar system in FIG. 3 and true parameters; and

[0035] FIGS. 7a-7d are further diagrams with parameters estimated using the radar system in FIG. 3 and true parameters.

DETAILED DESCRIPTION

[0036] FIG. 1 shows a vehicle 1 according to an exemplary embodiment of the invention during cornering. Vehicle 1 is shown in a Cartesian x-y coordinate system.

[0037] Vehicle 1 has four radar sensors 2.1, 2.2, 2.3, 2.4. Alternatively, it can also have only one, two, three, or more than four radar sensors 2. Each of the radar sensors 2.1, 2.2, 2.3, 2.4 experiences a velocity v, which can be represented as a vector. The velocity vector {right arrow over (v)} has a longitudinal velocity component v.sub.sensor,x, a lateral velocity component v.sub.sensor,y, and a vertical velocity component v.sub.sensor,z. The velocity vector {right arrow over (v)} of each radar sensor 2.1, 2.2, 2.3, 2.4 thus takes the form

[00001]$\stackrel{.fwdarw.}{v}=\left[\begin{array}{c}{v}_{\mathrm{sensor},x}\\ v_{\mathrm{sensor},y}\\ v_{\mathrm{sensor},z}\end{array}\right],$

where it is assumed hereinafter that v.sub.sensor,z=0.

[0038] The depicted velocity vectors {right arrow over (v)} of radar sensors 2.1, 2.2, 2.3, 2.4 are labeled with 3.1, 3.2, 3.3, 3.4 here. Due to the different installation position of radar sensors 2.1, 2.2, 2.3, 2.4 on vehicle 1, radar sensors 2.1, 2.2, 2.3, 2.4 pass through a different radius during cornering with respect to an object or reflector 10, which radar sensors 2.1, 2.2, 2.3, 2.4 perceive as a point target. Accordingly, the velocity vectors 3.1, 3.2, 3.3, 3.4 differ from one another. The curve radii 4.1, 4.2, 4.3, 4.4 of radar sensors 2.1, 2.2, 2.3, 2.4 and curve radius 5 of vehicle 1 are drawn accordingly. Also, a sideslip angle δ of vehicle 1 as the angle enclosed between the longitudinal axis of vehicle 1 and the direction of movement of vehicle 1 during cornering is marked accordingly in FIG. 1.

[0039] Whereas, on the one hand, the longitudinal velocity v.sub.x of vehicle 1 is determined by the most accurate possible estimate using the method presented here, the installation orientation of radar sensors 2.1, 2.2, 2.3, 2.4 can also be determined simultaneously by an accurate estimate. The installation orientation is estimated here as an alignment error from a parameterized installation orientation or position.

[0040] The alignment error is shown in FIG. 2 as the misalignment angle α between velocity vector 3 of radar sensor 2 as measured by radar sensor 2 and the true velocity vector 6. The true velocity vector 6 can also be referred to as the ground truth velocity vector {right arrow over (v)}.

[0041] FIG. 3 now shows a radar system 100 for one of the vehicles 1 from FIGS. 1 and 2 according to an exemplary embodiment of the invention with one or more radar sensors 2.1 . . . 2.N, whose velocity vectors {right arrow over (v.sub.1)} . . . {right arrow over (v.sub.N)} are evaluated in a module 14, in this case, in the form of a Kalman filter. In this regard, module 14 can be executed on a computing unit of radar system 100. In the following it shall be assumed that N=2; therefore, two radar sensors 2.1, 2.2 are evaluated.

[0042] Radar sensors 2.1, 2.2 detect one or more reflectors when vehicle 1 is cornering. Radar sensors 2.1, 2.2 independently determine the velocity vectors {right arrow over (v.sub.1)}, {right arrow over (v.sub.2)} from this using a sensor velocity determination. The relative velocity v.sub.r, which is measured by radar sensors 2.1, 2.2, is used for the sensor velocity determination. This is a radial velocity. It can be represented as follows with the aid of the sideslip angle δ of vehicle 1, a horizontal incidence angle φ of radar sensors 2.1, 2.2, and a vertical incidence angle ε of radar sensors 2.1, 2.2 for each of radar sensors 2.1, 2.2:

v.sub.r=−v.sub.x cos(φ−δ)cos(ε)−v.sub.y sin(φ−δ)cos(ε).

[0043] Here, v.sub.x, is the longitudinal velocity component of the respective radar sensor 2.1, 2.2 in the direction of the longitudinal vehicle axis of vehicle 1 and v.sub.y is the lateral velocity component of the respective radar sensor 2.1, 2.2, as has already been explained above.

[0044] FIG. 4 shows by of example the measurement results of the relative velocity v.sub.r of one of the two radar sensors 2.1, 2.2 for various stationary reflectors or targets 10. To determine the velocity vector {right arrow over (v)}, the normal vector 7 to the plane spanned by the stationary targets 10 is calculated and differentiated in the direction of the coordinates cos φ and −sin φ. The negative reciprocal gradient now gives the estimated longitudinal velocity component v.sub.x and the estimated lateral velocity component v.sub.y of the respective radar sensor 2.1, 2.2.

[0045] The determined velocity vector {right arrow over (v)} of each of the radar sensors 2.1, 2.2 can thereby be modeled by the relationship {right arrow over (v)}={right arrow over (w)}×{right arrow over (R)}+{right arrow over (v.sub.trans)}, where {right arrow over (w)} is the rotational velocity vector with

[00002]$\stackrel{.fwdarw.}{w}=\left[\begin{array}{c}0\\ 0\\ w\end{array}\right],$

{right arrow over (R)} is a position vector from the vehicle rear axle to the parameterized position of the respective radar sensor 2.1, 2.2 which has been parameterized beforehand, and {right arrow over (v.sub.trans)} is a velocity vector of vehicle 1 measured at its rear axle. Thus, knowledge of the velocity vector {right arrow over (v)} of each of the radar sensors 2.1, 2.2 and the position vector {right arrow over (R)} allows determination of the yaw rate or yaw velocity w and the longitudinal velocity or longitudinal vehicle velocity v.sub.x of vehicle 1.

[0046] In this regard, the determined velocity vector {right arrow over (v)} of each of the radar sensors 2.1, 2.2 can be transmitted to radar sensors 2.1, 2.2 among each other by inter-sensor communication. In particular, however, these are transmitted to module 14. An open or proprietary data channel, in particular the CAN bus 13, can be used for this purpose. Further, an odometry sensor 11 and a yaw rate sensor 12 transmit via CAN bus 13 an inaccurate longitudinal velocity v.sub.CAN and an inaccurate yaw rate w.sub.CAN, which are inaccurate in the sense that they do not correspond to ground truth or are true, but are beset with measurement errors.

[0047] In module 14, the measured values transmitted to it are combined into a measurement vector z, which can be represented as

[00003]$z=\left[\begin{array}{c}{v}_{x1}\\ v_{y1}\\ v_{x2}\\ v_{y2}\\ v_{\mathrm{CAN}}\\ w_{\mathrm{CAN}}\end{array}\right].$

The velocity components v.sub.x1, v.sub.x2, v.sub.y2 here are lateral velocity components and longitudinal velocity components of the two radar sensors 2.1, 2.2.

[0048] The result sought is a state vector to be estimated x, which is estimated and output by module 14. This state vector can be represented as

[00004]$x=\left[\begin{array}{c}{v}_x\\ {\stackrel{.}{v}}_x\\ w\\ \stackrel{.}{w}\\ {\alpha }_1\\ {\alpha }_2\\ \beta \\ \gamma \end{array}\right],$

where v.sub.x indicates the longitudinal velocity of vehicle 1, {dot over (v)}.sub.x indicates the longitudinal acceleration of vehicle 1, w indicates the yaw velocity or yaw rate of vehicle 1, {dot over (w)} indicates the yaw acceleration of vehicle 1, α.sub.1, α.sub.2 indicate the misalignment angles of radar sensors 2.1, 2.2, β is a scaling factor or correction factor for the inaccurate longitudinal velocity v.sub.CAN, and γ is a scaling factor or correction factor for an inaccurate yaw rate w.sub.CAN.

[0049] If radar sensors 2.1, 2.2 are mounted rotated in vehicle 1 due to tolerances, the velocity vector {right arrow over (v)} is rotated in the same way. The measured sensor velocity

[00005]$\left[\begin{array}{c}{v}_{x1}\\ v_{y1}\end{array}\right]=\left[\begin{array}{cc}\cos \left({\alpha }_1\right)& \sin \left({\alpha }_1\right)\\ -\sin \left({\alpha }_1\right)& \cos \left({\alpha }_1\right)\end{array}\right]\left[\begin{array}{c}{v}_x@\mathrm{Sensor}2.1\\ v_y@\mathrm{Sensor}2.1\end{array}\right]$

then corresponds to the rotated velocity

[00006]$\left[\begin{array}{c}{v}_x@\mathrm{Sensor}2.1\\ v_y@\mathrm{Sensor}2.1\end{array}\right]$

of vehicle 1 and can be summarized in a state-to-measurement equation h(x) as follows:

[00007]$z=\left[\begin{array}{c}{v}_{x1}\\ v_{y1}\\ v_{x2}\\ v_{y2}\\ v_{\mathrm{CAN}}\\ w_{\mathrm{CAN}}\end{array}\right]=h\left(x\right)=\left[\begin{array}{c}\cos \left({\alpha }_1\right)\left(v_x-\Delta y_1\omega \right)+\sin \left({\alpha }_1\right)\left(\Delta x_1\omega \right)\\ -\sin \left({\alpha }_1\right)\left(v_x-\Delta y_1\omega \right)+\cos \left({\alpha }_1\right)\left(\Delta x_1\omega \right)\\ \cos \left({\alpha }_2\right)\left(v_x-\Delta y_2\omega \right)+\sin \left({\alpha }_2\right)\left(\Delta x_2\omega \right)\\ -\sin \left({\alpha }_2\right)\left(v_x-\Delta y_2\omega \right)+\cos \left({\alpha }_2\right)\left(\Delta x_2\omega \right)\\ \beta v_x\\ \mathrm{\gamma \omega }\end{array}\right].$

[0050] Accordingly, the state-to-measurement matrix shown in FIG. 5 is obtained, the solution of which leads to the sought state vector x to be estimated.

[0051] FIGS. 6a through 6d and FIGS. 7a through 7d show, for the various parameters to be estimated, the result of the estimate, therefore, the estimated value, versus the ground truth, therefore, the true value. According to FIGS. 6a to 6d, it can be seen that the motion states were estimated for the entire time duration in an unbiased manner. According to FIGS. 7a to 7d, it can be seen that the mechanical system states after the Kalman filter has settled were also estimated in an unbiased manner.

[0052] The invention being thus described, it will be obvious that the same may be varied in many ways. Such variations are not to be regarded as a departure from the spirit and scope of the invention, and all such modifications as would be obvious to one skilled in the art are to be included within the scope of the following claims.