EP3837147A1 - Driving assistance method for a vehicle, control unit, driving assistance system and vehicle - Google Patents

Abstract

The present invention relates to a driving assistance method (T) for a vehicle (1), in which (i) an instantaneous speed (v) of the vehicle (1) and an instantaneous yaw rate (ω) of the vehicle (1) are determined (T1), and (ii) on the basis of the determined instantaneous speed (v) and the determined instantaneous yaw rate (ω) of the vehicle (1), a process of self-locating the vehicle (1) is carried out (T2), wherein, to do this, an instantaneous wheel circumference speed (V) of one or more wheels (4) of the vehicle (1) is directly sensed, evaluated and used as the basis for determining (T1) the instantaneous speed (v) and the instantaneous yaw rate (ω) of the vehicle (1).

Description

 description

title

 Driver assistance method for a vehicle, control unit, driver assistance system and vehicle

State of the art

The present invention relates to a driver assistance method for a vehicle, a control unit for a driver assistance system of a vehicle and a driver assistance system and a vehicle as such.

Driver assistance processes and systems for vehicles are enjoying increasing popularity. An essential aspect of such methods and systems is the ability of the vehicle to self-locate, for example with regard to a positioning and orientation of the vehicle in relation to a roadway, surrounding objects, the surrounding traffic and in particular in relation to a parking space or the like.

In conventional systems and methods, the

Real-time requirements of such driver assistance methods and systems rely on signals from so-called wheel impulse counters or WICs (English abbreviation "wheel impulse counter"). However, these deliver in

Low speed range, for example when parking at less than 5 km / h, values with often insufficient measuring accuracy.

Disclosure of the invention

The driver assistance method according to the invention with the features of

In contrast, claim 1 has the advantage that also in

Low-speed range measurement values are provided with sufficient measurement accuracy for self-localization of an underlying vehicle. This is achieved according to the invention with the features of claim 1 thereby achieved that a driver assistance method for a vehicle is created in which (i) a current speed of the vehicle and a current yaw rate of the vehicle are determined and in which (ii) on the basis of the determined current speed and the determined current yaw rate of the vehicle a process of self-localization of the vehicle is carried out, with a current one

The circumferential wheel speed of one or more wheels of the vehicle is recorded directly, evaluated and used to determine the current speed and the current yaw rate of the vehicle. The measurement of the wheel circumference speed of one or more wheels can be carried out with a higher measuring accuracy than the commonly used measurement values of wheel pulse counters. According to the invention, this also results in a higher overall accuracy in the self-localization of the underlying vehicle.

The subclaims show preferred developments of the invention.

In a preferred embodiment of the invention

Driver assistance procedure becomes a respective current one

Wheel circumferential speed detected and provided by a wheel circumferential speed sensor.

It is particularly advantageous if, in one embodiment of the

According to the driver assistance method according to the invention, a time delay of a detected instantaneous wheel circumferential speed is compensated for by temporal extrapolation of measured values at an earlier measurement time to a current evaluation time. This measure makes it possible to compensate for the fact that the measurement data is not always up-to-date or sufficient for many wheel circumference speed sensors due to a time delay, so that advantages of real-time use are preferred.

A particularly simple configuration of the compensation can be achieved if it (i) by time integration from the earlier measurement time to the current evaluation time, (ii) on the basis of one or more measurement values for an instantaneous acceleration of the vehicle and / or on the basis of a single-track model of the Vehicle. Another advantageous exemplary embodiment of the driver assistance method according to the invention results in a particularly advantageous and rapid feasibility of the individual with regard to an implementation

Processing steps if a process of Moore's pseudo-inversion is provided when determining the current speed and the current yaw rate of the vehicle and on the determined

Wheel peripheral speeds is applied. These measures provide an elegant and at the same time reliable

Relationship between different sizes, the state of the

Exploited vehicle, namely the relationship between the current speed and the current yaw rate of the vehicle on the one hand, which are to be determined, and the measurable values of the

Wheel speed.

This can be done in particular in that when and for determining the current speed and the current yaw rate of the vehicle, a Moore's pseudo inverse of a transformation matrix between a state of the underlying vehicle and a vector formed by the individually determined wheel circumferential speeds is generated and for providing the current speed and the current yaw rate of the vehicle on that determined by the individual

Wheel circumferential speeds formed vector is applied. The state of the vehicle describes in particular the current one

Speed and the current yaw rate of the vehicle. Optionally, the distance traveled by a wheel contact point, which can also be referred to as the current distance traveled by a wheel contact point, to one or more wheels of the vehicle can be taken into account.

An advantage of Moore's pseudo-inverses is their analytically specifiable representation and structure and their property of being able to inherently minimize or optimize an underlying norm without numerical methods or iteration.

A particularly high degree of accuracy in self-localization can be achieved if, in accordance with another advantageous development of the driver assistance method according to the invention, in addition to

Wheel circumferential speed a currently covered distance of the Contact point of one or more wheels of the vehicle is detected, evaluated and used to determine the current speed, the current yaw rate, a current position and / or a current orientation of the vehicle.

In this case, a respective currently covered distance of a respective contact point of a wheel of the vehicle can advantageously be recorded and made available via a respective wheel pulse counter, taking into account a value provided for the circumference of the wheel. This measure can therefore be provided in an ESP system anyway

Measured values are used according to the invention.

The accuracy can be further increased if, according to another advantageous embodiment of the invention

Driver assistance procedure a respective detected instantaneous

Wheel circumference speed of one or more wheels of the vehicle and a respective currently covered distance of the contact point of one or more wheels of the vehicle, a Bayesian filter and in particular an extended Kalman filter for evaluating, for plausibility checking and / or for determining a current position and / or current orientation of the vehicle be fed.

According to a further aspect of the present invention, a

Control unit for a driver assistance system of a vehicle is proposed, the control unit being set up, an inventive one

Initiate, execute, control and / or regulate driver assistance procedures.

The present invention furthermore also relates to a driver assistance system for a vehicle as such, the driver assistance system being set up to initiate, execute, control and / or regulate a driver assistance method according to the invention, and / or the driver assistance system being a control unit designed according to the invention or an active connection to such a control unit designed according to the invention.

The driver assistance method according to the invention and the driver assistance system according to the invention can be understood and implemented purely in terms of hardware, for example also as a device for controlling the operation of a vehicle. For example, implementation as an ASIC is conceivable. Alternatively, a purely procedural implementation, for example in connection with a computer implementation of the driver assistance method and driver assistance system according to the invention, is conceivable, preferably in connection with or as a method for controlling the operation of a vehicle. Combined or mixed systems are also conceivable, in which partial aspects of the invention are hardware and / or software or

are or will be implemented in terms of process engineering.

The present invention also provides a vehicle as such. The vehicle according to the invention is set up to be used with a driving assistance method designed according to the invention. Alternatively or additionally, the proposed vehicle is designed with a driver assistance system according to the invention.

Brief description of the figures

Embodiments of the invention will be described in detail with reference to the accompanying figures.

Figure 1 shows a kind of a schematic block diagram

 Vehicle designed according to the invention, in which an embodiment of the invention

 Driver assistance procedure is used.

FIG. 2 shows a flow diagram of an embodiment of a

 Driver assistance system according to the invention in the manner of a driver assistance method.

Figure 3 explains schematically various at a

 Embodiment of the driver assistance system or driver assistance method according to the invention used parameters in relation to the state of an underlying vehicle.

FIGS. 4 to 7 show graphs to explain the mode of operation of

 Embodiments of the invention

Driver assistance system and method. Preferred embodiments of the invention

In the following, with reference to FIGS. 1 to 7

Embodiments of the invention and the technical background described in detail. Identical and equivalent elements and components which have the same or equivalent effect are denoted by the same reference symbols. The detailed description of the designated elements and components is not reproduced in every case.

The features and further properties shown can be isolated from one another in any form and can be combined with one another as desired without leaving the essence of the invention.

Figure 1 shows a kind of a schematic block diagram

Vehicle 1 designed according to the invention, which uses an embodiment of the driver assistance system 100 according to the invention and / or an embodiment of the driver assistance method T according to the invention.

The vehicle 1 according to the invention is shown schematically with a

Body 2, wheels 4, a drive 30 with drive train 31 and a system 40 for steering and braking with a steering and / or brake train 41.

A control unit designed according to the invention is also provided

50 for the underlying driver assistance system 100 according to the invention, which can also be designed, for example, as part of a vehicle or engine control and which via a control and / or detection line

51 establishes a connection with the drive 30 and the system 40 for braking and steering.

Furthermore, the control unit 50 is also connected to sensors 10 via the control and / or detection line 51, namely a first sensor 10-1 in the form of a sensor for the circumferential wheel speed and a second sensor 10-2 in the form of a wheel pulse counter.

In operation of the vehicle 1, corresponding measurement signals with respect to the sensors 10, 10-1, 10-2 of the control unit 50 are Wheel circumferential speed or with respect to the wheel speed or in terms of the wheel rotation angle supplied and using a

Bayesian filter and in particular a Kalman filter 20 one

Subsequent processing and analysis subjected to values for the

Vehicle speed v and for the yaw rate w and on the other hand to generate values for the distance S covered by the respective contact point of a wheel 4, and to provide a position and / or orientation of the vehicle 1 in the environment with a high degree of reliability even at low speeds of the Provide vehicle 1.

FIG. 2 shows a flow diagram of an embodiment of an as

Driver assistance method T designed according to the invention

Driver assistance system 100.

The core of the embodiment of the driver assistance method T according to FIG. 2 (i) is subdivided into a step T1 for determining the speed v and yaw rate w of a vehicle 1, (ii) a step T2 of the self-localization of the vehicle 1 on the basis of the provided

Measured values and / or data derived therefrom, (iii) a step T3 of the

Detecting and evaluating the vehicle environment and (iv) a step T4 of controlling a vehicle assembly based on the self-localization and the evaluation of the vehicle environment.

Step T 1 of determining speed v and yaw rate "des

Vehicle 1 is divided into a number of sub-steps T1-1 to T1-3.

In the first sub-step T1-1, the wheel circumferential speed V is detected in relation to one or more wheels 4, in particular by direct measurement with a corresponding sensor 10-1 for the wheel circumferential speed V of an assigned wheel 4.

In the second sub-step T1 -2, any time delay that occurs when the wheel circumferential speed V is detected is compensated for, for example by temporal extrapolation into the future by means of time integration, as explained in detail below in connection with a preferred embodiment of the present invention. Finally, in the third sub-step T1 -3, speed v and yaw rate w of the underlying vehicle 1 are generated and made available.

The step T2 of the self-localization of the vehicle 1 can also be subdivided into a series of sub-steps T2-1 to T2-3 in one embodiment of the present invention.

In a first sub-step T2-1, the currently traveled path S of a wheel contact point to one or more wheels 4 is recorded, in particular by direct measurement and / or in connection with measurement data read out from a WIC sensor 10-2 on the basis of a wheel radius,

Wheel diameter and / or wheel circumference of a respectively assigned wheel 4 of the vehicle 1.

In a second sub-step T2-2, a Bayesian filter and in particular a Kalman filter 20 are applied to the acquired data, namely to the

Speed v and the yaw rate «of the underlying vehicle 1 on the one hand and on the data acquired on the currently covered distances S of the wheel contact points on the other hand.

This then results in a determination and / or plausibility check of the position and / or orientation of the underlying vehicle 1 in its surroundings in a further sub-step T2-3.

The data relating to the position and / or orientation of the vehicle 1 in its surroundings, which are generated with high reliability, are then used as a basis for evaluating the vehicle environment in step T3 and as a result of the control of at least one vehicle unit in step T4, for example in connection with the control a system 40 of steering and brake and / or a drive 30 of the vehicle 1.

These and other features and characteristics of the present invention are further illustrated by the following statements:

Precise vehicle localization An increased customer acceptance of automated parking systems leads to an increasing use of such systems. The user evaluates the performance of the overall system and the concept of

Self-localization plays a major role in this context.

Two critical and measurable aspects in the context of automated parking are (i) the presence or absence of curbs and (ii) the minimum required parking space for a given vehicle. The impact of these aspects can be reduced to improve the customer experience. However, to achieve such a goal, a more precise localization of the vehicle while parking is an essential requirement.

The present invention proposes a new method of using information that can be derived from ordinary ESP systems.

Through the procedure according to the invention, the performance is

Systems of automated driving and parking increased without the need for new or additional sensors and without the need to evaluate new and / or additional signals ESP systems.

Conventional self-localization algorithms use information at low speeds that is used in the ESP system

Wheel pulse counters (English: Wheel Impulse Counter or WIC) can be read out. Although the corresponding measured values are available with a known, fixed time delay, they are subjected to a comparatively high error for the evaluation of the vehicle speed and the yaw rate by quantization and are therefore inaccurate and therefore do not allow precise self-localization in a vehicle application, for example in automated driving or parking.

Conventionally, the more precise measured values from sensors for the wheel circumferential speeds or wheel rotational speeds (path length per unit of time) are not used.

This is due to, (A) that the measured values of the sensors for the wheel circumferential speed or wheel rotation speed below a certain threshold value of the detection time are not immediately available and

(B) that the measured values of the sensors for the wheel circumferential speed or wheel rotation speed are only available with a variable delay.

Both are based on related signal preprocessing and corresponding timeout conditions.

Estimation of vehicle speed and yaw rate

The following describes how accurate estimates for the

 Vehicle speed and for the yaw rate of the underlying vehicle 1 from the four available wheel circumferential speeds

V = (v ^FrL V ^FrR V ^RrL V ^RrR ) ^T GR ⁴ . _. , , , , ... ..

 V J or wheel speeds are generated.

According to the invention, vehicle 1 can generally have four-wheel steering. This means that, according to the invention, all four wheels 4 of vehicle 1 can be steerable.

In addition, in the method according to the invention and in the

Implementation as an algorithm real-time conversion can be achieved, although in one embodiment of the method according to the invention, a matrix inversion is included in the evaluation.

The dimension of the matrix to be inverted can be, for example, by

Use of an extended information filter compared to an extended Kalman filter or extended Kalman filter can be reduced to such an extent that the method according to the invention and that of the algorithm are less

Keep computing time real-time capable.

In addition, delay compensation can be introduced in other embodiments of the method according to the invention, so that In particular, measured values from sensors for wheel circumference speed or wheel rotation speed can be used.

Speed and yaw rate of a vehicle from the wheel speeds If the yaw rate w - i.e. the change in the yaw angle over time

Vehicle 1 - and the speed v of the vehicle 1 are given and as

State ^X ( ^uw ) ^ ^ can be shown, the

Wheel speeds V, also called wheel speeds or

Wheel circumferential speeds are designated (all terms are used synonymously) using a suitable transformation matrix

H {u) GR ^{4x 2} d _{U | -c} h the following expression can be represented:

V = H (u) z: (1.1.1)

H (u) = (cos (d 7) r dr sin ^x cos ^y · d (1.1,2)

Only values for measurements of the wheel circumferential speeds or wheel speeds V are given, but not the state ^c as such.

It is therefore desirable to find the best estimate for the state ^c , which minimizes the value of an underlying standard chosen as the quality measure of the evaluation - here, for example, the minimum standard - min || V— H {u) x \ (1.1.4)

This problem can be solved by determining and using the pseudoinverse P ^ ^nv (( ^u )) assigned to the matrix ^H ( ^u ) (instead of the inverses actually required). This is either the unique least-squares solution or the least-squares solution of the minimum norm according to 1.1.4: x = pinv (H (u)) · V (1.1.5)

Analytical solution of pseudoinverse

The elegance of the use of the pseudoinverse pinv (H (u)) according to the invention is that the pseudoinverse P ^inv ( ^H ( ^u )) for the matrix ^H ( ^u ) can be calculated analytically, whereby the calculation method can be easily implemented in a real-time application, for example on the basis of the following expressions 1.1.6:

Delay compensation for vehicle speed and yaw rate

At a point in time K = k-Ts, a direct speed measurement with the value V _{k is} not available due to the time delay of the signals. Assuming that measurements Vi are available at the time L = lT _s with k> I, the corresponding states ^c can be determined, for example according to (1.1.7): xi = pinv (H (ui)) V i ( 1.1.7)

One way to determine values at a later point in time K = kT _s can, for example, be to integrate the changes in state ^c over time, i.e. from point in time L to point in time K, for example according to expression (1.1.8)

The time derivative ^c to state ^c can be derived from

Determine acceleration measurements e ® and a single-track model, which provides the distance R. e R from the center of rotation or from the center of rotation. The expression (1.1.9) results:

This results in a representation for the state ^c at the time K according to the expression (1.1.10):

(1.1,10)

To calculate the values for such a representation according to expression (1.1.10), measured values for acceleration A and the center of rotation or center of rotation or the corresponding distance R of the center of rotation with respect to the single-track model must be available and known without significant delay. When using measurements from an ESP system with regard to acceleration, an offset estimate must be implemented.

Application of the concept - measurement and simulation

4 and 5, graphs 140, 150 show tracks 143-1 to 153-3 for various signals for yaw rate and speed as a function of time t, as part of a parallel parking scenario.

The solid tracks 143-1, 153-1 refer to

Reference system, which is used to represent the actual conditions. The measurements related to the

Reference systems were recorded with an inertial measurement unit coupled to the DGPS system to compensate for sensor errors, such as offset, drift and gain.

The values calculated from the wheel circumferential speeds or wheel speeds are shown in dashed lines as derived values or estimated values in tracks 143-2, 153-2. They have a time delay and are determined according to expression (1.1.7). The measurements compensated by acceleration measurements with a time delay are shown in dotted lines in tracks 143-3, 153-3, the corresponding values are generated according to the expression (1.1.10).

The time is plotted on the abscissa 141 and 151 in the graphs 140 and 150 of FIGS. 4 and 5. The yaw rate and the vehicle speed are plotted on the ordinates 142 and 152.

fusion concept

Using a Bayesian filter and in particular an extended Kalman filter (extended Kalman filter) together with the pseudo inverses described above for the actual transformation matrix ^H ( ^u )

Measurements of values of the wheel circumferential speed V and values of a distance traveled from wheel contact points or wheel contact points S are fused or connected to one another.

For this purpose, a system function / R ⁸ and a measurement function R ^{9 are} introduced. The system function describes how the vehicle speed ve M, the vehicle yaw rate we R and the distances traveled R ^{4 of} the wheel contact points develop over time. There is one

Representation according to expression (1.2.1):

In this representation, T _{s is} the sampling time. The sizes and P denote the contact point vectors. The size d denotes the vector of the individual wheel rotation angles. The following component representations (1.2.2) result for these quantities:

The measurement function h describes how the values of the measurements z can be determined as a function of the system states x and input values or input values u. The following component representation results (1.2.3):

In this representation, the following sizes occur according to the

Component display (1 .2.4) on: The size S 'denotes that of the corresponding one

Distance traveled by the wheel contact point, which can be determined on the basis of the corresponding circumference of the assigned wheel 4 and the value read from the WIC sensor. The quantities Vk and W _k denote the values or estimates of values of the speed or the yaw rate using the above-mentioned approach.

Prediction - condition:

Prediction - measurement: (1 .2.5) Update with measurement:

In the contexts according to (1.2.5), P denotes the system covariance, S the innovation covariance, K the Kalmangain, Q the system noise and R the

Measurement noise. Y is an auxiliary variable.

The matrices F and H are defined in connection with the expressions (1.2.6) and (1.2.7).

In a preferred embodiment, a Bayesian filter and in particular an extended Kalman filter according to the above scheme (1.2.5) is used to determine or estimate the values of v and w. In general, the complete state that contains S is determined.

However, S is only integrated from v and w. As a result, the position of the vehicle 1, which can be calculated from and w, can be determined outside the filter. This results in the representation according to the expression (1 .2.6):

 TT _j_ ie

Note that ⁿ '9x holds because the lower right (4 ^c 4) sub-matrix of H is not an identity matrix. The result is:

Here, s cannot be influenced directly by a measurement of S.

Nevertheless, s is corrected indirectly via the states or values of v and w. Thereby, the position of the vehicle 1 can be determined or estimated from the states or values of v and w and over the distance traveled, according to and in accordance with the path of the contact points.

Simulationserqebnisse

In FIGS. 6 and 7, the graphs 160 and 170

Simulation results for the same sequence as described above are shown.

The time is plotted on the abscissa 161 and 171 in graphs 160 and 170 of FIGS. 6 and 7. The yaw rate and the vehicle speed are plotted on the ordinates 162 and 172.

The solid tracks 163-1 and 173-1 again refer to reference measurements, the dashed tracks 163-2 and 173-2 relate to values which were generated using a Bayesian filter and in particular an extended Kalman filter 20 (EKF).

It can be seen that in particular during the time during which a pseudo measurement of the speed v and yaw rate is available, the final or final determination or estimation by the Bayesian filter and in particular the extended Kalman filter 20 are very good. The accuracy of the angular rate can be further improved by taking pseudo measurements

Using a single track model or simply by measuring the yaw rate.

Claims

Expectations 1. Driver assistance method (T) for a vehicle (1) in which  - A current speed (v) of the vehicle (1) and one  current yaw rate (w) of the vehicle (1) can be determined (T1) and - On the basis of the determined current speed (v) and the determined current yaw rate (w) of the vehicle (1), a process of self-localization of the vehicle (1) is carried out (T2), with an instantaneous circumferential wheel speed (V) of one or more Wheels (4) of the vehicle (1) are directly recorded, evaluated and used as a basis for determining (T1) the current speed (v) and the current yaw rate (w) of the vehicle (1). 2. Driver assistance method (T) according to claim 1, in which a respective current wheel peripheral speed (V) by a  Wheel circumference speed sensor is detected and provided. 3. Driver assistance method (T) according to one of the preceding claims, in which a time delay of a detected instantaneous  Wheel circumferential speed (V) by extrapolation from time  Measured values at an earlier measurement time to a current evaluation time, in particular (i) is compensated by temporal integration from the earlier measurement time to the current evaluation time, (ii) on the basis of one or more measurement values at an instantaneous acceleration (A) of the vehicle (1) and / or on the basis of a single-track model of the vehicle (1). 4. Driver assistance method (T) according to one of the preceding claims, in which when determining (T1) the instantaneous speed (v) and the instantaneous yaw rate (w) of the vehicle (1), a process of Moore's pseudo-inversion is provided and determined Wheel circumferential speeds (V) is applied. 5. Driver assistance method (T) according to one of the preceding claims, in which during and for determining (T1) the instantaneous speed (v) and the instantaneous yaw rate (w) of the vehicle (1) a Moore's pseudo inverse (pinv (H (u)) ) a transformation matrix (H (u)) between a state (x) of the vehicle (1) and a vector formed by the individually determined wheel circumferential speeds (V) and for providing the current speed (v) and the current yaw rate (w) of the Vehicle (1) is applied to the vector formed by the individually determined wheel circumferential speeds (V). 6. Driver assistance method (T) according to one of the preceding claims, in which  - Additionally a currently traveled distance (S) of the  Contact point of one or more wheels (4) of the vehicle (1) recorded, evaluated and determining (T1) the current speed (v), the current yaw rate (w), a current position and / or a current orientation of the vehicle (1) is taken as a basis and in particular  - A respective currently traveled distance (S) of a respective contact point of a wheel (4) of the vehicle (1) is recorded and made available via a respective wheel pulse counter, taking into account a value provided for the circumference of the wheel (4). 7. Driving assistance method (T) according to claim 6, in which a respective detected instantaneous wheel circumferential speed (V) of one or more wheels (4) of the vehicle (1) and a respective detected distance traveled (S) of the contact point of one or more wheels (4) of the vehicle (1) are fed to a Bayesian filter (20) and in particular an extended Kalman filter (20) for evaluation, for plausibility checking and / or for determining a current position and / or current orientation of the vehicle (1). 8. Control unit (50) for a driver assistance system (100) of a vehicle (1), the control unit (50) being set up to initiate, execute, control and / or regulate a driver assistance method (T) according to one of the preceding claims. 9. Driver assistance system (100) for a vehicle (1), the  Driver assistance system (100) is set up to initiate, execute, control and / or regulate a driver assistance method (T) according to one of Claims 1 to 7, and / or a control unit (50) according to Claim 8 or an active connection to such a control unit (50). 10. Vehicle (1) which is set up to be used with a driver assistance method (T) according to one of claims 1 to 7, and / or which is equipped with a driver assistance system (100) according to claim 9  is trained.