U.S. patent application number 14/344342 was filed with the patent office on 2014-12-18 for sensor system comprising a vehicle model unit.
This patent application is currently assigned to Continental Teves AG & Co. oHG. The applicant listed for this patent is Nico Steinhardt. Invention is credited to Nico Steinhardt.
Application Number | 20140371990 14/344342 |
Document ID | / |
Family ID | 46832406 |
Filed Date | 2014-12-18 |
United States Patent
Application |
20140371990 |
Kind Code |
A1 |
Steinhardt; Nico |
December 18, 2014 |
SENSOR SYSTEM COMPRISING A VEHICLE MODEL UNIT
Abstract
The invention relates to a sensor system for a vehicle,
comprising at least two wheel rotation speed sensor elements, at
least one steering angle sensor element and a signal processing
device which is designed to evaluate at least part of the sensor
signals of the sensor elements together. Said signal processing
device comprises a vehicle model unit which is designed to
calculate, from the sensor signals of the wheel rotation speed
sensor elements and the steering angle sensor elements, at least
the speed along a first defined axis, the speed along a second
defined axis and the rotation rate about a third defined axis.
Inventors: |
Steinhardt; Nico;
(Frankfurt, DE) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Steinhardt; Nico |
Frankfurt |
|
DE |
|
|
Assignee: |
Continental Teves AG & Co.
oHG
Frankfurt
DE
|
Family ID: |
46832406 |
Appl. No.: |
14/344342 |
Filed: |
September 12, 2012 |
PCT Filed: |
September 12, 2012 |
PCT NO: |
PCT/EP2012/067878 |
371 Date: |
August 15, 2014 |
Current U.S.
Class: |
701/41 ;
701/70 |
Current CPC
Class: |
G06K 9/00503 20130101;
G08C 19/24 20130101; B60T 2210/36 20130101; B60R 16/0231 20130101;
G01C 21/165 20130101; G06K 9/624 20130101; B60T 2250/06 20130101;
G01S 19/47 20130101 |
Class at
Publication: |
701/41 ;
701/70 |
International
Class: |
B60R 16/023 20060101
B60R016/023 |
Foreign Application Data
Date |
Code |
Application Number |
Sep 12, 2011 |
DE |
102011082535.5 |
Sep 12, 2011 |
DE |
102011082539.8 |
Sep 12, 2011 |
DE |
102011082548.7 |
Sep 12, 2011 |
DE |
102011082549.5 |
Sep 12, 2011 |
DE |
102011082551.7 |
Sep 12, 2011 |
DE |
102011082552.5 |
Nov 9, 2011 |
DE |
102011082534.7 |
Nov 21, 2011 |
DE |
102011086710.4 |
May 2, 2012 |
DE |
102012207297.7 |
Claims
1. A sensor system, for a vehicle, comprising at least two wheel
rotation speed sensor elements, at least one steering angle sensor
element and a signal processing device adapted to evaluate sensor
signals of the sensor elements, wherein the signal processing
device comprises a vehicle model unit adapted to calculate from the
sensor signals of the wheel rotation speed sensor elements and the
steering angle sensor element, at least a velocity of the vehicle
along a first defined axis, a velocity of the vehicle along a
second defined axis, and a rotation rate of the vehicle about a
third defined axis.
2. The sensor system as claimed in claim 1, wherein the first, the
second and the third defined axes form a generating system, and in
the process are oriented perpendicular to one another.
3. The sensor system as claimed in claim 1, wherein the vehicle
model unit is adapted to use a least-squared-error method for
solving an overdetermined system of equations for the
calculation.
4. The sensor system as claimed in claim 1, wherein in each case
one of the wheel rotation speed sensor elements is assigned to each
wheel of the vehicle, wherein the vehicle model unit is adapted to
calculate, directly or indirectly, velocity components or a
velocity of each wheel with respect to the first and second defined
axes based on the sensor signals of the wheel rotation speed sensor
elements and a steering angle, provided by the steering angle
sensor element, or a steering angle of each wheel, detected by the
at least one steering angle sensor element for one or in each case
for a plurality of steerable axles or by at least one model
assumption for one or more unsteerable axles, wherein, from these
velocity components, in relation to the respective wheels or the
velocities in each case with respect to the first and second
defined axes of the associated wheels, the velocity of the vehicle
along the first defined axis, the velocity of the vehicle along the
second defined axis and the rotation rate of the vehicle about the
third defined axis are calculated.
5. The sensor system as claimed in claim 1, wherein the sensor
system has four wheel rotation speed sensor elements, wherein in
each case one of the wheel rotation speed sensor elements is
assigned to each wheel of the vehicle, wherein the vehicle model
unit is adapted to calculate, directly or indirectly, velocity
components or the velocity of each wheel with respect to the first
and second defined axes based on the sensor signals of the wheel
rotation speed sensor elements and a steering angle, provided by
the steering angle sensor unit, or the steering angle of each
wheel, detected by the steering angle sensor element for the front
wheels and from a model assumption or at least by means of a
further steering angle sensor element for the rear wheels, wherein,
from these eight velocity components and/or the four velocities, in
each case with respect to the first and second defined axes, the
velocity of the vehicle along the first defined axis, the velocity
of the vehicle along the second defined axis and the rotation rate
of the vehicle about the third defined axis are calculated.
6. The sensor system as claimed in claim 1, wherein the vehicle
model unit is adapted to consider in its calculation, at least the
following physical variables or parameters: a) a steering angle of
each wheel, detected by the steering angle sensor for the two front
wheels, wherein the model assumption is applied whereby a steering
angle of the rear wheels is known, the steering angle of the rear
wheels is equal to zero or the steering angle of the rear wheels is
additionally detected, b) a wheel rotation speed or a variable
dependent thereon of each wheel, c) a direction of rotation of each
wheel, d) a dynamic radius or wheel diameter of each wheel or a
variable derived therefrom as a parameter, which is taken into
consideration or estimated or calculated as the constant value
which is known to the model, and e) a track width of each axle of
the vehicle or a wheelbase between the axles of the vehicle.
7. The sensor system as claimed in claim 6, wherein the vehicle
model unit is adapted to consider, in its calculation, at least one
of the following physical variables or parameters: f) a slip angle
of each wheel, calculated from transverse acceleration or the
acceleration in the direction of the second defined axis; or g) a
wheel slip, calculated from wheel forces or accelerations of each
wheel.
8. The sensor system as claimed in claim 1, wherein the signal
processing device comprises a tire parameter estimation unit, which
is adapted to calculate or estimate at least a radius, a dynamic
radius, of each wheel or a variable dependent thereon or derived
therefrom and provides this to the vehicle model unit as an
additional input variable.
9. The sensor system as claimed in claim 8, wherein the tire
parameter estimation unit is further adapted to calculate or
estimate a cornering stiffness, and a slip stiffness of each wheel
or a variable dependent thereon or derived therefrom and provides
it to the vehicle model unit as an additional input variable,
wherein the tire parameter estimation unit is adapted to use a
substantially linear tire model for the calculation of the tire or
wheel variables.
10. The sensor system as claimed in claim 1, wherein the vehicle
model unit is adapted to for each of its three calculated
variables, namely the velocity along the first defined axis, the
velocity along the second defined axis, and the rotation rate about
the third defined axis, calculate information relating to data
quality and provides this as an additional output variable, in each
case a variance.
11. The sensor system as claimed in claim 10, wherein the vehicle
model unit adapted to evaluate a validity of its own output
variables on the basis of the calculated variances and in the
process takes into consideration the respective variance of the
velocity along the first and along the second defined axis and of
the rotation rate about the third defined axis in the evaluation of
the validity of its own output variables.
12. The sensor system as claimed in claim 11, wherein the vehicle
model unit is adapted to check the respective variance of its three
output variables with respect to a or in each case one defined
limit value being overshot, wherein, in the event of one or more of
the variances being overshot, no validity of the present output
variables of the vehicle model unit is provided.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims priority to German Patent
Application Nos. 10 2011 082 534.7, filed Sep. 12, 2011, 10 2011
082 535.5, filed Sep. 12, 2011, 10 2011 082 539.8, filed Sep. 12,
2011, 10 2011 082 548.7, filed Sep. 12, 2011, 10 2011 082 549.5,
filed Sep. 12, 2011, 10 2011 082 551.7, filed Sep. 12, 2011, 10
2011 082 552.5, 10 2011 086 710.4, filed Nov. 21, 2011, 10 2012 207
297.7, filed May 2, 2012 and International Patent Application No.
PCT/EP2012/067878, filed Sep. 12, 2012.
FIELD OF THE INVENTION
[0002] The invention relates to a sensor system in accordance with
the preamble of claim 1 and to the use thereof in motor vehicles,
in particular in automobiles.
BACKGROUND OF THE INVENTION
[0003] Laid-open specification DE 10 2010 063 984 A1 describes a
sensor system, comprising a plurality of sensor elements and a
signal processing device, wherein the signal processing device is
configured in such a way that the output signals from the sensor
elements are evaluated jointly.
SUMMARY AND INTRODUCTORY DESCRIPTION OF THE INVENTION
[0004] The invention is based on the object of proposing a sensor
system which provides or enables a relatively high degree of
accuracy with respect to its signal processing.
[0005] This object is achieved by the sensor system described
herein.
[0006] Expediently, the sensor system is arranged in a vehicle, in
particular a motor vehicle, particularly preferably an
automobile.
[0007] It is preferable for the first, the second and the third
defined axes to form a generating system, and in the process to be
in particular oriented perpendicular to one another.
[0008] Preferably, the vehicle model unit is designed in such a way
that it uses a least-squared-error method for solving an
overdetermined system of equations for the calculation.
[0009] It is preferable for in each case one of the wheel rotation
speed sensor elements to be assigned to each wheel of the vehicle,
wherein the vehicle model unit is designed in such a way that, from
the sensor signals of the wheel rotation speed sensor elements and
the steering angle, provided by the steering angle sensor unit,
and/or the steering angle of each wheel, in particular detected by
the at least one steering angle sensor element for one or in each
case for the plurality of steered/steerable axles and/or by at
least one model assumption for one or more unsteered/unsteerable
axles, it calculates, directly or indirectly, the velocity
components and/or the velocity of each wheel along/with respect to
the first and second defined axes, wherein, from these velocity
components, in relation to the respective wheels and/or the
velocities in each case with respect to the first and second
defined axes of the associated wheels, the speed along a first
defined axis, the speed along a second defined axis and the
rotation rate about a third defined axis are calculated.
[0010] It is expedient for the sensor system to have four wheel
rotation speed sensor elements, wherein in each case one of the
wheel rotation speed sensor elements is assigned to each wheel of
the vehicle, wherein the vehicle model unit is designed in such a
way that, from the sensor signals of the wheel rotation speed
sensor elements and the steering angle, provided by the steering
angle sensor unit, and/or the steering angle of each wheel, in
particular detected by the steering angle sensor element for the
front wheels and from a model assumption or at least by means of a
further steering angle sensor element for the rear wheels, it
calculates, directly or indirectly, the velocity components and/or
the velocity of each wheel along/with respect to the first and
second defined axes, wherein, from these eight velocity components
and/or the four velocities, in each case with respect to the first
and second defined axes, the velocity along a first defined axis,
the velocity along a second defined axis and the rotation rate
about a third defined axis are calculated.
[0011] It is preferable for the steering angle of each wheel to be
determined or calculated from a steering wheel angle sensor
element, i.e. a sensor element which detects the steering angle
desired by the driver, and from information relating to the
steering ratio characteristic, which is in particular stored in the
vehicle model unit or in another part of the signal processing
device.
[0012] It is expedient for the vehicle model unit to be designed in
such a way that, in its calculation, it takes into consideration at
least the following physical variables and/or parameters
[0013] a) the steering angle of each wheel, in particular detected
by the steering angle sensor for the two front wheels, wherein the
model assumption is applied whereby the steering angle of the rear
wheels is known, in particular the steering angle of the rear
wheels is equal to zero or the steering angle of the rear wheels is
additionally detected,
[0014] b) the wheel rotation speed or a variable dependent thereon
of each wheel,
[0015] c) the direction of rotation of each wheel,
[0016] d) the dynamic radius and/or wheel diameter of each wheel or
a variable derived therefrom as parameter, which is in particular
taken into consideration or estimated and/or calculated as the
constant value which is known to the model, and
[0017] e) the track width of each axle of the vehicle and/or the
wheelbase between the axles of the vehicle.
[0018] Particularly preferably, the vehicle model unit is designed
in such a way that, in its calculation, it takes into consideration
at least one of the following physical variables and/or
parameters
[0019] f) the slip angle of each wheel, in particular calculated
from the transverse acceleration, i.e. the acceleration in the
direction of the second defined axis, and/or
[0020] g) the wheel slip, in particular calculated from wheel
forces and/or accelerations of each wheel.
[0021] It is preferable for the signal processing device to
comprise a tire parameter estimation unit, which is designed in
such a way that it calculates and/or estimates at least the radius,
in particular the dynamic radius, of each wheel or a variable
dependent thereon or derived therefrom and provides this to the
vehicle model unit as additional input variable.
[0022] Particularly preferably, the tire parameter estimation unit
is designed in such a way that it additionally calculates and/or
estimates the cornering stiffness, and the slip stiffness or
longitudinal slip stiffness of each wheel or a variable dependent
thereon or derived therefrom and provides it to the vehicle model
unit as additional input variable, wherein the tire parameter
estimation unit is designed in such a way that it uses in
particular a substantially linear tire model for the calculation of
the wheel/tire variables.
[0023] Expediently, the tire parameter estimation unit is designed
in such a way that it receives the wheel rotation speeds and the
steering angle as input variables, at least partially or completely
the output variables or values of the strapdown algorithm unit, in
particular the variances provided thereby in addition to the values
of the physical variables, and the variances of the fusion filter,
with respect to the physical variables which are the input
variables of the tire parameter estimation unit.
[0024] It is preferable for the vehicle model unit to be designed
in such a way that, for each of its three calculated variables,
namely the velocity along a first defined axis, the velocity along
a second defined axis, and the rotation rate about a third defined
axis, it calculates information relating to the data quality and
provides this as additional output variable, in particular in each
case a variance.
[0025] It is expedient for the vehicle model unit is designed in
such a way that it evaluates the validity of its own output
variables on the basis of the calculated variances and in the
process in particular takes into consideration the respective
variance of the velocity along the first and along the second
defined axis and of the rotation rate about the third defined axis
in the evaluation of the validity of its own output variables.
[0026] Particularly preferably, the vehicle model unit is designed
in such a way that it checks the respective variance of its three
output variables with respect to a or in each case one defined
limit value being overshot, wherein, in the event of one or more of
the variances being overshot, no validity of the present output
variables of the vehicle model unit is provided.
[0027] It is preferable for the vehicle model unit and/or the tire
parameter estimation unit to be designed in such a way that they
comprise at least one linearization. This linearization is in
particular only carried out or has the boundary parameter that the
total acceleration of the vehicle, that is to say the acceleration
relative to all three defined axes, is less than 5 m/s.sup.2 in
terms of magnitude.
[0028] It is expedient for the first, second and third defined axes
to be defined in relation to a coordinate system of the vehicle in
which the sensor system is implemented, as follows:
[0029] the first defined axis corresponds to the longitudinal axis
of the vehicle,
[0030] the second defined axis corresponds to the transverse axis
of the vehicle, and
[0031] the third defined axis corresponds to the vertical axis of
the vehicle. These three axes form in particular a Cartesian
coordinate system, in particular a vehicle coordinate system.
[0032] It is preferable for the vehicle model unit to be designed
in such a way that it carries out or supports a direct or indirect
measurement of the wheel loads and/or wheel contact forces and at
least provides this variable as output variable.
[0033] It is expedient for the vehicle model unit to be designed in
such a way that it comprises the modeling of a wheel suspension,
with regard to a kinematic and/or dynamic model, as a result of
which, taking into consideration this modeling, a steering angle is
or can be calculated with increased accuracy. Said steering angle
relates, in particular, to the steering angle of each wheel, which
is used in each case for the further calculation of the output
variables of the vehicle model unit.
[0034] The steering angle of the rear wheels is expediently
detected by means of at least one additional rear wheel steering
angle sensor element. In particular additionally or alternatively
preferably the actuator system of a rear axle steering system
provides the steering angle of the wheels of the rear axle.
[0035] The signal processing device of the sensor system
additionally preferably comprises a fusion filter. The fusion
filter provides a defined fusion data set in the course of the
joint evaluation of at least the sensor signals and/or signals
derived therefrom of the sensor elements, that is to say of the
odometry, and in particular additionally of the output signals of a
satellite navigation system and/or signals derived therefrom. Said
fusion data set has in each case data with respect to defined
physical variables, wherein the fusion data set comprises, with
respect to at least one physical variable, a value of said physical
variable and information about the data quality thereof, wherein
this information about the data quality is expressed as variance in
accordance with the example.
[0036] Preferably, the fusion data set comprises, as the value of
the at least one physical variable, a relative value, for example a
correction value, also referred to as offset value or change value
or error value.
[0037] The relative values of the respective physical variables of
the fusion data set are therefore expediently correction values and
variances.
[0038] The values of the physical variables of the fusion data set
are preferably calculated on a direct or indirect basis of the
sensor signals of the sensor elements and the satellite navigation
system, wherein at least some variables, for example the velocity
and the position of the vehicle in relation to the vehicle
coordinates, are detected and computed with redundancy.
[0039] The fusion filter is expediently designed as an error state
space extended sequential Kalman filter, that is to say as a Kalman
filter which comprises, in particular a linearization, and in which
the correction values are calculated and/or estimated and which
operates sequentially and in this case uses/takes into
consideration the input data available in the respective function
step of the sequence.
[0040] The vehicle model unit provides its output variables or
output data, that is to say at least the velocity along a first
defined axis, the velocity along a second defined axis and the rate
of rotation about a third defined axis, preferably to the fusion
filter, which takes into consideration or uses said output
variables or output data of the vehicle model unit in its
calculations, that is to say the calculations of the fusion
filter.
[0041] The dynamic radius of each wheel or the dynamic tire radius
r.sub.dyn is preferably defined as follows: effectively covered
distance during a tire revolution. The latter does not correspond
to the radius of the tire, since the radius of the tire effectively
decreases as a result of spring deflection under load. Influencing
variables that can alter the tire radius including during a journey
are, for example, traveling velocity, air pressure and
temperature.
[0042] The variable referred to as longitudinal slip .lamda. is
expediently defined as follows: under the influence of longitudinal
force, a slip movement arises as a result of the deformation of the
tread elements of the tire, without taking into consideration the
tire sliding on the road. Said slip movement has the consequence
that the tire, depending on the longitudinal force, rotates faster
or more slowly than would be expected over the tire radius. The
extent to which this effect is manifested is influenced principally
by the rubber mix and the type of tire and is characterized by the
longitudinal slip stiffness:
c .lamda. = F x .lamda. . ##EQU00001##
[0043] Skew running or the slip angle .alpha. is preferably defined
as follows: in a manner similar to that in the case of longitudinal
slip, lateral forces, perpendicular to the rolling direction, cause
a sideways movement of the tire as a result of the rubber
elasticity. This relationship is characterized by the cornering
stiffness:
c .alpha. = F y .alpha. . ##EQU00002##
[0044] In order to compensate for these disturbance variables, the
vehicle model unit preferably makes recourse to a linear tire model
of the tire parameter estimation unit or includes it in the
calculations. Said model is restricted in particular to
accelerations or total accelerations of the vehicle
< 5 m s 2 . ##EQU00003##
In this range, it is particularly preferably assumed, in particular
as a model assumption for the calculation, that the relationship
between longitudinal slip and skew running and the associated
forces is linear, and that the forces that can be transmitted rise
linearly with the contact force F.sub.R or normal force on the
tire. By canceling the vehicle mass, this allows a normalization of
the variables to accelerations. In this case, the vehicle masses
and accelerations are expediently related to individual wheels, but
assumed to be randomly distributed:
r dyn = 2 .pi. .DELTA. .PHI. wheel s x , absolute , tire
##EQU00004## .lamda. = F x F N c .lamda. = m Fzg a x , Fzg m Fzg g
c .lamda. = a x , Fzg g c .lamda. ##EQU00004.2## .alpha. = F y F N
c .alpha. = m Fzg a y , Fzg m Fzg g c .alpha. = a y , Fzg g c
.alpha. ##EQU00004.3##
In this case, the following preferably hold true:
.DELTA..phi..sub.wheel: Angle of rotation of the wheel measured
from wheel ticks S.sub.x,absolute,tire: Distance actually covered
over the road g: Acceleration due to gravity The following
preferably ensue therefrom for the distances covered:
.DELTA. x , wheel = 2 .pi. .DELTA. .PHI. wheel r dyn 1 - .lamda. a
1 + .lamda. b ##EQU00005## .DELTA. y , wheel = .DELTA. x , wheel
tan .varies. ##EQU00005.2##
In this case, the following preferably hold true: .lamda..sub.a:
Drive slip during acceleration .lamda..sub.a: Braking slip during
deceleration According to the example, therefore, it is the case
that the slip variable that is respectively not applicable in the
traveling situation=0.
[0045] Since the accelerations used are known from the navigation
calculation, the actual planar movement of the vehicle over the
roadway can be estimated in a model-based manner preferably given a
known tire radius and known cornering and longitudinal slip
stiffness. A possible skew of tire coordinates relative to the
vehicle coordinates is expediently taken into consideration by
means of the measured steering wheel angle and the known steering
ratio. The distances and velocities of the individual wheels are
preferably calculated as follows in the vehicle model unit:
[0046] calculation of accelerations and rates of rotation at the
center of gravity of the vehicle
[0047] transformation to tire coordinates
[0048] calculation of the velocities/distances using the tire model
and the wheel angular momenta or wheel rotation speeds
[0049] inverse transformation into vehicle coordinates.
[0050] Preferably, two measured variables (.DELTA.X.sub.wheel,
.gamma.Y.sub.wheel in vehicle coordinates) per wheel, that is to
say a total of eight measurement values, are available after the
conclusion of these steps.
[0051] It is preferable for the tire parameter estimation unit to
be designed in such a way that it carries out a method for
estimating tire parameters for a vehicle, comprising the following
steps:
[0052] measuring a reference movement of the vehicle;
[0053] modeling a model movement of the vehicle on the basis of a
model freed of the tire parameters to be estimated; and
[0054] estimating the tire parameters of the vehicle on the basis
of a comparison of the reference movement and the model
movement.
[0055] The method comprises in particular additionally the
following step:
[0056] detecting the real velocity of the vehicle at wheel contact
points of the vehicle.
[0057] It is preferred for the method to comprise the following
step:
[0058] establishing the model freed of the tire parameters to be
estimated on the basis of approximated tire parameters, and in
particular the following further step:
[0059] using the estimated tire parameters as approximated tire
parameters in the model, for estimating new tire parameters.
[0060] The method expediently comprises the following steps:
[0061] detecting a variance of the reference movement, and
[0062] estimating the tire parameters of the vehicle on the basis
of the detected variance.
[0063] The method is preferably developed by the estimated tire
parameters of the vehicle being regarded as valid if the reference
movement and/or the model movement exceed(s) a specific value.
[0064] The method expediently comprises the following step:
[0065] comparing the reference movement and the model movement on
the basis of an observer.
[0066] With regard to the method in the tire parameter estimation
unit the observer preferably a Kalman filter.
[0067] It is preferable for the signal processing device to have a
fusion filter, which provides a defined fusion data set in the
course of the joint evaluation of at least the sensor signals
and/or signals derived therefrom of the sensor elements, wherein
said fusion data set has in each case data with respect to defined
physical variables, wherein the fusion data set comprises, with
respect to at least one physical variable, a value of said physical
variable and information about the data quality thereof.
[0068] The fusion filter is preferably in the form of a Kalman
filter, alternatively preferably a particle filter or alternatively
an information filter or alternatively in the form of an
"unscented" Kalman filter.
[0069] It is preferable for the fusion filter to be designed in
such a way that the fusion data set comprises, as value of the at
least one physical variable, a relative value, in particular an
offset value and/or change value and/or correction value and/or
error value.
[0070] It is expedient for the relative values of the respective
physical variables of the fusion data set to be correction values,
to each of which scattering information or scattering or scattering
degree, in particular a variance, is assigned as information
relating to the data quality of said correction values.
[0071] It is preferable for the fusion filter to be designed in
such a way that the value of at least one physical variable of the
fusion data set is calculated on a direct or indirect basis from
sensor signals from a plurality of sensor elements, wherein these
sensor elements detect this at least one physical variable in a
direct or indirect manner, with redundancy. This redundant
detection is particularly preferably implemented as direct or
parallel redundancy and/or as analytical redundancy, from
computationally derived or deduced variables/values and/or model
assumptions.
[0072] The fusion filter is preferably in the form of a Kalman
filter which iteratively implements at least prediction steps and
correction steps and at least partially provides the fusion data
set. In particular, the fusion filter is in the form of an error
state space extended sequential Kalman filter, i.e. in the form of
a Kalman filter which particularly preferably comprises
linearization and in which error state information is calculated
and/or estimated and/or which operates sequentially and in the
process uses/takes into consideration the input data available in
the respective function step of the sequence.
[0073] It is expedient for the sensor system to have an inertial
sensor arrangement, comprising at least one acceleration sensor
element and at least one rotation rate sensor element, and for the
sensor system to comprise a strapdown algorithm unit, in which a
strapdown algorithm is implemented, with which at least the sensor
signals of the inertial sensor arrangement relating to in
particular corrected navigation data and/or driving dynamics data
are processed, on the basis of the vehicle in which the sensor
system is arranged.
[0074] It is particularly preferable for the strapdown algorithm
unit to provide its calculated navigation data and/or driving
dynamics data to the fusion filter directly or indirectly.
[0075] The sensor system preferably has an inertial sensor
arrangement, which is designed in such a way that it can detect at
least the acceleration along a second defined axis, in particular
the transverse axis of the vehicle, and at least the rotation rate
about a third defined axis, in particular the vertical axis of the
vehicle, wherein the first and third defined axes form a generating
system, and in the process are in particular oriented perpendicular
to one another, wherein the sensor system also has at least one
wheel rotation speed sensor element, in particular at least or
precisely four wheel rotation speed sensor elements, which detect
the wheel rotation speed of a wheel or the wheel rotation speeds of
in each case one of the wheels of the vehicle and in particular
additionally detect the direction of rotation of the assigned wheel
of the vehicle in which the sensor system is arranged,
[0076] wherein the sensor system additionally comprises at least
one steering angle sensor element, which detects the steering angle
of the vehicle, and
[0077] wherein the sensor system furthermore comprises a satellite
navigation system, which is designed in particular in such a way
that it detects and/or provides the distance data in each case
between the assigned satellite and the vehicle or a variable
dependent thereon and velocity information data in each case
between the assigned satellite and the vehicle or a variable
dependent thereon.
[0078] Particularly preferably, the inertial sensor arrangement is
designed in such a way that it can detect at least the
accelerations along a first, a second and a third defined axis and
at least the rotation rates about these first, second and third
defined axes, wherein these first, second and third defined axes
form a generating system, and in the process are in particular in
each case oriented perpendicular to one another.
[0079] It is preferable for the inertial sensor arrangement to
provide its sensor signals to the strapdown algorithm unit and for
the strapdown algorithm unit to be designed in such a way that it
at least calculates and/or provides
[0080] at least corrected accelerations along the first, the second
and the third defined axes,
[0081] at least corrected rotation rates about these three defined
axes,
[0082] at least a velocity with respect to these three defined
axes,
[0083] and at least one position variable,
[0084] as measured variables and/or navigation data and/or driving
dynamics data
[0085] from the sensor signals of the inertial sensor arrangement
and in particular at least fault state information and/or variance
and/or information on the data quality which is assigned to a
sensor signal or a physical variable and is provided by the fusion
filter.
[0086] It is expedient for the sensor system to be designed in such
a way that in each case at least one sensor signal and/or a
physical variable, as direct or derived variable
[0087] of the inertial sensor arrangement and/or the strapdown
algorithm unit,
[0088] of the wheel rotation speed sensor elements and the steering
angle sensor element, in particular indirectly via a vehicle model
unit,
[0089] and of the satellite navigation system, in this case in
particular distance data in each case between the assigned
satellite and the vehicle or a variable dependent thereon and
velocity information data in each case between the assigned
satellite and the vehicle or a variable dependent thereon,
[0090] are provided to the fusion filter and taken into
consideration by the fusion filter in the calculations it
performs.
[0091] It is particularly preferable for the vehicle model unit to
be designed in such a way that
[0092] the speed along the first defined axis,
[0093] the speed along the second defined axis
[0094] and the rotation rate about the third defined axis
[0095] are calculated from the sensor signals of the wheel rotation
speed sensor elements and the steering angle sensor element.
[0096] It is very particularly preferable for the vehicle model
unit to be designed in such a way that, for the calculation, a
least-squared-error method is used for solving an overdetermined
system of equations.
[0097] It is expedient for the vehicle model unit to be designed in
such a way that, in its calculation, it takes into consideration at
least the following physical variables and/or parameters
[0098] a) the steering angle of each wheel, in particular detected
by the steering angle sensor for the two front wheels, wherein the
model assumption whereby the steering angle of the rear wheels is
equal to zero or the steering angle of the rear wheels is
additionally detected is used,
[0099] b) the wheel rotation speed or a variable dependent thereon
for each wheel,
[0100] c) the rotation direction of each wheel,
[0101] d) the dynamic radius and/or wheel diameter of each wheel,
and
[0102] e) the track width of each axle of the vehicle and/or the
wheelbase between the axles of the vehicle.
[0103] The signal processing device is preferably designed in such
a way that the fusion filter calculates and/or provides and/or
outputs the fusion data set at defined times.
[0104] The fusion filter is preferably designed in such a way that
it calculates and/or provides and/or outputs the fusion data set
independently of the sampling rates and/or sensor signal output
times of the sensor elements, in particular the wheel rotation
speed sensor elements and the steering angle sensor element, and
independently of temporal signal or measured variable or
information output times of the satellite navigation system.
[0105] It is expedient for the signal processing device to be
designed in such a way that, over the course of a function step of
the fusion filter, the newest information and/or signals and/or
data available to the fusion filter
[0106] of the sensor elements, in particular of the wheel rotation
speed sensor elements and the steering angle sensor element, are
always updated, in particular asynchronously, directly or
indirectly, in particular by means of the vehicle model unit and
the satellite navigation system directly or indirectly,
sequentially and/or are recorded by the fusion filter and taken
into consideration in the calculation of the assigned function step
of the fusion filter.
[0107] It is preferable for the sensor system to have a standstill
identification unit, which is designed in such a way that it can
identify a standstill of the vehicle and, in the event of an
identified standstill of the vehicle, provides information from a
standstill model at least to the fusion filter, in this case in
particular the information that the rotation rates about all of the
three axes have the value zero and at least one position change
variable likewise has the value zero and in particular the
velocities along all three axes have the value zero.
[0108] It is preferable for the signal processing device to
calculate and/or use a first group of data of physical variables,
whose values relate to a vehicle coordinate system, and wherein the
signal processing device additionally calculates and/or uses a
second group of data of physical variables, whose values relate to
a world coordinate system, wherein this world coordinate system is
suitable in particular at least for describing the orientation
and/or dynamic variables of the vehicle in the world, wherein
[0109] the sensor system has an orientation model unit,
[0110] with which the orientation angle between the vehicle
coordinate system and the world coordinate system is calculated,
wherein
[0111] the orientation angle between the vehicle coordinate system
and the world coordinate system is calculated in the orientation
model unit at least on the basis of the following variables:
[0112] the velocity with respect to the vehicle coordinate
system,
[0113] the velocity with respect to the world coordinate system and
in particular the steering angles.
[0114] It is expedient for the following terms to be used
synonymously, i.e. have the same meaning when implemented
technically: offset value, change value, correction value and error
value.
[0115] Error state information is preferably understood to mean
error information and/or error correction information and/or
scattering information and/or variance information and/or accuracy
information.
[0116] The term variance is preferably understood to mean scatter,
wherein in particular in the case of a general fusion filter, said
filter in each case assigns scatter or a scatter value to each
value of a physical variable of the fusion filter, and in the case
of a Kalman filter as the fusion filter, in each case a variance is
assigned to each value of a physical variable of the fusion
filter.
[0117] It is expedient for the first, second and third defined axes
on the basis of a coordinate system of the vehicle in which the
sensor system is implemented to be defined as follows: the first
defined axis corresponds to the longitudinal axis of the vehicle,
the second defined axis corresponds to the transverse axis of the
vehicle, and the third defined axis corresponds to the vertical
axis of the vehicle. These three axes in particular form a
Cartesian coordinate system.
[0118] It is preferable for the fusion filter to be designed in
such a way that its data, in particular the physical variables or
the data of the physical variables of the fusion data set, are
divided into blocks which always have a constant size and which are
processed iteratively in any desired order in the fusion filter,
i.e. the fusion filter implements a sequential update with respect
to its input data. In this case, the fusion filter is particularly
preferably designed in such a way that the filter equations are
matched, with the result that the computational result of the
sequential update in each step of the fusion filter is an update,
i.e. a data update, for all measured variables of the input data of
the fusion filter.
[0119] The sensor system is expediently arranged in a vehicle, in
particular in a motor vehicle, particularly preferably in an
automobile.
[0120] The sensor system is preferably designed in such a way that
data of the satellite navigation system, in particular position
data, are assigned timestamp information, which substantially
describes the measurement time of these data. The timestamp
information of the respective datum of the satellite navigation
system is provided together with this respective datum to the
fusion filter and taken into consideration in the internal
calculation in the fusion filter.
[0121] Preferably, in addition such timestamp information is
likewise assigned to the data of further or all of the sensor
elements and/or the inertial sensor arrangement, which timestamp
information is likewise provided with the respective datum to the
fusion filter and is taken into consideration in the internal
calculation in the fusion filter. Expediently, the respective
timestamp information is generated by the satellite navigation
system itself with respect to the data of the satellite navigation
system.
[0122] It is preferable for the respective timestamp information to
be generated by the signal processing device in the case of the
additional timestamp information of the further sensor elements
and/or the inertial sensor arrangement, in particular depending on
the time measurement of the satellite navigation system.
[0123] Preferably, a function step of the fusion filter comprises
at least one prediction step and a correction step. The fusion
filter is in this case formed iteratively and performs iteratively,
one after the other, function steps. In particular, data or values
or signals are input within each function step of the fusion
filter, i.e. input data are taken into consideration, i.e. data or
values or signals are also output, i.e. provided as output
data.
[0124] The fusion filter is preferably designed in such a way that
the fusion filter implements a plurality of update steps within a
function step, wherein these update steps relate to loading or use
or updating of input data or signals. The fusion filter runs in
particular sequentially through all of the input variables or input
signals and checks in each case whether new information/data are
present. If this is the case, this information or data is
transferred into the filter or the information/data are updated in
the filter, and if this is not the case the present value is
maintained and the filter checks the next input or the next input
variable or the next input signal.
[0125] The strapdown algorithm unit preferably provides at least
absolute values of physical variables, in particular absolute
values for the acceleration, the rotation rate, the velocity, in
this case in each case in relation to the three axes, to the
vehicle and/or world coordinate system, and a position and the
orientation angle. The values with respect to these variables are
in this case particularly preferably all provided by the strapdown
algorithm unit as corrected values/variables.
[0126] It is expedient for the inertial sensor arrangement to clock
and/or trigger the fusion filter, in particular each fusion step
which is implemented by the fusion filter is triggered by the
inertial sensor arrangement or at least one output signal or output
datum.
[0127] It is preferable for the strapdown algorithm unit to be
designed in such a way that it has a start vector of physical
variables and/or a start value of the position, in particular with
respect to the start of the sensor system, particularly preferably
after each time the sensor system is switched on. The strapdown
algorithm unit particularly preferably receives this start vector
and/or this start position via the fusion filter from the satellite
navigation system.
[0128] It is expedient for the data of the fusion filter, in
particular the fusion data set thereof, to image a virtual sensor
or correspond to such a virtual sensor.
[0129] The term sensor elements is preferably understood to mean
the wheel rotation speed sensor elements, the at least one steering
angle sensor element, the sensor elements of the inertial sensor
arrangement and in particular additionally also the satellite
navigation system.
[0130] If, in general, a variable and/or value is specified in
respect of the three defined axes, it is preferable for this to be
intended with respect to the vehicle coordinate system and/or the
world coordinate system.
[0131] It is expedient for the fusion data set, which comprises
values of the physical variables, to comprise a relative value, for
example a correction value, also referred to as offset value, and
in particular to be provided to the strapdown algorithm unit. In
accordance with the example, this respective correction value
results in each case from the accumulated error values or change
values which are provided by the fusion filter.
[0132] In addition, the invention relates to the use of the sensor
system in vehicles, in particular motor vehicles, particularly
preferably in automobiles.
BRIEF DESCRIPTION OF THE DRAWINGS
[0133] Further preferred embodiments result from the description
below relating to an exemplary embodiment with reference to FIG.
1.
[0134] FIG. 1 shows a schematic illustration of an exemplary
embodiment of the sensor system, which is intended for arrangement
and use in a vehicle. In this case, the sensor elements and the
satellite navigation system as well as the most important signal
processing units of the signal processing device are illustrated as
function blocks and the interaction of said blocks with one another
is also illustrated.
DETAILED DESCRIPTION OF THE INVENTION
[0135] The sensor system comprises an inertial sensor arrangement
1, IMU, "inertial measurement unit", which is designed in such a
way that it can detect at least the accelerations along a first, a
second and a third defined axis and at least the rotation rates
about these first, second and third defined axes, wherein the first
defined axis corresponds to the longitudinal axis of the vehicle,
the second defined axis corresponds to the transverse axis of the
vehicle, and the third defined axis corresponds to the vertical
axis of the vehicle. These three axes form a Cartesian coordinate
system, the vehicle coordinate system.
[0136] The sensor system has a strapdown algorithm unit 2, in which
a strapdown algorithm is implemented, with which at least the
sensor signals of the inertial sensor arrangement 1 are processed
to give corrected navigation data and/or driving dynamics data.
These output data of the strapdown algorithm unit 2 include the
data of the following physical variables:
[0137] the velocity, the acceleration and the rotation rate in each
case of the vehicle, by way of example with respect to the three
axes of the vehicle coordinate system and, in accordance with the
example, additionally in each case in relation to a world
coordinate system, which is suitable for describing the orientation
and/or dynamic variables of the vehicle in the world. In addition,
the output data of the strapdown algorithm unit 2 comprise the
position with respect to the vehicle coordinate system and the
orientation with respect to the world coordinate system. In
addition, the output data of the strapdown algorithm unit have the
variances as information on the data quality of the abovementioned
physical variables, at least some of said variables. These
variances, in accordance with the example, are not calculated in
the strapdown algorithm unit, but are only used and passed on by
said strapdown algorithm unit.
[0138] The output data of the strapdown algorithm unit are
additionally, by way of example, the output data or signals 12 of
the entire sensor system.
[0139] The sensor system additionally comprises wheel rotation
speed sensor elements 3 for each wheel of the vehicle, in
accordance with example four, which detect the wheel rotation
speeds of in each case one of the wheels of the vehicle and in each
case additionally detect the direction of rotation, and
additionally a steering angle sensor element 3, which detects the
steering angle of the vehicle. The wheel rotation speed sensor
element and the steering angle sensor element form a sensor
arrangement 3 for odometry detection.
[0140] Furthermore, the sensor system has a satellite navigation
system 4, which is designed in such a way that it detects and/or
provides the distance data in each case between the assigned
satellite and the vehicle or a variable dependent thereon and
velocity information data in each case between the assigned
satellite and the vehicle or a variable dependent thereon. In
addition, the satellite navigation system 4, in accordance with the
example, provides a start position or start position information,
at least for starting or switching on the sensor system, to the
fusion filter.
[0141] The signal processing device of the sensor system also
comprises a fusion filter 5. The fusion filter 5 provides a defined
fusion data set 6 over the course of the joint evaluation of at
least the sensor signals and/or signals derived therefrom of the
sensor elements 3, i.e. the odometry, and the output signals of the
satellite navigation system 4 and/or signals derived therefrom.
This fusion data set has in each case data with respect to defined
physical variables, wherein the fusion data set 6 with respect to
at least one physical variable comprises a value of this physical
variable and information on its data quality, wherein this
information on the data quality is expressed as variance, in
accordance with the example.
[0142] The fusion data set 6 comprises, as value of the at least
one physical variable, a relative value, for example a correction
value, also referred to as offset value. In accordance with the
example, the correction value results in each case from the
accumulated error values or change values which are provided by the
fusion filter 5.
[0143] The relative values of the respective physical variables of
the fusion data set 6 are therefore correction values and
variances, in accordance with the example. In other words, the
fusion data set 6, in accordance with the example, calculates an
error budget, which is provided as input variable or input data set
to the strapdown algorithm unit and is taken into consideration at
least partially by said strapdown algorithm unit in its
calculations. This error budget comprises, as data set or output
data, at least correction values or error values of physical
variables and in each case a variance, as information on the data
quality, with respect to each value. In this case, at least the
correction values and variances with respect to the physical
variables velocity, acceleration and rotation rate, in each case in
relation to the vehicle coordinate system, i.e. in each case the
three components of these variables with respect to this coordinate
system, and IMU orientation or the IMU orientation angle between
the vehicle coordinate system and the coordinate system or the
installation orientation of the inertial sensor arrangement 1 and
the position in relation to the world coordinate system are
transmitted by the fusion filter to the strapdown algorithm
unit.
[0144] The values of the physical variables of the fusion data set
are calculated on a direct or indirect basis of the sensor signals
of the sensor elements 3 and the satellite navigation system 4,
wherein at least some variables, for example the velocity and the
position of the vehicle with respect to the vehicle coordinates,
are detected and used with redundancy with respect to the data of
the strapdown algorithm unit 2.
[0145] The fusion filter 5 is, in accordance with the example, in
the form of an error state space extended sequential Kalman filter,
i.e. in the form of a Kalman filter which comprises in particular
linearization and in which the correction values are calculated
and/or estimated and which operates sequentially and in the process
uses/takes into consideration the input data available in the
respective function step of the sequence.
[0146] The fusion filter 5 is designed in such a way that, over the
course of a function step of the fusion filter, the newest
information and/or signals and/or data available to the fusion
filter
[0147] of the sensor elements 3, i.e. the wheel rotation speed
sensor elements and the steering angle sensor element indirectly by
means of a vehicle model unit 7 and
[0148] of the satellite navigation system 4 directly or
indirectly
are always sequentially updated, asynchronously, and/or recorded in
the fusion filter and taken into consideration in the calculation
of the assigned function step of the fusion filter 5.
[0149] The vehicle model unit 7 is designed in such a way that it
calculates, from the sensor signals of the wheel rotation speed
sensor elements 3 and the steering angle sensor element 3, at
least
[0150] the velocity along a first defined axis,
[0151] the velocity along a second defined axis, and
[0152] the rotation rate about a third defined axis
[0153] and provides these to the fusion filter 5.
[0154] The sensor system has, in accordance with the example, four
wheel rotation speed sensor elements 3, wherein in each case one of
the wheel rotation speed sensor elements is assigned to each wheel
of the vehicle, wherein the vehicle model unit 7 is designed in
such a way that it calculates, from the sensor signals of the wheel
rotation speed sensor elements and the steering angle, provided by
the steering angle sensor unit, and/or the steering angle of each
wheel, in particular detected by the steering angle sensor element
for the front wheels and by means of at least one further steering
angle sensor element for the rear wheels or at least from a model
assumption for the rear wheels,
[0155] the velocity components and/or the velocity of each wheel,
along/with respect to the first and second defined axes directly or
indirectly,
[0156] wherein, from these eight velocity components and/or the
four velocities, in each case with respect to the first and second
defined axes,
[0157] the velocity along a first defined axis,
[0158] the velocity along a second defined axis, and
[0159] the rotation rate about a third defined axis
[0160] are calculated.
[0161] The sensor system or the signal processing device of said
sensor system also comprises a tire parameter estimation unit 10,
which is designed in such a way that it calculates at least the
radius, in accordance with the example the dynamic radius, of each
wheel and additionally calculates the cornering stiffness and the
slip stiffness of each wheel and provides these to the vehicle
model unit 7 as additional input variables, wherein the tire
parameter estimation unit 10 is designed in such a way that it uses
a substantially linear tire model for calculating the wheel/tire
variables. The input variables of the tire parameter estimation
unit in accordance with the example are in this case the wheel
rotation speeds 3 and the steering angle 3, at least partially or
completely the output variables or values of the strapdown
algorithm unit 2, in particular the variances provided thereby in
addition to the values of the physical variables, and the variances
of the fusion filter 5, with respect to the physical variables
which are the input variables of the tire parameter estimation unit
10.
[0162] The sensor system or its signal processing device also
comprises a GPS error identification and plausibilization unit 11,
which is designed in such a way that, in accordance with the
example, it receives, as input data, the output data or output
signals of the satellite navigation system 4 and at least partially
the output data or output signals of the strapdown algorithm unit 2
and takes these into consideration in its calculations.
[0163] In this case, the GPS error identification and
plausibilization unit 11 is additionally connected to the fusion
filter 5 and exchanges data therewith.
[0164] The GPS error identification and plausibilization unit 11 is
designed, by way of example, in such a way that it implements the
following method:
Method for Selecting a Satellite, Comprising:
[0165] measuring measurement position data of the vehicle with
respect to the satellite on the basis of the GNSS signal, i.e. the
global navigation satellite system signal, the output signal or the
output data of the satellite navigation system 4,
[0166] determining reference position data of the vehicle which are
redundant with respect to the measurement position data determined
on the basis of the GNSS signal; and
[0167] selecting the satellite when a comparison of the measurement
position data and the reference position data satisfies a
predetermined condition,
[0168] wherein, in order to compare the measurement position data
and the reference position data, a difference between the
measurement position data and the reference position data is
formed,
[0169] wherein the predetermined condition is a maximum permissible
error between the measurement position data and the reference
position data,
[0170] wherein the maximum permissible error is dependent on a
standard deviation, which is calculated on the basis of a sum of a
reference variance for the reference position data and a
measurement variance for the measurement position data,
[0171] wherein the maximum permissible error corresponds to a
multiple of the standard deviation such that a probability that the
measurement position data fall below a predetermined threshold
value in a scatter interval which is dependent on the standard
deviation.
[0172] The sensor system or its signal processing device also has a
standstill identification unit 8, which is designed in such a way
that it can identify a standstill of the vehicle and, in the event
of an identified standstill of the vehicle, provides information
from a standstill model at least to the fusion filter 5, in this
case in particular the information that the rotation rates about
all three axes have the value zero and at least one position change
variable likewise has the value zero and in particular the
velocities along all three axes have the value zero. The standstill
identification unit 8 is in this case designed, in accordance with
the example, in such a way that it uses the wheel rotation speeds
or wheel rotation speed signals as input data and the "raw" or
direct output signals of the inertial sensor arrangement 1.
[0173] The signal processing device calculates and/or uses, in
accordance with the example, a first group of data of physical
variables, whose values relate to a vehicle coordinate system and
in addition calculates and/or uses a second group of data of
physical variables, whose values relate to a world coordinate
system, wherein this world coordinate system is suitable in
particular at least for describing the orientation and/or dynamic
variables of the vehicle in the world, wherein the sensor system
has an orientation model unit 9, with which the orientation angle
between the vehicle coordinate system and the world coordinate
system is calculated.
[0174] The orientation angle between the vehicle coordinate system
and the world coordinate system in the orientation model unit 9 is
calculated at least on the basis of the following variables:
[0175] the velocity with respect to the vehicle coordinate system,
the velocity with respect to the world coordinate system and the
steering angle.
[0176] The orientation angle between the vehicle coordinate system
and the world coordinate system is calculated, in accordance with
the example, in the orientation model unit 9 additionally at least
on the basis of one or more of the following variables:
[0177] orientation information of the vehicle based on the world
coordinate system,
[0178] same or all of the correction values and/or variances of the
fusion filter and/or
[0179] the acceleration of the vehicle in relation to the vehicle
coordinate system and/or the world coordinate system.
[0180] The orientation model unit 9 uses some or all of the output
data and/or output signals of the strapdown algorithm unit 2 for
the calculation.
[0181] The orientation model unit 9 is designed, in accordance with
the example, in such a way that it calculates and provides, in
addition to the orientation angle, also information on the data
quality of this variable, in particular the variance of the
orientation angle, wherein the orientation model unit 9 provides
the orientation angle between the vehicle coordinate system and the
world coordinate system and the information on the data quality of
this variable to the fusion filter 5, and the fusion filter uses
this orientation angle in its calculations and particularly
preferably passes on the information on the data quality of this
variable, in particular the variance of the orientation angle, to
the strapdown algorithm unit 2.
[0182] While the above description constitutes the preferred
embodiment of the present invention, it will be appreciated that
the invention is susceptible to modification, variation, and change
without departing from the proper scope and fair meaning of the
accompanying claims.
* * * * *