DE102006009682A1 - Dual-tracked vehicle`s driving condition determining method, involves using tire or wheel forces in vehicle-transverse direction, direction of vehicle-vertical axis and direction of longitudinal direction as value measured at vehicle - Google Patents

Abstract

The method involves estimating a side slip angle by a mathematical model on the basis of an actual value measured at a vehicle and by an observer valuation based on the model. Tire forces or wheel forces in a vehicle-transverse direction, a direction of vehicle-vertical axis and a direction of vehicle-longitudinal direction are used as the value measured at the vehicle. Yaw acceleration of the vehicle is determined by formation of moment equilibrium around the vehicle-vertical axis in consideration of an estimated moment of inertia of the vehicle.

Description

The invention relates to a method for determining the driving state of a two-lane vehicle by estimating the slip angle via a mathematical model based on values currently measured on the vehicle and by means of an observer approach based on this model. The prior art is by way of example on the DE 43 25 413 C2 directed.

at in series driving dynamics control systems for motor vehicles (For example, the known ESP) to stabilize appropriate control interventions of the vehicle derived from a driving state calculation, the appropriate taken into account on the vehicle. In particular, these measured quantities are the Driving speed in connection with the respective wheel speeds, as well as the vehicle lateral acceleration and the yaw rate when cornering, given to the driver Steering angle under consideration the travel speed are related. It is measured often also the longitudinal acceleration of the vehicle.

generally known would be at least for usual driving conditions, i. those where the vehicle wheels are still have a certain grip, a relatively accurate statement about the current driving condition possible, if the slip angle of the vehicle would be known with sufficient accuracy. A Measuring the slip angle of a two-lane vehicle is for one Large-scale production not possible yet. estimation methods for determining the float angle are known in principle (see. Writing), but these are usually not sufficiently accurate or need too much computing capacity, so for that reason a reliable float angle estimation in Real time not possible is. In particular, driving situations with slow slip angle build-up, with the influence of external forces (wind, Buoyancy forces Friction) and the finding of different coefficients of friction per wheel or axis place high demands on the calculation algorithms of the slip angle. Today's metrics do not give all the necessary Information again, necessary in the situations mentioned are to be able to conclude clearly on the driving condition.

Here Now to show an improvement is the task of the present Invention.

The solution this task is for a method according to the preamble of claim 1, characterized that the values measured on the vehicle are the tire forces or wheel forces at least in the vehicle transverse direction, preferably also in the direction the vehicle vertical axis and in the vehicle longitudinal direction, be used, further at least one steering angle for both Wheels of the Front axle and possibly the rear axle, and further in particular to calibrate the observer approach the measured yaw rate or yaw acceleration as well as the longitudinal speed of the vehicle. The latter can, as usual from the wheel speeds be determined.

The Accuracy of an estimation method for the Vehicle slip angle can be determined with specific measured values for the the wheels or tires of the vehicle prevailing or applied forces against the the prior art has been significantly increased. Especially can through the use of actual and thus more accurate wheel forces on others Place of the estimation procedure Simplifications are made, which lead to the shortening of the calculation times and still provide sufficiently accurate results. Further, a ongoing comparison of the observer approach, at least with regard to the wheel forces acting on the wheel possible, the additional as basically known from the respective wheel-individual normal force (in the direction the Fzg. Hochachse) as well as over the respective wheel-specific slip angle using known tire parameters can be determined. The same applies to the yaw rate or the yaw acceleration of the vehicle when measured both or by temporal differentiation from a measured value gewonnnen as well as by forming a moment equilibrium around the Vehicle vertical axis under consideration the suitably determined mass moment of inertia of the vehicle is determined. This moment of inertia can this several times be determined by iterative improvement, by the formation one of the measured wheel forces derived moment equilibrium around that through the vehicle's center of gravity extending vehicle vertical axis determines the current yaw moment and this measured by or arising from a temporal Derivative of the measured yaw rate resulting yaw acceleration divided becomes. This is in a filed on the same day another patent application the applicant with the title "vehicle dynamics control system a two-lane vehicle ", to hereby expressly Reference is made and their content by reference also content of the present application.

Furthermore, a preferred vehicle model and a preferred observer approach will be closer wrote. The symbols used here are listed at the end of the description, and the accompanying figures (fig.) 1 - 4 directed.

The vehicle is described using a mathematical, parametric and time-invariant model. The model should include the following properties: A rigid vehicle body with mass m, width b, length l, and four tires with saturation behavior and a tire side force function = f (α, κ F _z ). In modeling, the in 1 coordinate systems and dimensions used.

For the notation of the mathematical model, a general, non-linear state space representation is used, which is defined by the following equations (1):

This includes process disturbances d _Sys and measurement noise d _meas .

In 2 the selected inputs u , the states z and the outputs y of the vehicle model are listed.

sowie

The slip angles at the front axle or at the rear axle of the vehicle are defined as:

such as

For the slip angles of the wheels assumed to be the same for the right and left wheels of an axle, the following applies: α V = δ V - β V (Equation 4) α H = δ H - β H (Equation 5)

The lateral force of the tire is inter alia a function of the tire normal force as well as the slip angle and is approximated by the following equation. Combined slip states are initially not considered. The following tire model with saturation behavior and the basically known tire parameters k _ia and k _ib can be used:

=> KFix = cosδi·RFix – sinδi·RFiy (Gleichung 8) => KFiy = sinδi·RFix + cosδi·RFiy (Gleichung 9) Thus, the tire lateral forces are present in wheel coordinates and must be transformed into the body-fixed vehicle coordinate system in the next step:

=> K F ix = cosδ i · R F ix - sinδ i · R F iy (Equation 8) => K F iy = sinδ i · R F ix + cosδ i · R F iy (Equation 9)

With the wheel forces in the coordinate system K of the vehicle then the resulting forces on the previously based on the measured wheel forces easily determinable by the current vehicle focus Force sums are determined in the appropriate direction. Here, the other externally acting forces such as. The friction and aerodynamic forces can be neglected.

On the basis of the auxiliary quantities from equations (2) to (10), the equations for the state vector z (cf. 2 ) are formed. For the first component z ₁ = ψ, the moment equilibrium is formed around the z-axis (= vertical axis) of the vehicle coordinate system K for this purpose:

This results in the yaw acceleration as the first component of the vector z 2 to:

For the second component of the vector z, which (cf. 2 ) is the slip angle β, the set of center of gravity is set up in track coordinates in the vehicle transverse direction: m · B a Cy = -Sinβ K F cx + cosβ K F Cy = (β. + ψ.) · B v cx (Equation 13)

This results in the slip angle to:

The third component of vector z 2 results from the center of gravity set in the vehicle coordinate system in the longitudinal direction m · K a cx = K F cx = m · ( K v. cx - ψ. K v Cy ) With K v Cy = tan K v cx (Equation 16) to

The equations (11), (13) and (17) thus form the system function f = [ψ ..β. K v. cx ] T (Equation 18)

The output of the model contains the states ψ, _K v _Cx and the transverse forces _R F _iy on all tires and can be given by means of equations (2) to (9) as a function of the input u and the output y : H = [ψ K v cx K F 1y K F 2y K F 3y K F 4y ] T (Equation 19)

Hereupon, an observer approach is applicable in a manner known per se. This observer receives the measured inputs u and outputs y of the system and uses the mathematical model to determine the system state ẑ . 3 illustrates this process. The real system is affected by faults, which are subdivided into process faults d _sys and measurement noise d _meas .

K = P·CT·Q–1Mess (Gleichung 21) P . = A·P + P·AT – P·CT·Q–1Mess ·C·P + QSys (Gleichung 22) The observer chosen here, for example, is a nonlinear, time-continuous Kalman filter, which is referred to in the literature as the Extended Kalman Bucy Filter. The system state z (t) is calculated using the following equations:

K = P · C T · Q -1 measuring (Equation 21) P. = A · P + P · A T - P · C T · Q -1 measuring · C · P + Q Sys (Equation 22)

Except for the constant chosen variance matrices Q Sys = E {d Sys · d T Sys } and Q measuring = E {d measuring · d T measuring } are all quantities in the above equations (20) - (21) functions of time. The matrices A and C are Jacobian matrices and are calculated from the nonlinear state and output functions as follows:

In order to calculate the matrix P from the differential equation (22) for the Kalman gain K, an initial value P ₀ must be chosen. The simulation shows that the same matrix P always results for different values of P ₀ at the end of the simulation, so that the choice of P _{0 has} no appreciable influence on the result of the estimation. It is therefore possible to choose the zero matrix for P ₀ .

Inside, the general Extended Kalman Bucy Filter has the in 4 shown structure. From the determined system state results in an output, which is compared within the filter with the true, ie determined by measurement variables, namely the Rad-lateral force (in y-direction) and the yaw rate and the longitudinal velocity. The respective difference between the estimated value and the actual value is used to suitably correct the previously calculated state.

Since in the present model the inputs do not act directly on the outputs, as can be seen from the output equation (19), the input u in the output function h ( z (t), u (t), t) is omitted.

A comparison of the slip angle determined according to the described approach with float angles suitably measured in tests in specific driving situations revealed a high degree of agreement. In particular, in accordance with equations (4) and (5), a slip angle for the front axle or for the rear axle of the vehicle can then be determined from the float angles for the front axle or rear axle of the vehicle determined by the described approach, and the driving condition can be further determined therefrom , If the front axle slip angle is greater than the rear axle slip angle, then understeer is present; otherwise (α _H > α _v ) oversteer. A precisely known float angle is basically advantageous for a large number of systems in the vehicle which have an influence on the driving characteristics. The more accurately the float angle is known, the better z. For example, the vehicle dynamics control systems are able to place their interventions at the right time and in the right place.

Symbols:

• (Rarely used symbols and terms as well as different Meanings are explained in the text).

General spellings:

• a ⇀

  vector

  a

  column matrix

  A

  matrix

  _K a _BC

  Local column matrix from the point B to C, indicated in the coordinate system K

  ^IK A

  Transformation matrix transformed the column arrays from the coordinate system K into the system I.

  a ^

  Estimated size

Latin notation:

• t

  Time

  v

  Speed [m / s]

  _R F _ix

  Power on the wheel No. i in the x-direction of the coordinate system R, i = 1,..., 4

  m

  Vehicle mass [kg]

  J _zz

  Mass moment of inertia of the vehicle about the vertical axis z [kgm ² ]

  l

  Wheelbase [m]

  b

  Gauge [m]

  l _v

  Distance of the front axle to the main focus [m]

  l _H

  Distance of the rear axle to the main focus [m]

  b _R

  Distance of the right Wheels to Focus [m]

  b _l

  Distance of the left Wheels to Focus [m]

  z

  Status

  f

  state function

  H

  output function

  u

  entrance

  y

  output

Greek notation:

• ψ

  Yaw angle [rad]

  β

  Swing angle [rad]

  δ

  Steering angle [rad]

  α

  Slip angle [wheel]

Abbreviations:

• V

  ahead

  H

  behind

  R

  Right

  L

  Left

Coordinate systems (cf 1 ):

• R

  wheel coordinate

  K

  Body-fixed vehicle coordinate system

  B

  Railway coordinate system x-axis along the railway tangent

Claims (3)

Method for determining the driving state of a two-lane vehicle by estimating the slip angle via a mathematical model based on values currently measured on the vehicle and using an observer approach based on this model, characterized in that the tire forces or wheel forces, at least in the vehicle transverse direction, are measured as values measured on the vehicle, Also preferably used in the direction of the vehicle vertical axis and in the vehicle longitudinal direction, further at least one steering angle for the two wheels of the front axle and optionally the rear axle, and further to adjust the observer approach the measured yaw rate or yaw acceleration and the longitudinal velocity of the vehicle. Method according to claim 1, characterized in that that the yaw acceleration of the vehicle by forming a moment equilibrium around the vehicle's vertical axis, taking into account the suitably estimated mass moment of inertia of the vehicle is determined. Method according to one of the preceding claims, characterized characterized in that by comparing the slip angle at the front axle the vehicle with those on the rear axle on understeer or oversteer is closed.