SIMPACK User Meeting 2014

October 8 – 9, Augsburg, Germany

Simulation of Vehicle Dynamics Control by

active Steering Systems
Volker Dorsch, Faculty of Mechanical Engineering
 Overview

● Development scenario
● Validated mechanical multibody simulation (MBS) vehicle model
● Implementation of active steering at rear and front wheels
● Control strategy
● Co-Simulation of Simpack and Matlab/Simulink®
● Simulation Results
● Conclusions

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -2-
 Development Scenario

Vehicle Model Controller-Function Model

MiL SiL

Hi
L
Validation Code

Real Vehicle Hardware (ECU, Network)

Real System

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -3-
 MBS Model: Coordinates

Input:
δV – front steering angle
δH – rear steering angle
(v - velocity)
v - velocity

Output:
.
ψ – yaw rate
β – side slip angle β – side slip angle
ay – lateral acceleration

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -4-
 Validation

Suspensions: simulation of Kinematics and Compliance (K&C) measurements

Complete vehicle model: step steering input maneuver

driver objective, no driver influence

velocity 100 km/h
front wheel steering sudden steering angle with more than
angle 200 deg/s
steady-state lateral acceleration is 0.4 g
road surface dry, µ = 0.9

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -5-
 Simpack model validation:
Step Steering Input

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -6-
 Simpack model validation:
Step Steering Input

good correlation – realistic response of vehicle model

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -7-
 MBS Model: steering systems

Active front and rear

wheel steering

Simple implementation by


additional rotational degrees of
freedom at the corresponding
wheels

control of these degrees of
freedom by external input
of the co-simulated
Matlab/Simulink controller
model

Study of effects on vehicle dynamics

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -8-
 Vehicle stabilization by adjusted
steering angles
Active rear wheel steering:
e

e
rc

rc
adjust tire slip angle α by rear
fo

efo
e
sid

wheel steering angle

sid
ire

e
tir
et

se
as

ea
re

cr
inc

de

Side force FS
e
e

rc
rc

fo
efo

e
sid
sid

Slip angle α
ire
e
tir

et
se

as
ea

re

Active front wheel steering:

cr

inc
de

adjust tire slip angle α by

superimposed steering angle

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -9-
 Control algorithm: bicycle model

Equations of motion with linearized tire behavior: FS = Cα α

C αH l H C αV l V C α H l 2H +C αV l V2 C αV l V Cα H lH
Ψ̈ = β Ψ̇ + δV δH
Jz J zv Jz Jz
C αV +C α H C l C l C C
β̇ = β +[ α H H 2 αV V 1] Ψ̇ + αV δV + α H δ H
mv mv mv mv

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -10-
 Control algorithm: bicycle model

In case of steady-state cornering (i.e. Ψ̈=0 and β̇=0 ) without rear wheel angle
v x δw
Ψ̇= 2
vx
(l V +l H )(1+ 2 )
v ch

The characteristic velocity vch can be computed (or be identified in tests)

2
C αV C αH l
v ch =
m(C α H l H C α V l V )

Furthermore the maximum yaw rate is limited

a y ,max
∣̇∣
vx

The yaw rate Ψ̇ is used as desired yaw rate in the control algorithm.

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -11-
 Control algorithm: predict rear wheel
steering angle

Setting β̇=0 and β=0 in the differential equations of the bicycle model yields

δH (s) 1+T z s C αV C α H l H l C α V l V m v 2
=K δ with Kδ = ,
δ V (s) H
1+T 1 s H
C αV C α H l V l +C α H l H m v 2
Jzv
T z= ,
C α H l H l l V mv 2
Jzv
T 1= 2
C αV l V l l H m v
[Woernle, C.: Fahrmechanik.
Lecture notes,
University of Rostock]

The steady-state case with KδH results in a characteristic diagram,

that is used as a first prediction of the rear wheel steering angle.

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -12-
 Control algorithm: predict rear wheel
steering angle

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -13-
 Active rear wheel steering:
Control structure with co-simulation

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -14-
 Active front wheel steering:
Control structure with co-simulation

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -15-
 Steering angle orientation

Active rear wheel steering

∣Ψ̇ des∣ ∣Ψ̇∣<0 oversteer ∆Ψ

∣Ψ̇des∣ ∣Ψ̇∣≈0 neutral

∣Ψ̇ des∣ ∣Ψ̇∣>0 understeer neutral?
no

no
oversteer?
Ψ̇>0 left turn
Ψ̇<0 right turn yes

yes increase side force decrease side force

Tolerance function to avoid hard increase tire slip angle decrease tire slip angle

switch on – switch off effect yes

no no
left turn? left turn?

100% yes yes

no steering + − + −
angle increment steering angle steering angle
change change
full control no control full control ∆Ψ

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -16-
 Co-Simulation: Input - Output

u-vector:
y-vector:
δH, ∆δV
- velocity
- yaw rate
- δH, δV

Co-simulation:
- Simpack and Matlab/Simulink® run parallel
- Data exchange at each millisecond
- Relatively slow motion of the whole vehicle: uncritical

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -17-
 Maneuver steady-state cornering

Driver Simpack driver model, driver

influence
Velocity increasing up to 75 km/h
Track Circle with 80 m radius
Steering closed loop, by driver
wheel
angle input
Road dry, µ = 0.9
surface

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -18-
 Steady-state cornering: simulation results

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -19-
 Steady-state cornering: simulation results

Change of cornering behavior by change of bicycle model

parameters in the control algorithm

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -20-
 Maneuver step steering input

Driver objective, no driver influence

Velocity 80 km/h
Steering sudden steering angle step with
wheel more than 200 deg/s,
angle input steady-state lateral acceleration
is 0.4 g
Road dry, µ = 0.9
surface

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -21-
 Step steering input: simulation results

Peak reduction

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -22-
 Step steering input: simulation results

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -23-
 Maneuver sine with dwell

Driver objective, no driver influence

Velocity 80 km/h
Steering single sine of 0.7 Hz with dwell
wheel of 500 ms after ¾ of period,
angle input steady-state lateral acceleration of
amplitude is 0.4 g
Road dry, µ = 0.9
surface

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -24-
 Sine with dwell: Simulation results

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -25-
 Sine with dwell: Simulation results

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -26-
 Sine with dwell: Simulation results

Active rear wheel steering seems advantageous

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -27-
 Sine with dwell on slippery road

Driver objective, no driver influence

Velocity 80 km/h
Steering single sine of 0.7 Hz with dwell
wheel of 500 ms after ¾ of period,
angle input steady-state lateral acceleration of
amplitude is 0.4 g (for
vehicle without control)

Road slippery, µ = 0.35

surface

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -28-
 Sine with dwell on low µ: Simulation results

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -29-
 Sine with dwell on low µ: Simulation results

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -30-
 Sine with dwell on low µ: Simulation results

Active steering makes this maneuver driveable,

rear wheel steering is most effective

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -31-
 Maneuver slalom

driver Simpack driver

model, driver
influence
velocity 90 km/h
front wheel steering closed loop, by
angle driver
road surface dry, µ = 0.9

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -32-
 Slalom: Simulation results

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -33-
 Slalom: Simulation results

Less oscillations

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -34-
 Slalom: Simulation results

Rear wheel steering is advantageous

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -35-
 Conclusion and Outlook


Validated mbs vehicle model as a basis

Simple implementation of active steering systems by additional dofs:
rear wheel steering, front wheel steering

Co-Simulation of Simpack and Matlab/Simulink® for integration of
control systems

Systems work effectively, effects on vehicle dynamics can be
studied, active rear wheel steering seems more powerful


Optimization of both systems still needed (optimize control
parameters, take side slip limit and wheel slip into account)

Mechanical design of rear wheel steering and
implementation of an actor model

Implementation of a steering actor model in case of
front wheel steering

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -36-
 Acknowledgment

The author would like to thank Yangfang Yu for doing a lot of excellent modeling
and simulation work.

Volker Dorsch: Simulation of Vehicle Dynamics Control by active Steering Systems -37-