Safety and Performance Enhancement: The Bosch Electronic

Safety and Performance Enhancement: The Bosch Electronic
Safety and Performance Enhancement:
The Bosch Electronic Stability Control (ESP)
Product (GDP) [1]
E. K. Liebemann
K. Meder
J. Schuh
G. Nenninger
Robert Bosch GmbH
Paper Number 05-0471
In spite of improvements in passive safety and efforts to alter
driver behavior, the absolute number of highway fatalities in
2002 increased to the highest level since 1990 in the US.
ESP is an active safety technology that assists the driver to
keep the vehicle on the intended path and thereby helps to
prevent accidents. ESP is especially effective in keeping the
vehicle on the road and mitigating rollover accidents which
account for over 1/3 of all fatalities in single vehicle
In 1995 Bosch was the first supplier to introduce electronic
stability control (ESC) for the Mercedes-Benz S-Class sedan.
Since then, Bosch has produced more than 10 million systems
worldwide which are marketed as ESP - Electronic Stability
the world, the development of the mobility shows a clear
correlation to the gross domestic product (Fig. 1). With
further economical growth, we will see more increase in
mobility and in traffic density throughout the world. This will
require additional efforts to furthermore enhance the road
The statistics for the European Union demonstrate alarming
results. They show a total of 1.3 million accidents for the year
2000 with 1.7 million injured persons and more than 40.000
fatalities. The target of the eSafety Initiative of the European
Union for 2010 is set to reduce road deaths by 50%, e.g. by
the promotion of intelligent active driving safety systems (Fig.
In this report Bosch will present ESP contributions to active
safety and the required adaptations to support four wheel
driven vehicles and to mitigate rollover situations.
- 50%
Target of the eSafety Initiative of
the European Union for 2010:
Reduction of road deaths by
50% by the promotion of
intelligent active driving-safety
Worldwide traffic is increasing with more and more vehicles
on the road. Considering the different regions of
Traffic safety situation European Union (status year 2000):
1 300 000 accidents, 40 000 deaths, 1 700 000 injured
Traffic performance/inhabitants [pkm]
Fig. 2: European eSafety initiative
North America
Eastern Europe
AUS, NZ, Japan
Latin America
South-East Asia
Japan has set a similar target and also NA is actively pursuing
advances in road safety.
EU (with CH, N)
South Asia
Middle East /
North Africa
East Asia (China)
GDP/inhabitants, US$ (1980)
The progress of crash energy absorbing car body design and
the standard fitting of airbags significantly improved the
passive safety especially combined with the use of seat belts.
But many of the serious accidents happen through loss of
control in critical driving situations. When the vehicle goes
Fig. 1: Development of mobility depending on Gross Domestic
into a skid, a side accident is the frequent result. With a
reduced protection zone for the occupants compared to front
crashes, these accidents show an amplified severity.
Especially with vehicles of an elevated center of gravity like
sport utility vehicles (SUV) and light trucks (LT) the loss of
control with subsequent skidding may even lead to a rollover.
Most of the rollovers are caused either by tripping at an
obstacle or in the soil. The severity of rollover accidents is
extremely high. Accounting for only 2% of the total crashes,
they contributed in 2002 with 10.656 fatalities to one third of
all occupant fatalities (Fig. 3) in the US.
US Accident fatality statistics
Total Accidents
Involved Vehicles:
Occupant Fatalities:
10.6 Mio
Point of Impact
Severity (by fatalities)
Frontal crash:
46 %
Frontal crash:
39 %
Side crash:
29 %
Side crash:
23 %
33 %
Fig. 3: North America accident fatality statistics
A study performed by the University of Iowa at the National
Advanced Driving Simulator showed a strong impact of ESP
on vehicle stability [2]. The primary question was “Does the
presence of an ESP system aid the driver in maintaining
control of the vehicle in critical situations?”. Based on all
analyses completed there was a 24.5 percentage point
reduction between situations in which the drivers lost control
with the system present and situations without ESP. This
constitutes an 88% reduction in loss of control. Looking at the
data from an improvement standpoint, 34% more drivers
retained control with ESP than without. Based on the study
results it was concluded that there is significant and
meaningful safety benefit associated with driving a vehicle
equipped with an ESP system.
The results of the studies show a consistent picture of the ESP
with remarkable safety benefits. Further potential is available
especially with functional extensions for SUV and light trucks
concerning rollover mitigation and four wheel drive
However it is important to say that ESP cannot prevent all
accidents or adjust for all driver errors. Essential for a safe
road traffic are still appropriate driving practices, common
sense and a good traffic judgement.
In critical driving situations most drivers are overburdened
with the stabilizing task. According to Foerster [4] the average
driver can neither judge the friction coefficient of the road nor
the grip reserves of the tires. The drivers are typically startled
by the altered vehicle behavior in in-stable driving situations;
as a result, a well-considered and thought-out reaction of the
driver can not be expected. For that reason the ESP has to be
designed to stabilize the vehicle even in situations with panic
reactions and driving failures like exaggerated steering.
The reason why stabilizing a vehicle in critical situations is so
challenging can be shown by considering the physical effects.
Steering of a vehicle yields in a yaw moment which results in
a directional change. The effect of a given steering angle
depends on the actual side slip angle [5, 6]. Only slight
alterations of the yaw moment are possible at large side slip
angles even for extensive steering interventions which can be
seen in Fig. 4.
The characteristic side slip angles, where the steerability of
the vehicle is vanishing, are dependent on the road friction
coefficient. On dry asphalt it is around ±12° as shown in Fig.
4, whereas on polished ice it is in the range of ±2°. The driver
experiences in all day traffic situations side slip angle values
of typically not more than ±2°.
Supporting conclusions are drawn by VW [1]. Based on their
accidentology, ESP is considered to avoid 80% of the
accidents caused by skidding. VW concludes that the safety
benefit of ESP is even greater than that of the Airbag.
According to VW a 100% installation rate would result in
Germany in a 20% reduction of road fatalities and this even
with an ESP installation rate of already 53% in 2003.
Based on the analysis of traffic accidents statistics, Toyota [3]
estimated that the accident rate of vehicles with ESP for more
severe accidents is approximately reduced by 50% for single
car accidents and reduced by 40% for head-on collisions with
other automobiles. The casualty rate of vehicles with ESP
showed approximately a 35% reduction for both types of
Fig. 4: Influence of side slip angle on yaw moment for different steering
angles at high tire-road friction [5, 6].
Special adaptations of the ESP system and the control concept
are required for the cooperation with a four wheel drive
(4WD) power train.
So one of the main tasks of ESP is the limitation of side slip
angle dependent on the actual friction coefficient.
Even in the range of characteristic side slip angles, where the
effectiveness of steering is rather limited, ESP can exercise
remarkable yaw moments by brake interventions. The tire
characteristic determines the longitudinal slip value 0 where
the maximum brake force is generated. The slip value 0 is
typically in the range
of 10%. Considering
the left front wheel
cornering (Fig.
wheel 1), the resulting
wheel force in free
FR( =0) is in lateral
direction. By adjusting
the tire slip to Λ0, the
maximum brake force
FB( 0) is applied and
by this means the
lateral force is reduced
FS( 0).
resulting force vector
FR( 0)
relative to the tire
thereby modifying the
longitudinal and the
lateral forces.
Several center coupling concepts are used in the various types
of four wheel driven vehicles. Most of them can be combined
with an ESP system.
The major element of a four wheel driven (4WD) vehicle is
the center coupling. The objective is to distribute drive torque
to the front and rear axle and at the same time to permit
different axle velocities that occur as soon as the vehicle
drives around a bend (Fig. 6).
Engine Management
Differential Lock Management
Brake Management
Engine interventions
Lock interventions
MCar, RA , vCar, RA
MCar, FA , vCar, FA
Fig. 6: Control concept for four wheel drive trains with engine (1), center
coupling (2), brake (3), differential front/rear axle (4/5).
The classic solution for a 4WD drive train is the open center
differential. Its disadvantage is - analogous to a transversal
axle differential - the drive torque limitation of an axle if the
other one shows increased slip. In the worst case a 4WD car
with an open center differential does not move if only one
wheel is spinning.
Fig. 5: Turning of resulting wheel force by tire slip control.
The required yaw moment can be applied by controlling the
longitudinal tire slip and in that way employing it as a vehicle
dynamics control variable. This approach is utilized with
anti-lock and traction slip control, yaw rate control with
restricted side slip angle and with a limitation of lateral
acceleration for rollover mitigation functionality.
With an ESP system available, this drive train concept can be
supported by the brake interventions of the traction slip
control without the necessity to install additional longitudinal
and transversal lock devices (Fig. 7). The longitudinal
differential lock controller in the ESP restrains the difference
speed between both axles through a symmetric brake
intervention on both wheels of one axle. The transversal
differential lock controls the difference speed on one axle
through wheel individual brake interventions.
During the last few years the segment of four wheel driven
vehicles got more and more popular. The main focus of
attention is the range of SUV and LT vehicles that are suitable
for use on public roads but also have qualities under off-road
conditions. Part of the off-road capacities are due to the
elevated center of gravity which augments the susceptibility to
rollover. This makes SUV and LT the preferred target for ESP
M RR ,v RR
MCar , v Car
Transversal differential
Transversal differential
Trans. lock controller
Trans. lock controller
MCar, FA , vCar, FA
MCar, RA , vCar, RA
Fig. 7: Four wheel drive with longitudinal and transversal brake lock
Another class of differential locks or center couplings are selflocking devices, where the locking degree depends on torque
or rotation speed differences between the two driven axles.
Examples are Torsen - for Torque-sensing - or viscous
coupling. If their locking potential is exceeded, the above
described longitudinal differential lock via brake intervention
will support and secure the lock functionality.
A 100% mechanical differential lock is useful for heavy offroad applications, as it prevents any axle speed differences.
Since ESP relies on a wheel individual slip control, a
cooperation with a mechanically locked center differential is
not feasible unless the lock is opened either manually or
electronically. Even anti-lock control (ABS) is deactivated or
distinctively reduced.
Apart from the mentioned devices that have a system inherent
locking effect, there are center couplings that can be fully
influenced by an external controller – so called Center
Coupling Control (CCC). In this case an electric or hydraulic
actuator operates a clutch, providing adjustable locking
torque. In combination with vehicle dynamics signals, as
vehicle speed and wheel speeds, yaw rate, lateral acceleration
and engine torque, the locking torque can be adjusted to tune
to the desired vehicle dynamics behavior suitable for the
specific driving conditions (Fig. 8).
Fig. 8: Influence of drive torque distribution on vehicle dynamics behavior
like over-steering and under-steering. Shown is the maximum possible
flexibility of drive torque distribution; actual flexibility depends on drive
train configuration.
Even in critical driving situations the variable drive torque
distribution can positively influence the road behavior of the
vehicle. By shifting drive torque to the rear axle, the understeering behavior of a vehicle can be reduced; by shifting
drive torque to the front axle, the over-steering behavior can
be trimmed down (Fig. 8). Overall a more responsive vehicle
handling can be achieved.
The ESP is well suited to extend the brake and engine torque
interventions with a center coupling torque interface to
optimize the dynamic behavior of the vehicle. One example is
shown in Fig. 9. The ESP detects an understeering situation
and requests a reduction of the coupling torque transferred to
the front axle. Beside this drive torque transfer an additional
ESP brake intervention on the curve inner rear wheel supports
in case of strong understeering to achieve the desired vehicle
yaw rate.
start of understeering
desired ...
circumference speed. Alternatively the driver may select the
off-road adaptations via a switch setting, the activation of a
countershaft gearbox or the vertical adjustment of a level
control system.
end of understeering
...and actual vehicle
yaw rate
start of CCC intervention
against understeering
target coupling
torque ...
...and actual coupling
torque to front axle
start of brake intervention
against understeering
(pressure on right rear wheel)
500 ms
Fig. 9: Understeering intervention with a shift of drive torque followed by a
supporting brake intervention. Sporty SUV vehicle with 4WD, center
coupling control and ESP8.
For vehicle dynamics and traction optimization a controllable,
well defined opening and closing of the coupling is necessary.
On the other hand, during a wheel individual brake
intervention, a fully or partially locked center coupling would
result in an unintended torque transfer. Therefore a fast
opening must also be demanded during stabilizing brake
interventions and an active ABS function. In some instances,
it may also be necessary during partial braking to allow the
“Electronic Brake Distribution” function to prevent the overbraking of the rear axle. This requires the clutch to be opened
in less than 100ms.
Additional adaptations support off-road functionality. The
off-road features of the ESP controller improve robustness
and maintain superior traction under off-road conditions.
These features are:
Adaptation of start of control thresholds for vehicle
dynamics under off-road conditions; increased yaw rate
target allowed.
Self tuning of traction target slip dependent on the road
surface and terrain.
Lessening of engine torque reductions to maintain
traction even under difficult drive conditions.
Adaptive pre-control for the brake torque controller.
Enhanced vehicle speed estimation under off-road
conditions even without use of longitudinal acceleration
Robustness measurements for the ABS controller with
increased target slip under off-road conditions.
The off-road situation can be detected automatically by a
special function of the ESP. Based on wheel speed sensor
signals, the off-road detection function analyses wheel
excitations and looks for specific oscillations in the wheel
In powerful ESP systems for 4WD vehicles, even different
performance settings can be selected by the driver. This can
be as simple as disabling the engine torque reduction triggered
by the ESP to allow for full driver control of the propulsion.
Other possibilities are terrain specific adaptations to surfaces
like ice, snow, grass, sand, mud or bedrock.
Some drive train concepts allow a flexible configuration by
switching from rear wheel drive or front wheel drive to 4WD.
Even 4WD with locked center differential is possible. With a
cooperating ESP system, the stabilizing and traction control
functionality can be automatically adjusted to the selected
drive train concept.
In cooperation with four wheel drive train concepts, ESP
delivers the expected safety benefits and excellent off-road
functionality. Since most of the respective vehicles are
characterized by an elevated center of gravity, road safety can
be further improved by implementing rollover mitigation
The complex events of automobile crashes involve three main
contributing factors and their interactions [7]:
the driver,
the driving environment like weather, road condition,
time of day,
and the vehicle.
In the US, about 10% of all road accidents are non-collision
crashes, but approximately 90% of such single-vehicle
crashes account for fatalities [8]. The SUV and LT with their
elevated center of gravity (CoG) show an amplified rollover
propensity. This is reflected in their increased rollover rates.
Due to the ever increasing popularity of these vehicles, the
percentage of fatal rollover crashes escalated significantly
within the last decade.
A vehicle rollover occurs when the lateral forces create a large
enough moment around the longitudinal roll axis of the
vehicle for a sufficient length of time.
Critical lateral forces can be generated under a variety of
conditions. The vast majority of rollover crashes take place
after a driver lost control over the vehicle. By skidding off the
road, the vehicle may get in lateral contact with a mechanical
obstacle like a curb, a pot hole or a plowed furrow which
yields a sudden large roll moment. This results in a so called
tripped rollover in contrast to an un-tripped or friction
rollover. The latter takes place on roads during severe steering
maneuvers solely as a result of the lateral cornering forces.
Although the ratio of un-tripped to tripped rollovers is small,
the un-tripped rollovers account for the most severe crashes.
Accident analysis has shown that the ratio of the track width T
and the height of the center of gravity hCoG gives a first
indication for the rollover propensity of vehicles.
2 ⋅ hCoG
Static Stability Factor
The SSF is an important parameter affecting vehicle rollover
risk and is both relevant for tripped as well as un-tripped
rollover. The track width is a fixed parameter while the center
of gravity height varies with subject to different load
conditions. Through a one rigid body model - which means
no distinction between the mass of the chassis and the sprung
mass of the vehicle body – the SSF relates geometrical vehicle
data to the level of lateral acceleration that will result in a
A one rigid body model cannot predict time dependent details
of an on-road rollover critical situation. For transient
maneuvers involving high lateral accelerations, many vehicle
design parameters have an effect on the vehicle handling
behavior like e.g. front to rear roll couple distribution, roll
axis location, tire behavior, suspension characteristics and roll
resonant frequency. These handling characteristics
significantly influence the ability of the driver to maintain
control in an emergency situation.
To assess a vehicle’s handling performance with reference to
rollover, the SSF is complemented by metrics derived from
dynamic testing which can be partially influenced by
electronic stability control. In the US, beginning with the
rollover ratings for model year 2004, the National Highway
Traffic Safety Administration (NHTSA) will combine the SSF
measurement of the vehicle with the dynamic performance in
the so-called fishhook or road edge recovery maneuver [8].
New criterion:
in terms of SSF:
**** <= 0.1
>= 1.4532
**** in [0.1; 0.2]
in [1.1764; 1.4531]
*** in [0.2; 0.3]
in [1.0743; 1.1763]
in [0.3; 0.4]
in [1.0194; 1.0742]
> 0.4
<= 1.0193
> 1.45
in [1.25; 1.44]
in [1.13; 1.24]
in [1.04; 1.12]
< 1.03
Fig. 10: NHTSA star rating in case of a positive dynamic test compared
with the previously static SSF rating only. Table derived from [8].
If the Fishhook test is passed successfully due to a highly
effective vehicle stabilizing system, the corresponding
Rollrate may result in a better NHTSA star rating compared
with the static evaluation only and more, the rollover risk for
the vehicle is essentially reduced.
The load condition influence on the rollover propensity is
shown in figure 11 in a simplified manner for different types
of cars and loading conditions. The static stability factor for
typical passenger cars is far above the lateral acceleration
which can be transferred by the maximum tire grip. This is the
reason why passenger cars are usually not subject to untripped rollovers even in extreme loading conditions. If the
adhesion limit between the tires and the road surface is
reached before the lateral acceleration gets rollover critical,
the vehicle starts to skid over the front wheels.
The situation is different especially for light commercial
vehicles, where elevated loading may play a major role.
To improve the relationship between the real world rollover
risk and the SSF-based rollover prediction, the NHTSA
defined a new indicator called Rollrate.
RollrateSSF =
1+ e
(c1 + c2 *ln ( SSF −0.90 ))
The parameters c1=2.7546 and c2=1.1814 are derived from a
detailed analysis of U.S. crash data using a logistic regression
Based on the result of the dynamic test, the static Rollrate
value is either increased or decreased. In case of a positive
test result, the Rollrate is evaluated with the parameters
c1=2.8891 and c2=1.1686 based on crash data analysis; for a
failed test, the parameters are c1=2.6968 and c2=1.1686.
Therefore, the dynamic Rollrate replaces the static SSF to get
the star rating for a single vehicle according to the following
table (Fig. 10).
Fig. 11: Typical critical lateral accelerations for rollover dependent on
loading conditions reflecting different types of vehicles
At the physical limit the tire behavior is extremely nonlinear
and the linearized tire-wheel-brake system is even unstable.
As a result, the vehicle may suddenly spin and the driver is
caught by surprise.
Changing the direction of the resultant tire forces of
individual wheels by specific wheel slip demands applies a
stabilizing yaw moment (see Fig. 5) . Besides standard ESP,
active steering can be used as well to increase the vehicle’s
tracking stability [9]. Both concepts mentioned as well as
Active Roll Control [10] or Electronic Damper Control [11]
can in general help to avoid critical situations and as a result
indirectly help to reduce the rollover risk.
The discrete states represent the different defined phases
within highly-dynamical steering maneuvers: one possible set
of discrete states comprises
Besides the classification according to the rollover reason,
rollover scenarios can be divided into highly dynamic
maneuvers, e.g. obstacle avoidance, or quasi stationary
maneuvers like circular driving with steadily increasing
steering wheel angle. The latter can arise while driving on a
highway exit with excess speed.
an Initial state taken if no roll-stabilizing intervention is
a Pre-fill state to apply the brake pads to the brake discs
thereby reducing the pressure build up time,
a Hold state for first turn maneuvers with a high lateral
a Steer-back state with special pre-fill measures for
steering back in highly dynamical maneuvers, and
a Counter-fly state for the second steady steering interval
in multi-directional maneuvers.
The Bosch Rollover Mitigation Functions (RMF) are based on
the standard ESP sensor set and provide a scalable structure
concerning the determination of rollover critical situations and
brake/engine control (Fig. 12). Other solutions additionally
use a roll rate sensor [12].
driv ing
driv e
Fig. 12: Structure of the entire vehicle stabilizing system with the basic
Electronic Stability Program ESP and the Hybrid Rollover Mitigation
Controller( HRMC) with discrete (D) and continuous (C) dynamics parts.
Considering well-known obstacle avoidance maneuvers or
severe steering maneuvers like the NHTSA “Fishhook”, a
classification can be made in
Transitions between the discrete states are essentially
influenced by the driver’s input and the vehicle reaction.
Continuous states vary over time dependent on the discrete
state. They are influenced by continuous inputs like the
steering wheel angle, the lateral acceleration, the yaw rate, the
longitudinal velocity, the body slip angle, and other reference
variables essential for the rollover prediction. Ackermann and
Odenthal propose a rollover coefficient based on the tire
vertical loads [9] which are usually not available in a standard
ESP systems with the required accuracy. The Bosch approach
uses only existing sensor signals and estimated values to
predict the vehicle’s rollover propensity. For example, based
on the well-known single-track model, an early lead for a
subsequent high lateral acceleration is given by
c pre = ψ& ⋅ v x − a y ≈ − β& ⋅ v x
ψ& : yaw rate
a y : lateral acceleration
first turn maneuvers (e.g. J-turn, decreasing radius turn,
first steering input of single or double lane change, or
NHTSA Fishhook),
second turn maneuvers (constant radius turn with
additional steering input, second steering input of a single
or double lane change, or NHTSA Fishhook), and
further turn maneuvers (third or further steering input of
a double lane change or slalom).
Each turn or even a subset of the corresponding time interval
is characterized by a set of typical driver’s inputs as well as a
typical vehicle response. Consequently, each dynamic steering
maneuver can be divided into several time slots which follow
each other in a specific manner. To get an appropriate
stabilization, the controller must provide suitable intervention
strategy and strength for each of the described phases.
This is why for the detection of severe steering maneuvers
and a suitable anti-rollover control, a hybrid dynamical system
is used (Fig. 12). The input, output and state of such a system
is composed of a discrete and a continuous part; the discrete
dynamics D and the continuous dynamics C are connected by
adequate interfaces (for details on hybrid dynamical systems,
e.g. see [13]).
v x : longitudinal velocity
β& : change in body slip angle
With a rapid change of the body slip angle weighted with vx,
the lateral acceleration will heavily increase short after.
The Hybrid Rollover Mitigation Controller outputs derived
from its states are e.g. the brake torque and brake slip values
for the appropriate wheels. The general control strategy is a
fast active brake pressure increase at the curve outside wheels
especially at the front axle initiated by suitable brake slip and
brake torque target values. This reduces the lateral forces as
well as the longitudinal speed of the vehicle and results in an
increased curve radius. Subsequently the track can be
regained due to the reduced speed. In these special situations
the brake intervention is usually combined with a cut back on
engine torque.
In general, the hydraulic braking system must provide a fast
pressure increase over a wide temperature range. For that, the
brake caliper size, the brake tube dimensions, and the
characteristics of the utilized brake fluid are very important.
As an example, a NHTSA Fishhook maneuver with a sporty
SUV model is taken to illustrate the rollover mitigation by a
hybrid controller (Fig. 13). The steering input is depicted in
terms of steering wheel angle whereas the vehicle reaction is
expressed in terms of lateral acceleration and yaw rate. The
stepped variable at the top of the chart indicates the discrete
states of the hybrid controller. The curves at the bottom show
the target brake torque values for the left and right wheel.
During severe steering back a brake torque pre-control at the
curve inside wheel (right wheel) is used to apply the brake
pads to the brake discs to reduce pressure build up time (see
Fig. 13, dotted lines).
Such a hybrid controller can easily be extended beyond the
previously mentioned discrete states to cover other driving
situations like e.g. slalom driving.
designed. It uses the vehicle’s mass and the estimated CoG
position to adjust the threshold for brake interventions. This
ensures timely interventions with the correct intensity and
minimized comfort impairment.
The results of several independent studies show a consistent
picture of the ESP with remarkable safety benefits and proof
the positive impact. Further potential is available with
functional extensions especially for SUV and light trucks
concerning rollover mitigation and 4WD adaptations. The
ESP with Rollover Mitigation functions helps the driver to
stay on the road and to avoid tripping obstacles by a specific
yaw control. It also supports the driver with an optimized
lateral acceleration control to manage rollover critical on-road
situations. In cooperation with four wheel drive train
concepts, ESP delivers at the same time the expected safety
benefits and excellent off-road and handling functionality.
Figure 13: Example of a severe steering maneuver: NHTSA Fishhook with a
sporty SUV model with ESP 8; entrance velocity vF=72 km/h.
Since the major parameter to recognize rollover-critical
driving situations is the measured lateral acceleration ay
relative to the center of gravity. This value plays an important
role in the execution and release of roll-stabilizing
interventions and in the determination of the suitable strength.
However, only the measured lateral acceleration is not
sufficient to clearly detect rollover-critical situations in due
time and to prevent incorrect interventions at high lateral
accelerations in otherwise uncritical driving situations. Beside
the lateral acceleration ay, a lead in the form of the lateral
acceleration gradient, the steering angle velocity and the
steering angle itself are used to calculate a so-called effective
lateral acceleration. In the Fishhook example above, the
effective lateral acceleration is plotted indicating the rollover
propensity during this severe steering maneuver.
If the fixed release threshold dependent on the beforehand
mentioned effective lateral acceleration is used to execute
roll-stabilizing interventions, an improved behavior can be
realized for the empty as well as fully laden vehicle with a
minimized comfort impairment due to early braking
interventions. For vehicles with a high variance of the center
of gravity height, an adaptive rollover mitigation strategy is
Rabe, M.; VW-Research, Germany, 5. Symposium
Automatisierungs- und Assistenzsysteme für Transportmittel, Braunschweig, Germany, (17-Feb-2004).
2. Papelis, Y.E.; Brown T.; Watson G.; Holtz, D.;Pan, W.;
University of Iowa, National Advanced Driving
Simulator, Document ID: N04-003-PR.
3. Aga, M.; Okada, A.; Toyota, Japan, Paper No. 541, JSAE
Automotive Engineering Exposition, Yokohama, May
4. Foerster: “Der Fahrzeugführer als Bindeglied zwischen
Reifen, Fahrwerk und Fahrbahn”, VDI-Berichte, Nr. 916
5. Van Zanten, A. et al.: “Control Aspects of the BoschVDC”. International Symposium on Advanced Vehicle
Control AVEC ‘ 96, 1996.
6. Van Zanten, A. T.: “Bosch ESP systems: 5 years of
experience”. SAE 2000-01-1633, 2000.
7. Chen, B.-C.; Peng, H.: “Differential braking based
rollover prevention for Sport Utility Vehicles with
Dynamics, Vol. 36, No. 4-5, pp. 359-389, 2001.
8. National Highway Traffic Safety Administration
(NHTSA): Final Policy Statement on NCAP
Rollover Resistance Rating, Consumer Information,
9. Ackermann, J.; Odenthal, D.: “Damping of vehicle roll
dynamics by gain scheduled active steering”.
Proc. European Control Conference, Karlsruhe,
Germany, 1999.
10. Sampson, D.J.M.: “Active Roll Control of Articulated
Heavy Vehicles”. Ph.D. thesis, Cambridge University
Engineering Department, UK, 2000.
11. BMW EDC, see
12. Brown, T. A. et al.: “Rollover Stability Control for an
Automotive Vehicle”. US patent No. 6,263,261 B1.
13. Branicky, M.S.: “Studies in Hybrid Systems: Modeling,
Analysis, and Control”. PhD Thesis, Laboratory for
Information and Decision Systems, MIT, Cambridge,
USA, 1995.
Dr. E. K. Liebemann, Robert Bosch Corporation, Chassis
Systems, email: [email protected]
Electronic Stability Program
Sport Utility Vehicle
Light Truck
Four Wheel Drive
Anti-Lock Control
Center Coupling Control
Center of Gravity
Static Stability Factor
National Highway Traffic Safety Administration
Rollover Mitigation Function
Hybrid Rollover Mitigation Controller
Electronic Stability Control
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF