ANTILOCK-BRAKING SYSTEM USING FUZZY LOGIC

ABSTRACT
This paper deals with study and tests on
an experimental car with antilock-
braking !t"# $ABS% an& '"(icl"
)""& "ti#ation *ing +*,,! logic.
Vehicle dynamics and braking systems
are complex and behave strongly non-
linear which causes difficulties in
developing a classical controller for
ABS. u!!y logic" however facilitates
such system designs and improves
tuning abilities. The underlying control
philosophy takes into consideration
wheel acceleration as well as wheel slip
in order to recogni!e blocking
tendencies. The knowledge of the actual
vehicle velocity is necessary to calculate
wheel slips. This is done by means of a
+*,,! "ti#ator- which weighs the
inputs of a longitudinal acceleration
sensor and four wheel speed sensors. #f
lockup tendency is detected" magnetic
valves are switched to reduce brake
pressure. $erformance evaluation is
based both on computer simulations and
an experimental car. To guarantee
realtime ability %one control cycle takes
seven milliseconds& and to relieve the
electronic control unit %'()&" all fu!!y
calculations are made by the +*,,!
co)roc"or SAE ./C00A.
*easurements in the experimental car
prove the functionality of this
automotive fu!!y hardware system.
INTRO1UCTION
u!!y (ontrol" a relatively new"
intelligent" knowledge based control
techni+ue performs exceptionally well in
nonlinear" complex and even in not
mathematically describable systems.
Thus the use of fu!!y logic for an
antilock-braking system %ABS& seems to
be promising.
ANTILOCK-BRAKING
SYSTEMS
The aim of an ABS is to minimi!e brake
distance while steerability is retained
even under hard braking. To understand
the underlying physical effect which
leads to wheel-blocking during braking"
consider igure ,a- (oefficient of
friction is shown as a function of wheel
slip" relating to the terms given in igure
,b.
Figure 1: a) Friction characteristics
b)Wheelmodel
FZ: Wheel load
R: Wheel radius
w: Angular wheel frequency
v: Velocity of wheel center
F: ongitudinal force
(alculat
ing the
wheel
slip by
"
the
longitudi
nal
wheel
force
results in
.
At the beginning of an uncontrolled full
braking" the operating point starts at s .
/" then rises steeply and reaches a peak
at s . s max. After that" the wheel locks
within a few milliseconds because of the
declining friction coefficient
characteristic which acts as a positive
feedback. At this moment the wheel
force remains constant at the low level
of sliding friction. Steering is not
possible any more.
Therefore a fast and accurate control
system is re+uired to keep wheel slips
within the shaded area shown in igure
,a.
2E3ICLE S4EE1
A crucial point in the development of
wheel slip control systems is the
determination of the vehicle speed.
There are several methods possible- until
now the velocity is measured with
inductive sensors for the wheel
rotational speed. 'specially in the case
of brake slips the measured speed does
not correspond with reality. To obtain
very accurate results" optical or
microwave sensors take advantage of a
correlation method. 0owever" these
sensors are very expensive and will not
be used for ABS.
SENSORS AN1 ACTUATORS
The experimental car was fitted with
sensors and actuators shown in igure 1.
'ach wheel is connected to a metallic
gearwheel" which induces a current
within an attached sensor. The fre+uency
of the rectangular shaped current is
proportional to the angular fre+uency w i"
2 and can be evaluated by a
microcontroller. #n addition to common
ABS fitted cars" a capacitive acceleration
sensor for measuring the longitudinal
acceleration ax is implemented. urther
more igure 1 depicts the hydraulic unit
including main brake cylinder" hydraulic
lines and wheel brake cylinders. By
means of two magnetic two-way valves
each wheel" braking pressure pi" 2 is
modulated.
Three discrete conditions are possible-
decrease pressure" hold pressure firm
and increase pressure %up to main brake
pressure level only&. 'ach valve is
hydraulically connected to the main
brake cylinder" to the wheel brake
cylinders and to the recirculation.
Figure !: "ensors and actuators of the e#$erimental
car
(3- (enter of gravitiy
a
x
- 4ongitudinal acceleration
w
i"2
- Angular wheel fre+uency
0)- 0ydraulic )nit
p
i"2
- 5heel brake pressure
i- l.left" r.right
2- f.front" r.rear
ESTIMATION OF 2E3ICLE
S4EE1 USING FUZZY LOGIC
As described in the first chapter" the
knowledge of the actual vehicle speed
over ground is vital in order to calculate
wheel slips correctly. 6ai7 and 8iencke
9,: presented an estimation system based
on 8alman-ilter which performs well"
but is not suitable because of very high
performance re+uirements. #n this
approach the speed estimation uses
multisensor data fusion that means
several sensors measure vehicle speed
independently and the estimator decides
which sensor is most reliable. igure ;
represents the schematic structure of the
fu!!y estimator. The signals of the four
wheel speed sensors w i"2 are used as well
as the signal of the acceleration
sensor ax.
igure ;- 'stimation of car velocity
#n a data pre-processing block the
measured signals are filtered by a
lowpass and the inputs for the fu!!y
estimator are calculated- four wheels slip
" and an acceleration value 6 va. The
applied formulas are-
a
n
d
"
whereby a<ffset is a correction value
consisting of an offset and a road slope
part. #t is derived by comparing the
measured acceleration with the
derivative of the vehicle speed v u!"
which is calculated with the fu!!y logic
system. After this subtraction" the signal
is lowpass filtered to obtain the constant
component a<ffset
.
v u!%k-,& is the
estimated velocity of the previous cycle.
A time-delay of T is expressed by the
term ,=!.
The fu!!y estimator itself is divided into
two parts. The first %4ogic ,& determines
which wheel sensor is most reliable" and
the second %4ogic 1& decides about the
reliability of the integral of the
acceleration sensor" shown in igure >.
This cascade structure is chosen to
reduce the number of rules.
igure >-structure of the fu!!y estimator
Starting at block ?4ogic ,@ and ?4ogic
1@ the crisp inputs are fu!!ificated.
igure A shows the input-membership-
functions %#*& with four linguistic
values %%egative& Zero& 'ositive and
Very('ositive&.
igure A- input membership functions
The rule base consists of ;A rules
altogether. To classify the present
driving condition vehicle
acceleration is taken into
consideration. This should be
explained for three situations-
6 va $ositive- Braking situation"
all wheels are weighted low
because of wheel slips appearing.
6 va Bero- #f wheel speeds tend
to constant driving the
acceleration signal is low
weighted in order to ad2ust the
sensor.
6 va Cegative- The experimental
car was rearwheel driven
therefore rear wheels are less
weighted than front wheels.
igure D- output membership
functions
igure D depicts the output-membership-
functions %<*&. 0ere" three linguistic
values are sufficient. The output of the
estimation is derived as a weighted sum
of the wheel measurement plus the
integrated and corrected acceleration-
.
T3E FUZZY-ABS
ALGORIT3M
The u!!y-
(ontroller uses two
input values- the
wheel slip SB-
and
the
wheel
accele
ration
-
with wheel speed v5heel and vehicle
speed vu!" which is given by the u!!y-
'stimator.The input variables are
transformed into fu!!y variables slip and
dvwheel=dt by the fu!!ification process.
Both variables use seven linguistic
values" the slip variable is described by
the terms
sli$ . E)ero& very small& too
small& smaller than o$timum&
o$timum& too large& very largeF"
and the acceleration dvwheel*dt by
dvwheel*dt . Enegative large&
negative medium& negative small&
negative few& )ero& $ositive
small& $ositive largeF.
As a result of two fu!!y variables" each
of them having seven labels" >G different
conditions are possible. The rule base is
complete that means" all >G rules are
formulated and all >G conditions are
allowed. These rules create a nonlinear
characteristic surface as shown in igure
H.
igure H- fu!!y characteristic
surface
)sing this characteristic surface" the two
fu!!y input values sli$ and dvwheel*dt
can be mapped to the fu!!y output value
$ressure. The labels for this value are-
$ressure . E$ositive fast& $ositive
slow& )ero& negative slow&
negative fastF
The structure of the fu!!y ABS
controller is shown in igure I.
igure I- structure of fu!!y ABS
controller
The optimal breaking pressure results
from the defu!!ification of the linguistic
variable $ressure. inally a three-step
controller determines the position of the
magnetic valves" whether the pressure
should be increased" hold firm or
decreased.
igure G summari!es the total amount of
fu!!y calculations. Cumbers within a
rectangle indicate the +uantity of fu!!y
rules.
igure G- fu!!y calculations
#t should be noted that linguistic
variables and rule tables can be designed
with numerical optimi!ation methods"
for example described in 91:. #n this
work they were created using expert
knowledge and analysis of measured
data during ABS braking action.
SIMULATION OF A FULL
BRAKING
After implementation of the whole
system in S#*)4#C8" a full braking on
high-m -road was carried out" with and
without the fu!!y ABS.
5ithout fu!!y ABS the braking pressure
reaches a very high level and the wheels
block within short. This results in an
unstable behavior" the vehicle cannot be
steered any more and the stopping
distance increases.
5ith fu!!y ABS controller activated"
steerability is not only retained during
the whole braking maneuver" but the
slowing down length was considerably
shortened as well.
The following graphs show the steady
decline of the vehicle speed" the
fluctuating decline of the wheel speed of
the left front wheel as an example and
the fluctuating level of the wheel slip.
The applied braking pressure is depicted
in the last diagram. The other wheels
behave approximately similar.
igure ,/a- simulations of a braking
igure ,/b- simulations of a full
braking
igure ,/c- simulations of full
braking
IM4LEMENTATION OF T3E
FUZZY ABS CONTROLLER
The fu!!y ABS controller uses the
microprocessor SAB I/(,DD together
with the fu!!y coprocessor SA' I,(GGA
9;:. 6ue to the implementation of u!!y
algorithms into the hardware of the
coprocessor" the calculation speed of the
host processor increased significantly.
5hile the control cycle time was set to a
standard value of H msec" the
computation time was only /.A msecJ
This offers facilities for implementation
of extended vehicle dynamics control .
The flexibility of the coprocessor is
considerable" up to D> rule bases are
possible" each of them having up to 1AD
inputs and rules. urthermore an
interface to most commonly used
microprocessors is available. Arbitrary
shapes of membership functions"
different defu!!ification modes
including K(enter of 3ravity@" an
enormous rule engine with up to ,/
million rule calculations per second
makes this device a very interesting
product in the field of real time fu!!y
control.
TEST RESULTS
After the whole system was carefully
simulated" tests on an experimental car" a
B*5 ;1Ii" were carried out. igure ,,
shows a full braking with ABS on dry
asphalt.
igure ,,- Lesults of test brake
The first diagram displays the decreasing
estimated speed of the vehicle vu! and
the fluctuating decrease of the speed of
the left front wheel vl"f. 5heel
acceleration and wheel slip are shown in
the second and third graph. The slip
value is limited successfully by means of
the output of the ABS controller" which
is the driver current of the magnetic
valve" presented in the next diagram.
inally the system performance is
proved by the last graph. The
longitudinal acceleration ax is near the
physical limit.
CONCLUSION
The basis of the controlling algorithm
consists of a nonlinear characteristic
surface" which was created by fu!!y
logic. The convincing advantage of
fu!!y logic is the ability to modify and
tune certain parts of this characteristic
surface easily and carefully. Must the
linguistic rules or variables need to be
varied. This simplifies the development
and shortens the development time
considerable. #mplementation of the
fu!!y ABS leads to excellent results of
braking behavior of the test vehicle. The
deceleration level and steer ability is
comparable to commercially available
systems.
BIBLIOGRA43Y
9,: 6ai7" A. and 8iencke" ).-
+stimation of Vehicle "$eed , Fu))y,
+stimation in -om$arison with
.alman,Filtering" >
th
#''' ((A"
Cew Nork" ,GGA.
91: <stertag" *.- "tru/turierte
0$timierung technischer 'ro)esse
am 1eis$iel der .FZ
-rasher/ennung" #nstitute for
#ndustrial #nformation Systems"
)niversity of 8arlsruhe" $h. 6.
dissertation" ,GGD.
9;: 8lein" L.- Realisierung einer Fu))y,
A1",Regelung mit dem
2i/rocontroller "A1 34-155 und
dem Fu))y,-o$ro)essor "A+
31-66A" $ro2ect work at the #nstitute
for #ndustrial #nformation Systems"
)niversity of 8arlsruhe" ,GGA.
9>: 6ai7" A.- 1eobachtung
fahrdynamischer Zust7nde und
Verbesserung einer A1", und
Fahrdynami/regelung" #nstitute for
#ndustrial #nformation Systems"
)niversity of 8arlsruhe" $h. 6.
dissertation" ,GGD.