See discussions, stats, and author profiles for this publication at: [Link]

net/publication/325833794

From the standard PID to the CRONE first generation controller: Application
to an anti-roll system for Electric Vehicles.

Article  in  IFAC-PapersOnLine · January 2018

DOI: 10.1016/[Link].2018.06.209

CITATIONS READS

3 91

4 authors, including:

Hussein Termous Francis Clovis

University of Bordeaux Lebanese University
5 PUBLICATIONS   28 CITATIONS    99 PUBLICATIONS   687 CITATIONS   

SEE PROFILE SEE PROFILE

Hassan Shraim
Lebanese University
58 PUBLICATIONS   463 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Thesis View project

Towards Global Chassis Control View project

All content following this page was uploaded by Hussein Termous on 05 February 2020.

The user has requested enhancement of the downloaded file.

Proceedings of the 3rd IFAC Conference on Control
Advances in Proportional-Integral-Derivative
Proceedings
Advances of the 3rd IFAC Conference on Control
in Proportional-Integral-Derivative
Ghent, Belgium, May 9-11, 2018
Proceedings of the 3rd IFAC Conference on
Advances
Ghent, in Proportional-Integral-Derivative
Belgium,
Proceedings of May
the 9-11,
3rd IFAC2018 on Control
Available
Conference online at [Link]
Advances in
Proceedings Proportional-Integral-Derivative
of the 3rd IFAC Conference on Control
Ghent, Belgium,
Advances May 9-11, 2018
in Proportional-Integral-Derivative Control
Ghent, Belgium,
Advances May 9-11, 2018
in Proportional-Integral-Derivative Control
Ghent, Belgium, May 9-11, 2018
Ghent, Belgium, May 9-11, 2018 ScienceDirect
From
From the
the standard
standardIFAC PID to
to the
PapersOnLine
PID 51-4 CRONE
the (2018) 733–738 first generation controller:
CRONE first generation controller:
From
From the
the standard
Application
standard toPID
an
PID to
to the
anti-roll
the CRONE
systemfirst
system
CRONE for generation
for Electric
first generation controller:
Vehicles.
controller:
From Application
the standard toPID
an anti-roll
to the CRONE firstElectric
generation Vehicles.
controller:
From Application
the standard
Application to
to an
PID
an anti-roll
to the
anti-roll system
CRONE
system for
first
for Electric
generation
Electric Vehicles.
controller:
Vehicles.
Application
Hussein to an
Termous*/**,
Hussein Termous*/**, anti-roll
Xavier Moreau**,system
Clovis for Electric
Francis *, Vehicles.
Hassan Shraim*
Application to anXavier Moreau**,
anti-roll Clovisfor
system Francis *, Hassan
Electric Shraim*
Vehicles.
Hussein Termous*/**, Xavier Moreau**, Clovis Francis *, Hassan Shraim*
* Lebanese University, Hussein Termous*/**,
Faculty of Engineering, Xavier Moreau**,
Scientific ClovisCenter
Research Francis *, Hassan Shraim*
in Engineering (CRSI), Campus Rafic Hariri,
* Lebanese University, Hussein Termous*/**,
Faculty of Engineering, Xavier Moreau**,
Scientific ClovisCenter
Research Francis *, Hassan Shraim*
in Engineering (CRSI), Campus Rafic Hariri,
** Lebanese Beirut, Lebanon,
Hussein
University, Email:
Faculty of husseintermous@[Link];
Termous*/**,
Engineering,Xavier Moreau**, cfrancis@[Link];
Clovis Francis *, [Link]@[Link]
Hassan Shraim*
Lebanese Beirut, Lebanon,
University, Email:
Faculty Engineering, Scientific
Scientific Research
of husseintermous@[Link]; Center
Center in in Engineering
cfrancis@[Link];
Research (CRSI),
(CRSI), Campus
[Link]@[Link]
Engineering Campus Rafic Rafic Hariri,
Hariri,
* Lebanese University,
Beirut, Lebanon, Faculty
Email: of husseintermous@[Link];
Engineering, Scientific Research Center in Engineering
cfrancis@[Link]; (CRSI), Campus Rafic Hariri,
[Link]@[Link]
* Lebanese Beirut, Lebanon,
University, Email:
Faculty husseintermous@[Link];
of husseintermous@[Link];
Engineering, Scientific Research cfrancis@[Link];
Center in Engineering [Link]@[Link]
(CRSI), Campus Rafic Hariri,
Beirut, Lebanon, Email: cfrancis@[Link]; [Link]@[Link]
Beirut, Lebanon, ** University
Email: of Bordeaux, Lab IMS CNRScfrancis@[Link];
husseintermous@[Link]; 5218, Talence, Bordeaux, France
[Link]@[Link]
** University of Bordeaux, Lab IMS CNRS 5218, Talence, Bordeaux, France
** Email: [Link]@[Link]
** University
University of of Bordeaux, Lab
Lab IMS
IMS CNRS
CNRS 5218,
Email: [Link]@[Link]
Bordeaux, 5218, Talence,
Talence, Bordeaux,
Bordeaux, France France
** University of Bordeaux, Email: Lab IMS CNRS 5218, Talence, Bordeaux, France
[Link]@[Link]
** University of Bordeaux, Email: [Link]@[Link]
Lab IMS CNRS 5218, Talence, Bordeaux, France
Abstract: In this paper, the standardEmail: [Link]@[Link]
PID controller and the fractional first generation CRONE controller
Abstract: In this paper, the standardEmail: [Link]@[Link]
PID controller and the fractional first generation CRONE controller
are applied
Abstract: on the anti-roll moment system to improve ride comfort forfirst
passengers in the framecontroller
of global
are appliedIn
Abstract: In this
on
thisthepaper,
anti-roll
paper, the
the standard
moment
standard PID
system
PID controller
to improve
controller and the
ride fractional
the comfort
andstudy fractional forfirst generation
passengers
generation in theCRONE
CRONEframecontroller
of global
chassis
Abstract:
are appliedcontrol
Inonthis of
the electric
paper,
anti-roll vehicles.
the moment
standard A
PID
system comparative
controller
to improveandstudy is done
the comfort
ride fractional forshowing
first the
generation
passengers performance
in CRONE
the frame and
controller
of the
global
chassis
are applied
Abstract: control
In on
this of
the electric
anti-roll
paper, the vehicles.
moment
standard A
system
PID comparative
to improve
controller and ride
the is done
comfort
fractional forshowing
passengers
first the
generation performance
in the
CRONEframe ofand the
global
controller
robustness
are applied
chassis of
control the
on the
the
of two controllers,
anti-roll
electric moment
vehicles. in frequency
system
A and
to improve
comparative time domain.
ride domain.
study comfort
is done It
for is shown
passengers
showing that
the in the first
the frame
performance generation
ofand
global
the
robustness
chassis
are applied of
control
on the of two controllers,
electric
anti-roll vehicles.
moment in Afrequency
system comparative and
to improve time
study is
rideorder done
comfort It
for is shown
showing
passengers that
the the first
performance
in theperturbedgeneration
and
frame ofand the
global
CRONE
chassis
robustness control-system
control
of theoftwo
electricis able
controllers,to provide
vehicles. in A robust
comparative
frequency fractional
and study
time [Link]
is done It showing
is for uncertain
shown the
that performance
the first plants.
generation the
CRONE
robustness
chassis control-system
of
control the two
oftwo
electricis able
controllers,to provide
vehicles. in A robust
frequency
comparative fractional
and time order
domain.
studydomain. controller
It is
is doneItshowing for
shown uncertain
that
the the perturbed
first
performance plants.
generation
and the
robustness of (International
theCRONE
control-system controllers,
is able to in frequency andControl)
time by is shown that Allthe firstreserved.
generation
CRONE
© 2018, IFAC
CRONE
Keywords:
robustness PID,
control-system
of the two is able
controllers,to provide
Federation
, fractional order
provide
in
robust
robust fractional
of Automatic
controller,
frequency fractional
and time
order
robustness,
order
domain.
controller
Hosting
anti-roll,
controller
It is
for
for uncertain
Elsevier
comfort,
shown Ltd. global
uncertain
that the
perturbed
rights
chassis
perturbed
first
plants.
control.
plants.
generation
Keywords: PID, CRONE , fractional order controller, robustness, anti-roll,
CRONE control-system is able to provide robust fractional order controller for uncertain perturbed plants. comfort, global chassis control.
CRONE control-system is able to provide robust fractional order controller for uncertain perturbed plants.
 Keywords: PID,
Keywords: CRONE,, fractional
PID, CRONE fractional order
order controller,
controller, robustness,
robustness, anti-roll,
anti-roll, comfort,
comfort, globalglobal chassis
chassis control.
control.
 Keywords: PID, CRONE , fractional order controller, robustness, anti-roll, comfort, global chassis control.
 Keywords: 1. INTRODUCTON
PID, CRONE, fractional order controller, robustness, also ananti-roll,
issue for railroad
comfort, globalvehicles
chassiswith high center of
control.
 1. INTRODUCTON also an issue for railroad vehicles with high center of
1. INTRODUCTON gravity.
also an issue for railroad vehicles with high

Since
 its introduction, 1. INTRODUCTON
PID controllers have been widely
gravity.
also an issue for railroad vehicles with high center
center of of
Since its introduction, 1. INTRODUCTON
PID controllers have been widely also an
gravity. issue for railroad vehicles with high center of
used in industrial 1. INTRODUCTON
control applications (Bennett, 2001). For gravity.
also
To an issue the
improve for roll
railroad
dynamics,vehiclesseveral
with highvehiclescenter of
were
Since gravity.
Sinceinits
used
years, its
it
introduction,
industrial
introduction,
was the
controlPID
most PID
popular
controllers
applications
controllers
controller,
have
(Bennett,
have been
2001).
been
mainly
widely
widely
because
For To improve the roll dynamics, several vehicles were
gravity.
equipped with active suspension components controlled by
Since
used inits introduction,
itindustrial control PID controllers
applications have
(Bennett, been
2001). widelyFor To
years,
useditsin
Since
of
was the most
industrial
its introduction,control popular
PID controller,
applications
controllers mainly
(Bennett,
have because
2001).
been widelyFor To improve
equipped
improvewiththe active
the roll dynamics,
suspension
rollcontrollers
dynamics, several
components vehicles
severalet al., controlled
vehicles were
2016),were
by
used
years,
of itsinsimple
it was
simple
parameters
industrial
the control
most
parameters
and
popular
and
simple
applications structure
(Bennett,
controller,
simple structuremainlythat
2001).
that
need
because
need
to
For
to
different
To
equipped
different
methods
improvewith
methods the
activeof
rollcontrollers
of dynamics,
suspension (Chokor
severalet al.,
components
(Chokor vehicles
controlled
2016),
like
were
by
like
years,
used init was the most to popular controller, mainly because equipped withthe activerollsuspension components controlled by
be
of
be
used,
years,
its
used, itindustrial
in addition
was
simple
in
control
the most
parameters
addition to
theapplications
and
the
ease
popular simple
ease with
(Bennett,
withstructure
which
controller,
which
2001).
engineers
mainlythat
engineersbecause
need
For
can
to
can
To improve
the Hydractive,
equipped
different
the with active
methods
Hydractive,
CRONE
of
CRONE
dynamics,
(Moreau
suspension
controllers
(Moreau
several
et al., et
components
(Chokor
et al.,
vehicles
n.d.)
n.d.)
(Moreau
controlled
al., 2016),
(Moreau
wereet
by
like
et
of its simple
years,
implement it was itparameters
the most
using and simple
popular
current computerstructure
controller, mainlythat because
technology need(De to different
equipped
al., 2009),methods
with
and activeof controllers
BOSE suspension
(Brown, (Chokor
components
2005). et
Oneal., 2016), like
controlled
example by
of
of
be its
used,
implementsimple
in itparameters
addition
using to and
the
current simple
ease with
computerstructure
which that
engineers
technology need to
can
(De different
the
al., methods
Hydractive,
2009), and of
CRONE
BOSE controllers
(Moreau
(Brown, (Chokor
et
2005).al., et al.,
n.d.)
One 2016),
(Moreau
example like
et
of
be
of used,
its
Oliveira in
simple addition
parameters
andaddition
Karimi, to the
and
2012). ease
simplewith which
has structure
Itcomputer high engineers
that needcan
performance can
to
in the Hydractive,
different
active methodsCRONE
suspension ofsystem (Moreau
controllersis the et
oneal.,
(Chokor n.d.)
et al., (Moreau
introduced 2016), et
in like
the
be used,
implement
Oliveira in it using
andaddition to the
current ease with which engineers
technology (De the
al., Hydractive,
2009), and CRONE
BOSE (Moreau
(Brown, et
2005).al., n.d.)
One (Moreau
example et
of
implement
be
manyused, in
control it Karimi,
problems,
2012).
using tocurrent
the ease
but
Itcomputer
the
has high
with which performance
technology
engineers
performance/robustness (Dein
can active
the
Active
suspension
al., Hydractive,
2009),
Wheel and BOSE
CRONE
(Laurent
system
et
is the2005).
(Brown,
(Moreau
al., 2000)
oneal.,introduced
et
(Fig.1).One
n.d.) example
(Moreauin the
of
et
implement
Oliveira
many control and it using
Karimi,
problems, current
2012). It
but Itthecomputer
has high technology
performance
performance/robustness (Dein al.,
active
Active2009), and
suspension
Wheel BOSE
(Laurent system (Brown,
is
et(Brown, the
al.,is2000) 2005).
one One
introduced
(Fig.1). example in of
the
Oliveira
implement
analysis and
of it Karimi,
using
closed-loop 2012).
current systems has
computer high
with performance
technology
PI and (De
PIDin active
al., suspension
2009), and BOSE system the2005).
one introduced
One example in the
of
Oliveira
many and
analysiscontrolof Karimi,
problems, 2012).but Itthehas high
with performance
performance/robustness in active
Active suspension
Wheel (Laurent systemet is the one introduced in the
many
Oliveira control
and closed-loop
problems,
Karimi, thebut
2012). systems
Itthe PI and PID
performance/robustness Active Wheel (Laurent et al.,
al., 2000)
2000) (Fig.1).
(Fig.1).
controllers
many
analysis control
controllers of revealed
problems,
closed-loop
revealed the but thehas
existing
systems
existing
high performance
trade-off
performance/robustness
with PI
trade-off
between
and
between PID
in active suspension
Active Wheel (Laurent systemet al.,is2000)
the one introduced in the
(Fig.1).
analysiscontrol
many
performance of and closed-loop
problems,
robustness systems
but the
(Garpinger with
et PI2014).
al., and PID
performance/robustness Active Wheel (Laurent et al., 2000) (Fig.1).
analysis
controllers of and closed-loop systems with
et al.,PI and PID
controllersof revealed
performance
analysis revealed
closed-loop
the
robustness existing
(Garpinger
the systems
existing with trade-off
PI2014).
trade-off between
between
and PID
controllers revealed
performance and the (Garpinger
robustness existing trade-off
et al., 2014).between
performance
controllers
In order to achieve and
revealedrobustness
the (Garpinger
existing
more challenging et al.,
trade-off
control 2014).between
performance
In order to achieve and robustness (Garpinger
more challenging et al.,requirements,
control 2014).
performance
the fractional andorder
robustness (Garpinger
controllers have al.,requirements,
etreceived2014). a great
In
In order
theorder to
fractional achieve
torecently.
achieveorder more
more challenging
controllers
challenging havecontrol requirements,
received
control a great
requirements,
attention
In
theorder torecently.
fractional achieveorder This
more kind
challenging
controllers of controllers
havecontrol are aa more
requirements,
received great
attention
the
In fractional
order to achieveorder This
more kind
controllers
challenging of controllers
have received
control are
requirements, more
great
flexible
the
attention than
fractional the
order
recently. integer-order
controllers
This kind ones.
of have One approach
received
controllers are a for
great
more
flexible
attention
the than
fractional the
recently.
orderinteger-order
This kind
controllers ones.
of One
controllers
have approach
received are a morefor
great
fractional
attention
flexible order
recently.
than the controllers
This kindnamed
integer-order namedof
ones. CRONE,
controllers
One a
are
approach French
morefor
fractional
flexible
attention order
than the
recently. controllers
integer-order
This kind ones.
of CRONE,
One
controllers a
approach French
areEntier”
morefor Fig. 1. Active wheel from Michelin.
acronym
flexible
fractional of
than “Commande
order the controllers Robuste
integer-order namedones. d’Ordre Non
One approach
CRONE, a French for
acronym
fractional
flexible of
than “Commande
order the controllers Robuste
integer-order namedones. d’Ordre
CRONE,
One Non a
approach Entier”
French for Fig. 1. Active wheel from Michelin.
which means non-integer order robust controller. The Fig. 1.
fractional
acronym
which
acronym of
meansorder
oforder
controllers
“Commande
non-integer
“Commande
fractionalmethodology
CRONE controllers
named
Robuste
order
Robuste
namedd’Ordre
is a frequency-domain
CRONE,
d’Ordre
robust
CRONE,
Non
Non
a Entier”
controller. French
a Entier”
approach
The
French for This kind ofFig. active1. Active
Active wheel from
wheelswheel is mostly
Michelin.
fromused
Michelin.
in light electrical
which
acronym
CRONE means non-integer
of “Commande
methodology is a order
Robuste robust
frequency-domaind’Ordre controller.
Non
approach The
Entier” for This kind ofFig.
active1. Active
wheels wheel
is fromused
mostly Michelin.
in light electrical
which
the design
acronym means
of non-integer
of output
“Commande feedback order
robust
Robuste robust controller.
controllers.
d’Ordre Non It is usedThe
Entier” cars, where Fig.
two 1. Active
in-wheel wheel
30-kW from Michelin.
motors motorize it. The
which
CRONE
the design meansof non-integer
methodology
output is a
feedback order
robust robust
frequency-domain controller.
controllers. approach
It is The
for
used This
cars,
This kind
where
kind of
of active
two
active wheels
in-wheel
wheels is
30-kW
is mostly
mostly used
motors
used in light
lightanelectrical
motorize
in it. The
electrical
CRONE
which
for methodology
means
continuous-time non-integer
oris discrete-time
a frequency-domain
order robustproblems, approach
controller. and for
The
for suspension system is composed of a spring and electric
This
cars, kind
where ofsystem
active
two wheels
in-wheel is mostly
30-kW used
motors in lightanelectrical
motorize it.
CRONE
the
for
the
CRONE
design
design
perturbed
methodology
of
continuous-time
of output
output
methodology
SISO
oris discrete-time
a frequency-domain
feedback
feedback robust
robust
is a frequency-domain
LTIfeedback
systems. For the
controllers.
problems,
controllers.
approach
SISOapproach
It
It
case,
andis
is
for
used
for
used
for
three
suspension
cars,
This
engine. where
kind two
of two
This active
system
is composed
in-wheel
wheels 30-kW
is mostly
offers
of
new
a spring
motors
used in and
motorize it. The
electric
The
light electrical
perspectives in the
the
for design of
continuous-time
perturbed SISO outputLTI or robust
discrete-time
systems. For controllers.
the problems,
SISO It
case,andis used
for
three cars,
engine. where
suspension Thissystem in-wheel
systemis composed 30-kW
offers of
new motors
a spring motorize
and
perspectives an it. The
electric
in the
for
the
CRONE continuous-time
design of output
Control-System or discrete-time
feedback robust
Design problems,
controllers.
(CSD) methods Itandis for
used
have suspension
cars, where
suspension system
two
control, is composed
in-wheel which30-kW of a
indeed spring
motors and
motorize
offers an electric
it. The
important
for continuous-time
perturbed
CRONE SISO LTI
Control-System or discrete-time
systems.
Design For the
(CSD)problems,
SISO case,
methods andthree
havefor suspension
engine.
suspension Thissystem system
control, is composed
offers
which of a spring
new
indeed and an
perspectives
offers electric
in
importantthe
perturbed SISO
for continuous-time
been developed, LTI systems.
or For
discrete-time
successively the
extending SISO
problems,
the case, three
andthree
application for engine.
suspension This
opportunities system system
insystem is offers
composed
global chassis new
of a
control. perspectives
spring and
Itoffers an
can provide in the
electric
safer
perturbed
CRONE
been SISO LTI
Control-System
developed, systems.
successively Design For the
(CSD)
extending SISO case,
methods
the applicationhave engine.
suspension This
opportunities control,
in global offers
which
chassis new
indeed
control. perspectives
It can provide in
importantthe
safer
CRONE
perturbed
field Control-System
SISO LTI
(Oustaloup, Design
systems.
1991)(Lanusse, For (CSD)
the SISO
2010). methods
thiscase,
Inthe have
paper,three we suspension
engine.
steering This
usingcontrol,
system which
offers
the chassis indeed
new
active control.
roll offers
perspectives
control important
system in the
by
CRONE
been
field Control-System
developed,
(Oustaloup, successively Design
1991)(Lanusse, (CSD)
extending
2010). methods
Inthe application
thisapplication have
paper,have we suspension
opportunities
steering usingcontrol,
in global
the which
active indeed
roll It offers
can
control important
provide
system safer
by
been
CRONE
studied developed,
the successively
Control-System
roll stability Design
enhancement extending
(CSD)
of an methods
electric vehicle opportunities
suspension
compensating in global
control,
roll chassis
which
dynamics control.
indeed
when It can
offers provide
manoeuvring. safer
important
The
been (Oustaloup,
field
studied developed,
the roll successively
1991)(Lanusse,
stability enhancement extending
2010).of anInthe
this application
electricpaper,
vehicle we opportunities
steering
compensating using in global
rollthe chassis
active
dynamics control.
roll
when It can
control provide
[Link] safer
by
The
field (Oustaloup,
beenmeans
by developed,
of two 1991)(Lanusse,
successively
controllers, 2010).
extending
the standard Inthe
this paper,
PIDapplication
and we
first steering
opportunities
principle using
of the the chassis
incontroller
global active roll control
control.
is to provide It can
an system
provide
anti-roll by
safer
moment
field
studied
by (Oustaloup,
means the roll
of 1991)(Lanusse,
stability
two enhancement
controllers, the2010).of an
standard Inelectric
this
PID paper,
andvehicle we
first steering
compensating
principle using
of the rollthedynamics
controller active
is to roll
when
provide control
manoeuvring.
an system
anti-roll moment by
The
studied
field the roll
(Oustaloup,
generation CRONE stability enhancement
1991)(Lanusse,
controller. 2010).of anIn electric
this vehicle
paper, we compensating
steering
opposes theusing
one rollthe
resultsdynamics
active
from when
roll
vehicle manoeuvring.
control
manoeuvring. system The
by
studied
by the roll stability enhancement of an electric vehicle compensating
principle of the roll dynamics
controller is to when
provide manoeuvring.
an anti-roll momentThe
by means
generation
means
studied the
of
CRONE
of
roll
two
two controllers,
controller.
controllers,
stability enhancement
the
the standard
standard
of an
PID
PID and
electric and first
first
vehicle opposes the
principle
compensating oneroll
of the results
controller from
dynamics is tovehicle
provide
when manoeuvring.
an anti-roll moment
manoeuvring. The
principle
opposes of the controller is tovehicle
provide an anti-roll moment
by
by
Roll
means of
generation
generation
means
stability
CRONE
CRONE
of
two controllers,
istwo
controller.
controller.
onecontrollers,
of the mainthe
the standard PID and first
standard
dynamic PIDthat
factors andneeds first opposes the
principle
Depending the one
one
of the
results
results
on controller
the
from
from
design is to vehicle
provide
manoeuvring.
manoeuvring.
methodologies an anti-roll moment
introduced in
generation CRONE controller. opposes
Depending the one
on the results from
design vehicle manoeuvring.
Roll
to be
stability
generation
considered
is one controller.
CRONE in
of the main dynamic factors that needs
the ground vehicle. It presents a high opposes
(Abi ZeidtheDaou
one results
et al., from methodologies
2013; vehicle
Morand et al.,
introduced in
manoeuvring.
2016), the PID
Roll
bestability is one in of
thethe main Depending on the design
2013; methodologies introduced in
to
Roll considered
stability
challenge foris one of
vehicle the main dynamic
ground
safety.
vehicle. Itfactors
dynamic
Geometric
presents
factors that
thataneeds
dimensions,
high
needs (Abi
and
Zeid Daou
Depending
the first on et al.,
the
generation design CRONE
Morand et al., 2016),
methodologies
controllers introducedthe PID
are designed in
Roll
to be stability
challengeconsidered
foris one
in of
the
vehicle the main
ground
safety. dynamic
vehicle.
Geometric factors
It presentsthat a
dimensions,needs
high Depending
(Abi
and Zeid
the first on
Daou the
et
generation design
al., 2013;
CRONE methodologies
Morand et al.,
controllers introduced
2016),
are the
designed in
PID
to be
Roll considered
stability
suspension is one in the
of
characteristics,theground
main vehicle.
anddynamic It
maneuvering presents
factors a
thataneeds
conditions high (Abi
to Zeid
Depending
control Daou
on et
the
thegeneration
roll al., 2013;
design
dynamics Morand et
methodologies
of al.,
thecontrollers
vehicle’s 2016),
introducedthe
chassis. PID
in
The
to be
challengeconsidered
suspension for in the
vehicle
characteristics, ground
safety.
and vehicle.
GeometricIt
maneuvering presents
dimensions,
conditions high (Abi
and
to Zeid
the
control Daou
first
the et
roll al., 2013;
dynamics CRONE Morand
of the et al.,
vehicle’s 2016),
are the
chassis. PID
designed
The
challenge
to be considered
influence for
the vehicle
in the roll
dynamic safety.
ground
behavior Geometric
vehicle. dimensions,
of aIt vehicle.
presents a high
This is and
(Abi the
Zeid
synthesis first
Daou
model generation
et al.,
is dynamics
implementedCRONE
2013; Morandcontrollers
et al.,
forcontrollers are
2016),
an anti-rollare designed
the
system PID
that
challenge
suspension for vehicle
characteristics, safety.
and Geometric dimensions, and the first generation CRONE designed
and maneuvering conditions
influence the dynamic to
to control
synthesis the
model roll
is dynamics
implemented of
of the vehicle’s
forcontrollers
an anti-rollarechassis.
system The
that
suspension
challenge
suspension for vehicleroll
characteristics,
characteristics,
behavior
safety.
and
of a vehicle.
maneuvering
Geometric
maneuvering
This is
conditions
dimensions,
conditions
and control
synthesis
to control
thegeneration
the first
model
the
roll
is
roll implemented
dynamics CRONE of
the
for
the
vehicle’s
an anti-roll
vehicle’s
chassis.
system
chassis.
The
designed
that
The
influence the dynamic roll behavior of a vehicle. This is synthesis
influence ©the
suspension
Copyright
dynamic
characteristics,
2018
roll behavior
IFAC roll behaviorand maneuveringof a vehicle. This is 733 to
conditions controlmodel
synthesis the roll
model
is dynamics
implemented
is implemented
for vehicle’s
of the an anti-roll system The
for an anti-rollchassis.
that
system that
influence
Copyright
the
©the
2018
dynamic
IFAC
of a vehicle. This is
influence©
2405-8963 2018, IFAC (International
dynamic roll behavior Federation of Automatic
of a vehicle. is 733Hosting
This Control) synthesis modelLtd.
by Elsevier is implemented
All rights reserved. for an anti-roll system that
Copyright
Peer review© 2018 responsibility
IFAC 733Control.
Copyright ©under
2018 IFAC of International Federation of Automatic
733
Copyright © 2018 IFAC
10.1016/[Link].2018.06.209 733
Copyright © 2018 IFAC 733
IFAC PID 2018
734
Ghent, Belgium, May 9-11, 2018 Hussein Termous et al. / IFAC PapersOnLine 51-4 (2018) 733–738

is embedded in the electric vehicle, and the main objective

of this control process is to reduce the body roll dynamics
for ride comfort and decreasing rollover propensity. The
effectiveness of both controllers will be studied in the
frame of their performance and robustness against
uncertain parameters. The results obtained in the
frequency and time domains are compared, revealing the
advantages and drawbacks of each controller.

After this introduction, section 2 gives the context. Section

3 introduces the synthesis model. Section 4 shows the
robustness and performance analysis. Finally, a conclusion
and the prospects.

2. CONTEXT

A high-level controller named “The supervisor” is Fig. 3. Vehicle situations.

responsible for computing the needed anti-roll moment. The controller variables associated with the centre of
This moment is then transferred to every suspension gravity of the vehicle are: The longitudinal
actuator embedded in the Active Wheels as reference velocity 𝑉𝑉𝑥𝑥 (𝑡𝑡) = 𝑉𝑉𝑥𝑥𝑟𝑟𝑟𝑟𝑟𝑟 (𝑡𝑡) , the yaw rate 𝜓𝜓̇(𝑡𝑡) = 𝜓𝜓̇𝑟𝑟𝑟𝑟𝑟𝑟 (𝑡𝑡),
forces to provide the needed moment around the roll axis.
Fig. 2 shows the complete Global Chassis Control and the main three dynamics of the chassis: the velocity of
diagram, where the supervisor takes as an input all the vertical displacement 𝑧𝑧̇𝐺𝐺 (𝑡𝑡), the pitch rate 𝜑𝜑̇ 𝐺𝐺 (𝑡𝑡), and the
measured signals needed, and provides as an output the roll rate 𝜃𝜃̇𝐺𝐺 (𝑡𝑡) such that:
reference for the different actuators (steering, traction,
braking, and local active suspensions). The objective of zG (t )  0, G (t )  0, and G (t )  0. (1)
the supervisor is to estimate first the life situation of the In this study we are focusing in one the above goals, which
vehicle based on a set of available measurements.
is 𝜃𝜃̇𝐺𝐺 (𝑡𝑡) = 0. This should achieve a good compensation of
roll dynamics around the roll axis that consequently will
Supervisor improve ride comfort. Assuming that the vehicle is
operating in the green region (Fig. 3), then the longitudinal
Global Reference velocity is considered constant (𝑉𝑉𝑥𝑥 (𝑡𝑡) = 𝑐𝑐𝑐𝑐𝑐𝑐, 𝑎𝑎𝑥𝑥 (𝑡𝑡) = 0),
and a stable maneuvering, with a fixed radius 𝑅𝑅, leads to a
Global Control
Function Level constant value of lateral acceleration estimated by:
Measurements

Measurements

UGC
Vx2 (t )
Control Allocation
a y (t )   cst  0.4 g. (2)
Steering W. Traction Brakes Suspension R
Local
Reference
2.1. Synthesis Model
Local Controllers
Component Level
(Actuator Technology) Consider the following equation that represents the roll
ULC
Yr dynamics (Newton’s second law):
dm t
Vehicle
Ym 1
I xx 0
G (t )  G (t )  c x ( )d  G (0),
Steering Wheel
Pedals
d
(3)
Driver

dm: measurable disturbances Yr: measurements of local outputs

(driver inputs) to control where 𝐼𝐼𝑥𝑥𝑥𝑥 is the inertia moment of the sprung mass with
d: non-measurable disturbances Ym: all measurements available respect to the roll axis, and 𝑐𝑐Σ𝑥𝑥 (𝜏𝜏) is the algebraic sum of
(slope, wind gust..) from the vehicle the exterior torques applied on the roll axis of the vehicle.
Fig. 2. Global Chassis Control diagram. It consists of two terms:

As shown in Fig.3, according to the estimated value or the c x (t )  cr 0 (t )  car (t ). (4)

measured values of the lateral acceleration 𝑎𝑎𝑦𝑦 and the
longitudinal acceleration 𝑎𝑎𝑥𝑥 of the vehicle, the supervisor 𝑐𝑐𝑟𝑟0 (𝑡𝑡) is the roll couple introduced to the system as a
can estimate the life situation of the vehicle. Then the perturbation and is equal to the following:
supervisor selects the best control strategy associated with cr 0 (t )  H M T a y (t ), (5)
the detected field of operation of the vehicle, and finally
generates the reference signals for all actuators. where 𝐻𝐻 is the height of the center of gravity of the
vehicle, and 𝑀𝑀𝑇𝑇 is the total mass of the vehicle. The term

734
 IFAC PID 2018
Ghent, Belgium, May 9-11, 2018 Hussein Termous et al. / IFAC PapersOnLine 51-4 (2018) 733–738 735

𝑐𝑐𝑎𝑎𝑎𝑎 (𝑡𝑡) is the anti-roll torque provided by the anti-roll

system. It results from the summation of two inputs:
 The feedback control input:

CFB ( s)   K ar ( s) G ( s). (6)

 The feedforward command input:

cFF (t )   H nom M Tnom   (t ). (7)

Knowing that 𝐾𝐾𝑎𝑎𝑎𝑎 (𝑠𝑠) is the controller, 𝛺𝛺𝐺𝐺 (𝑠𝑠) is the output
roll rate, g is the gravity, 𝐻𝐻𝑛𝑛𝑛𝑛𝑛𝑛 = 0.55 𝑚𝑚 is the nominal
height of the chassis, 𝑀𝑀𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 = 600 𝑘𝑘𝑘𝑘 is the nominal
total mass of the vehicle, 𝜆𝜆 = 6.11 𝑚𝑚𝑠𝑠 −2 𝑟𝑟𝑟𝑟𝑑𝑑 −1 , and 𝜃𝜃̃𝑣𝑣 (𝑡𝑡)
is the measured steering angle. The controlled plant, as
formulated in (3) with zero initial conditions, is 𝑃𝑃(𝑠𝑠) =
1/(𝐼𝐼𝑥𝑥𝑥𝑥 𝑠𝑠) as in Fig.4, where 𝑁𝑁(𝑠𝑠) represents the noise
measurements. For passive system, the anti-roll torque
𝑐𝑐𝑎𝑎𝑎𝑎 (𝑡𝑡) is given by:
t
Fig. 6. Bode diagrams of uncertain models.
car (t )  b G (t )  k  G ( )d  car (0), (8)
0

where 𝑏𝑏𝜃𝜃 is the coefficient of viscous friction at the centre

of gravity resulting from the four dampers of the vehicle in
[Nm s/rad], and 𝑘𝑘𝜃𝜃 is the stiffness coefficient resulting
from the four springs and 4 anti-roll bars in [Nm/rad].
Then the passive suspension can be presented as shown in
Fig. 5, by a transfer function 𝐾𝐾(𝑠𝑠) similar to one of the PI
regulator as the following equation with 𝐶𝐶𝑝𝑝 = 𝑏𝑏𝜃𝜃 and
𝜔𝜔1𝑝𝑝 = 𝑘𝑘𝜃𝜃 /𝑏𝑏𝜃𝜃 : Fig. 7. Introduced perturbations into the three plants.

k  1  s / 1 p 
K ( s)  CPI ( s)  b   Cp   . (9) 2.2. Signal Modelling
s  s /
 1p 
In order to achieve our objective, the reference value of
roll rate chosen to be zero (𝜔𝜔𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺 = 0). The perturbations
𝑐𝑐𝑟𝑟0 (𝑡𝑡) (rolling couple) introduced to the systems have a
ramp profile, results from a left manoeuvring for a given
steering angle as shown in Fig. 7.

2.3. User Constraints (nominal case)

Fig. 4. Anti-roll system diagram. Different user constraints need to be specified in order to
design the two controllers. For this application, the
constraints are given as:
 The crossover frequency 𝜔𝜔𝑢𝑢 = 2𝜋𝜋10 𝑟𝑟𝑟𝑟𝑟𝑟/𝑠𝑠, which is
about four times rapid than that of the frequency of
the driver. Knowing that the range of driver frequency
is [0 – 2.5] Hz.
 The phase margin 𝑀𝑀𝜙𝜙 > 𝑀𝑀𝜙𝜙𝜙𝜙𝜙𝜙𝜙𝜙 = 40°.
Fig. 5. Passive anti-roll system diagram.
 Zero static error.
Fig. 6 shows the Bode responses for the three plants  Torque control limit : 𝑚𝑚𝑚𝑚𝑚𝑚 [𝑐𝑐𝑎𝑎𝑎𝑎 (𝑡𝑡)] < 16 000 𝑁𝑁.
showing the gain variation results from the variation of 𝐼𝐼𝑥𝑥𝑥𝑥
shown in Table 1, where the phase angles are constant. 3. CONTROLLERS
Table 1. Variation of Uncertain parameters. To obtain a significant comparison between PID and first
min med max generation CRONE controllers, two requirements should
𝑀𝑀𝑇𝑇 [𝑘𝑘𝑘𝑘] 600 750 900 be respected: Iso-rapidity, which means same crossover
frequency 𝜔𝜔𝑢𝑢 of the open loop system. Second, the Iso-
𝐼𝐼𝑥𝑥𝑥𝑥 [𝑘𝑘𝑘𝑘. 𝑚𝑚2 ] 150 225 300

735
IFAC PID 2018
736
Ghent, Belgium, May 9-11, 2018 Hussein Termous et al. / IFAC PapersOnLine 51-4 (2018) 733–738

degree of stability, which means same 𝑀𝑀𝜙𝜙 . In this paper, that for nominal case, both have same rapidity (𝜔𝜔𝑢𝑢 ) and
phase margin is given as 𝑀𝑀𝜙𝜙 = 45°. For both controllers same degree of stability (𝑀𝑀𝑀𝑀).With CRONE controller
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃, where 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 is the nominal plant. (Fig. 10-b), the phase margin remains constant for a range
of frequencies where 𝜔𝜔𝑢𝑢 can vary. This illustrates the
3.1. PID Controller stability degree robustness versus gain uncertainties in
frequency domain. Fig. 11 and 12 present the different
According to (Abi Zeid Daou et al., 2013; Morand et al., sensitivity functions obtained with PID and CRONE
2016) and for 𝜔𝜔𝑚𝑚 = 𝜔𝜔𝑢𝑢 and 𝜙𝜙1 = 60°, the transfer controller. It is obvious that with CRONE, 𝑇𝑇(𝑠𝑠) and 𝑆𝑆(𝑠𝑠)
function of 𝑃𝑃𝑃𝑃𝐷𝐷 controller obtained is as follows: have a robust factor of resonance.

 1  s /108.8  1  s / 48.2 
CPID  s   3616   . (10)
 s /108.8  1  s /82 
3.2. First Generation CRONE Controller

As ordered in (Abi Zeid Daou et al., 2013; Morand et al.,

2016), and choosing 𝜔𝜔𝐵𝐵 = 𝜔𝜔𝑢𝑢 , 𝐴𝐴 = 𝐵𝐵 = 10, and 𝑚𝑚𝑙𝑙 =
𝑚𝑚ℎ = 1, the computed fractional first generation transfer
function is as follows:
0.5
 1  s / 3.94  1  s / 3.94   1 
C f rac ( s)  37607     .
 s / 3.9  1  s / 628   1  s / 628  (11)

Then, the parameters and the recursive zeros and poles of

the rational transfer function are shown in Table 2. Fig. 8. Bode diagram representing the two controllers in
frequency domain.
Table 2. Parameters of CRONE controller.
4.2. Robustness analysis in time domain
Value Value
Parameter Parameter
(𝑟𝑟𝑟𝑟𝑟𝑟/𝑠𝑠) (𝑟𝑟𝑟𝑟𝑟𝑟/𝑠𝑠) Fig. 13 shows the command input 𝑐𝑐𝑎𝑎𝑎𝑎 (𝑡𝑡) of each controller
𝑁𝑁 4 𝜔𝜔𝑧𝑧3 68.5 while introducing a disturbance having a step profile of
𝛼𝛼 𝜂𝜂 3.55 𝜔𝜔𝑧𝑧4 243 magnitude 𝑐𝑐𝑟𝑟0 = 500𝑁𝑁 to the closed loop system. It is
𝛼𝛼 1.884 𝜔𝜔𝑝𝑝1 10.2 shown in Fig.13-b, with CRONE controller, that the three
𝜂𝜂 1.884 𝜔𝜔𝑝𝑝2 36.3 models have the same first overshoot and that the damping
5.4 𝜔𝜔𝑝𝑝3 129 remains constant. This illustrates the stability degree
𝜔𝜔𝑧𝑧1
robustness versus gain uncertainties in time domain.
𝜔𝜔𝑧𝑧2 19.3 𝜔𝜔𝑝𝑝4 458
4.3. Performance of the Anti-Roll System with 14 DOF Model
Finally, the rational transfer function 𝐶𝐶𝑟𝑟𝑟𝑟𝑟𝑟 (𝑠𝑠) can be
represented as: In the following, a comparison done between the
 1  s / 3.94  1  s / zi 
N 4
1  performance of the passive anti-roll system, an anti-roll
Crat ( s)  37607    . (12) system with PID controller, and an anti-roll system with
 s / 3.94  i 1  s /  pi  1  s / 628 
first generation CRONE controller. Fig. 14, 15 and 16
Fig.8 shows the bode diagrams of both controllers where it show the effectiveness of the controlled systems over the
can be seen that they present the same gain and phase at 𝜔𝜔𝑢𝑢 . passive system in compensating roll dynamics, the roll
rate, the roll angle and roll acceleration respectively. In
4. PERFORMANCE ANALYSIS addition to that, Table 3 presents the numerical results
obtained comparing the results of controlled systems with
In this section, the robustness analysis is presented for both respect to the passive one. It is clear that both controllers
controllers with the feedback control input only (without the achieved a satisfied performance, with very high
feedforward part), in the frequency and time domain. Then compensation, in which the human body cannot perceive
the performance analysis is presented for the controlled them. Moreover, the command signals produced by the
anti-roll systems, including the feedforward control part, controllers are also presented in Fig. 17, in comparison
applied on 14 DOF model (Morand et al., 2016) with the perturbation introduced.

4.1. Robustness analysis in frequency domain 4. CONCLUSION

Fig. 9 shows the Bode plot responses of the open loop This article presents the effectiveness of the first
transfer functions for both controllers. Fig. 10 represents generation CRONE control in achieving robustness in time
the Nichols loci obtained with two controllers. It is shown and frequency domain, in spite of variation of the plant

736
 IFAC PID 2018
Ghent, Belgium, May 9-11, 2018 Hussein Termous et al. / IFAC PapersOnLine 51-4 (2018) 733–738 737

parameters. The two controllers show their good

performance in compensating the roll dynamics, which
improve ride comfort for passengers. The next step of this
work is to implement these controllers with the other local
controllers in the frame of Global Chassis Control.
Table 3. Time domain numerical results.
(a) (b)
Roll motion Fig. 13. Step responses obtained with: (a) PID, (b) CRONE.
(max[controller]/max[Passive]
Acceleration Rate Angle
PID 0.1035 0.0327 0.0271 ACKNOWLEDGEMENT
CRONE 0.1242 0.0579 0.0706
This work was supported by the Lebanese research
program and the AUF-CNRSL-UL program.

REFEFRENCE

Abi Zeid Daou, R., Moreau, X., Francis, C., 2013.

Regulation of a pump-tank-sensor system using two
fractional-order controllers. J. Appl. Nonlinear Dyn.
2, 355–371.
Bennett, S., 2001. The past of PID controllers. Annu. Rev.
Control 25, 43–53. [Link]
(a) (b) 5788(01)00005-0
Fig. 9. Open loop bode diagram: (a) PID, (b) CRONE Brown, S.N., 2005. Vehicle suspension.
Chokor, A., Talj, R., Charara, A., Shraim, H., Francis, C.,
2016. Active suspension control to improve
passengers comfort and vehicle’s stability, in: 2016
IEEE 19th International Conference on Intelligent
Transportation Systems (ITSC). pp. 296–301.
[Link]
De Oliveira, V., Karimi, A., 2012. Robust and gain-
scheduled PID controller design for condensing
boilers by linear programming. IFAC Proc. Vol. 2,
(a) (b) 335–340. [Link]
Fig. 10. Open-loop Nichols plot with: (a) PID, (b) CRONE. 3014.00057
Garpinger, O., Hägglund, T., Åström, K.J., 2014.
Performance and robustness trade-offs in PID
control. J. Process Control 24, 568–577.
[Link]
Lanusse, P., 2010. CRONE control system design, a
CRONE toolbox for Matlab.
Laurent, D., Sebe, M., Walser, D., 2000. Assembly
comprising a wheel and a suspension integrated into
the wheel.
(a) (b)
Morand, A., Moreau, X., Melchior, P., Moze, M.,
Guillemard, F., 2016. CRONE cruise control system.
Fig. 11. T(s) obtained with: (a) PID, (b) CRONE. IEEE Trans. Veh. Technol. 65, 15–28.
[Link]
Moreau, X., Oustaloup, A., Nouillant, C., n.d. From
analysis to synthesis of vehicle suspensions : the
CRONE approach. Eur. Control Conf. ECC’99.
Moreau, X., Rizzo, A., Oustaloup, A., 2009. Application
of the CRONE control-design method to a low-
frequency active suspension system. Int. J. Veh.
Auton. Syst. 7, 172–200.
Oustaloup, A., 1991. La commande CRONE. Hermès,
(a) (b) Paris.
Fig. 12. S(s) obtained with: (a) PID, (b) CRONE.

737
IFAC PID 2018
Ghent, Belgium, May 9-11, 2018
738 Hussein Termous et al. / IFAC PapersOnLine 51-4 (2018) 733–738

(a) (b) (c)

Fig. 14. Roll rate (a) Passive, (b) PID , and (c) CRONE.

(a) (b) (c)

Fig. 15. Roll angle (a) Passive, (b) PID , and (c) CRONE.

(a) (b) (c)

Fig. 16. Roll acceleration: (a) Passive, (b) PID , and (c) CRONE.

(a) (b) (c)

Fig. 17. Torques 𝑐𝑐𝑟𝑟0 (𝑡𝑡) ‘ − −‘ and 𝑐𝑐𝑎𝑎𝑎𝑎 (𝑡𝑡) ‘— ‘ : (a) Passive, (b) PID , and (c) CRONE.

738
View publication stats