art. bianchi
TRANSCRIPT
-
8/10/2019 Art. Bianchi
1/9
-
8/10/2019 Art. Bianchi
2/9
The contribution of the present work is twofold. First, we pro-
pose a fullyintegrated controller, achieving the tracking of yaw
rate, lateral velocity and roll dynamics (angle and velocity), by
means of three actuators, in nominal conditions. Second, we
propose a methodology, based on a hybrid controller, for the
management of the actuators in saturation conditions, based on
fixed priorities on the states to be tracked, according to theirimportance for the vehicle attitude.
The paper is organized as follows. In Section 2, the mathemat-
ical model of the vehicle is presented, and the control problem
is stated. In Section 3, the nonlinear technique is developed
and the saturation management is described. In Section 4, the
proposed controller is tested with simulations and comparisons.
Some comments conclude the paper.
2. MATHEMATICAL MODEL AND PROBLEM
FORMULATION
For simplicity, we consider the model of a rearwheel driven
vehicle. The actuators considered in this work are
1. Active Front Steer (AFS), which imposes an incremental
steer angle on top of the drivers input. The control is then
actuated through the front axle tire characteristic.
2. Rear Torque Vectoring (RTV), which distributes the
torque in the rear axle, usually to improve vehicle trac-
tion, handling and stability. The control is then actuated
through the rear axle tire characteristics.
3. Semiactive suspensions (SAS), which are able to change
the damping coefficient of the shock absorber in a contin-
uous interval, differently from passive systems.
The mathematical model is derived under the following as-
sumptions, which are verified in a large number of situations
and which mitigate the complexity of the vehicle dynamics
The vehicle moves on a horizontal plane;
The longitudinal velocity is constant, so that vehicle shak-
ing/pitch motions can be neglected;
The steering system is rigid, so that the angular position
of the front wheels is uniquely determined by the steering
wheel position;
The wheels masses are much lower than the vehicle one,
so the steering action does not affect the position of thecentre of mass of the entire vehicle;
The vehicle takes large radius bends and the road wheel
angles are small (less than 10), i.e. the bend curveradius is much higher than the vehicle width;
The aerodynamic resistance and the wind lateral thrust are
not considered;
The tire vertical loads are constant;
The actuators are ideally modelled.
As a consequence of the previous assumptions, the vehicle
model has four degrees of freedom: lateral velocity, yaw rate,
roll angle and velocity. Following Guiggiani [2007], one can
consider the lateral acceleration of the vehicle
m(vy+ vxz) =(Fy,f+Fy,r) +mshdx (1)
where m is the vehicle mass, vy, z, xare the lateral, yaw, androll velocities, hd = h d,his the center of gravity height, disthe roll center height,ms is the sprung mass, is the roadtirefriction coefficient, andFy,f, Fy,r the lateral (front/rear) tireforces. The lateral forcesFy,f = Fy,f(f),Fy,r = Fy,r(r)depend on the slip angles
f =c+f,0= c+d+fx vy+lfz
vx
r =rxvy lrz
vxwhere c, d are the control, driver components of the wheelangle,f, r are front/rear sensitivities with respect to the rollanglex, andlf,lr are the distances from the center of gravityto the front/rear axle. A simple but accurate representation of
the lateral force functions is given by (Pacejka [2005])
Fy,f(f) =Cy,fsin(Ay,farctan(By,ff))
Fy,r(r) =Cy,rsin(Ay,rarctan(By,rr))(2)
where Ay,f, Ay,r , By,f, By,r , Cy,f, Cy,r are experimentalconstants.
The vehicle yaw dynamics can be expressed considering the
presence of RTV actuators
Jzz =(Fy,flf Fy,rlr) +Mz+Jzxx (3)
whereJz is the vehicle inertia momentum about thez axis,Jzxis the product of inertia about the axes z, x, andMzis RTV yawmoment.
The vehicle roll angular acceleration can be expressed as
x = x
Jrx = bxx (kx msghd)x+Jzxz
+mshd(vy+ vxz)
(4)
where Jx is the vehicle inertia momentum about the x axis,Jr = Jx + msh2d, bx is the suspension roll damping, kx isthe suspension roll stiffness, and g is the gravity accelerationconstant.
From (1), (3), (4) we get the mathematical model of a vehicle
with yaw, lateral and roll dynamics
z = 1
Jz,e
(Fy,flf Fy,rlr) +Mz
+kmkzhe(Fy,f+Fy,r)
kmkz
bxx+ (kx msghd)x
vy =zvx+ 1
me(Fy,f+Fy,r)
+kmkzhe
(Fy,flf Fy,rlr) +Mz
kmhe
bxx+ (kx msghd)x
x= km
bxx+ (kx msghd)x
+kmkz
(Fy,flf Fy,r lr) +Mz
+kmhe(Fy,f+Fy,r)
x= x
(5)
-
8/10/2019 Art. Bianchi
3/9
wherekm= 1/Jx,eand
Jx,e=JrJ2zx
Jz
m2sh2d
m
me= m
1 +mskmhdhe, Jz,e =
Jz1 +kmkzJzx
he=ms
mhd, kz =
JzxJz .
As anticipated, we considervx constant. Moreover, the controlinputs that we consider areMz , and the differences
Fy,f =Fy,f Fy,f,0, Fy,f,0:= Fy,f(f,0)
bx= bx bx,0
with bx,0 the damping when the SAS system is not active.Clearly, the real active front input is the control angle c whichcan be determined by inverting (2), obtaining
c =
d+ vy+lfz
vx+F1y,f(
F) if|F| Fy,f(f,sat)
d+ vy+lfz
vx f,sat otherwise
with Fa fixed value to be imposed by the AFS.
The control aim is to track asymptotically some bounded refer-
ences, with bounded derivatives, for z,vy,x andx. Moreprecisely, the reference generator is
z,ref= 1
Jz,ref
ref(Fy,f,reflf Fy,r,reflr)
+kmkzheref(Fy,f,ref+ Fy,r,ref)
kmkz
bxx,ref+ (kx msghd)x,ref
vy,ref=z,refvx+ 1
meref(Fy,f,ref+ Fy,r,ref)
+kmkzhe
ref(Fy,f,reflf Fy,r,reflr)
kmhe
bxx,ref+ (kx msghd)x,ref
x,ref=km
bx,refx,ref+ kx,refx,ref
+kmkz
ref(Fy,f,reflf Fy,r,reflr)
+kmheref(Fy,f,ref+ Fy,r,ref)
x,ref=x,ref
(6)
where Jz,ref, ref are appropriate parameters and Fy,f,ref,Fy,r,refare ideal curves, depending on
f,ref=d+fx,refvy,ref+ lfz,ref
vx
r,ref=rx,refvy,ref lrz,ref
vx.
In particular, we set
Fy,j,ref(j,ref) =
=
Fy,j(j,ref) +F
y,j(j,ref)(j,ref+ j,ref)j,ref j,ref
withj = f, r, where j,ref is a limit value for the slip angle,above which the control action is needed.
3. DESIGN OF A STATEFEEDBACK LINEARIZING
CONTROL LAW WITH SATURATION MANAGEMENT
In this section a statefeedback linearizing control, imposing
desired behaviors for yaw, lateral and roll dynamics, is de-
signed. Then a saturation management algorithm is proposed.
3.1 StateFeedback Linearizing Control Law
Let us impose the following control requirements
z =d1:= z,ref k1(z z,ref)
vy =d2:= vy,ref k2(vy vy,ref)x= d3:= x,ref k3(x x,ref) k4(x x,ref)
(7)
ki > 0,i = 1, , 4. To fulfill (7), the following control lawhas to be imposed
u1
u2
u3
=
Fy,f
Mz
bx
=C1
r1
r2
r3
(8)
where
C=
kmkzhe+ lfJz,e
1Jz,e
kmkzx
1me
+kmkzhelf
kmkzhe kmhex
km(he+kzlf) kmkz kmx
(9)
and
r1= d1 kmkzheFy 1
Jz,eTy+kmkzDy
r2= d2+zvx 1
meFy kmkzheTy+ kmheDy
r3= d3 kmheFy kmkzTy+kmDy
Fy =(Fy,f,0+Fy,r)
Ty =(Fy,f,0lf Fy,rlr)
Dy =bx,0x+ (kx msghd)x.
In the following we assume that
kmk2zJz,e = 1, meh
2ekm= 1
which is reasonable since this is verified for an appropriate
choice of the vehicle parameters. Therefore, since the firstcolumn ofCdepends on and the third onx, Cis invertible ifand only if= 0andx= 0. The inverse of (9) is
-
8/10/2019 Art. Bianchi
4/9
C1 =
0 1
me1meh2ekm
Jz,e1kmk2zJz,e
melf
1meh2ekm
1x
kzJz,e1kmk2zJz,e
1x
mehe1meh2ekm
1
mehe1meh2ekm
mehelf1meh2ekm
kzJz,e1kmk2zJz,e
1x
(1meh2
ek2
mk2
zJz,e)km(1meh2ekm)(1kmk
2zJz,e)
.
Therefore, it is clear that when 0 and/or x is zero,the control law (8) is not computable. Physically, when these
situations occur, some actuators saturate.
3.2 Saturation Management
Saturations may occur when = 0,x = 0. In the followingwe propose a hybrid approach to the design of a procedure to
manage saturation in the actuators. Withja = 1we denote theAFS, with ja = 2 the RTV, and with ja = 3 the SAS. If thejtha actuator saturates,ja = 1, 2, 3, the corresponding control isuja =uja,max.
Considering the control requirements expressed by (7), we
assign higher priority to the fulfillment of those for z, andlower priority to those forx. The actuators that do not saturatewill ensure the fulfillment of the requirements with higher
priority. To show the procedure in general, let
xi = di, i= 1, , m (10)
be the set of control requirements, generalizing (7), orderedaccording to the priority (first = higher priority, last = lower
priority), and let
Cu= C
u1...
um
=
r1...
rm
=R (11)
be the general equation for determining the control law satis-
fying (10). We partition u into a vector of nonsaturating andsaturating inputs
u=
unsus
, uns R
mns, us Rms, mns+ms = m.
We operate the corresponding partition on C, so that (11)becomes
Cns Cs
unsus
=R.
Finally, to select the firstmnscontrol requirementsCns Cs
=
Imnsmns 0mnsms
Cns Cs
R=
Imnsmns 0mnsms
R
so that one calculates the nonsaturating controls
uns= C1ns
R Csus
.
To illustrate the proposed saturation management, let us con-sider some examples in the case under study. Let us assume that
the SAS actuator, calculated according to (8), saturates. Hence,
Fy,f
Mz
=
kmkzhe+
lfJz,e
1Jz,e
1me
+kmkzhelf
kmkzhe
1
r1r2
bx,max
kmkzxkmhex
.
Note that if the saturation is due to the fact that x = 0, thiscontrol law reduces
Fy,f
Mz
=
kmkzhe+
lfJz,e
1Jz,e
1me
+kmkzhelf
kmkzhe
1
r1r2
whilebx= bx,max.
Analogously, if both SAS and AFS saturate, one has
Mz =Jz,e
r1 Fy,f,max
kmkzhe+
lfJz,e
bx,maxkmkzx
and if the saturations are due to= 0,x= 0, this control lawreduces to
Fy,f =Fy,f,max, Mz =Jz,er1, bx= bx,max.
4. SIMULATION RESULTS
In this section, we provide simulation results of the proposed
control technique. We consider two simulation sets. First, we
compare the proposed integrated solution with a linearizing
feedback controller (Borri [2007], Bianchi [2008]) based on a
lower number of actuators. Second, we show the improvement
ensured by the hybrid saturation management scheme with
respect to a classical fixed saturation scheme.
The parameters of the vehicle are
m= 1550 kg ms= 150 kg = 1
lf = 1.17 m lr = 1.43 m hd= 0.5 m
Jz = 2300 kg m2 Jx = 350 kg m2 Jzx = 50 kg m2
Ayf = 1.81 Byf = 7.2 Cyf= 8854
Ayr = 1.68 Byr = 11 Cyr = 8394
kx= 150, 000 Nm/rad bx,0= 7, 000 Nm rad/s
f =0.05 r = 0.05
while for the reference generation we have considered the
following set of values
Jz,ref=Jz ref= f,ref= 0.08
r,ref= 0.04 kx,ref=kx bx,ref= 11800.
The control inputs are restricted to the following intervals (see
Spelta [2010])
Fy,f[0.95Cyf Fy,f,0, 0.95Cyf Fy,f,0]
Mz [10000, 10000], bx [2500, 35000].
4.1 Comparison with other controllers
The first test maneuver is a step steer of120
with longitudinalvelocity of28m/s (100.8 km/h). It is useful to compare the per-formance of the controller with the AFS, RTV, SAS actuators,
-
8/10/2019 Art. Bianchi
5/9
defined in Section 3.1, with a state feedback linearizing control
law with two actuators (AFS and RTV), as defined in Borri
[2007], Bianchi [2008], designed to track lateral velocity and
yaw rate. Thanks to the presence of the SAS, whose primary
goal is the reduction of the oscillating behavior, the proposed
controller obviously performs better results for the roll track-
ing, namely x, x track more closely the references x,ref,x,ref(see Figures 3, 4), as expected. Here the reference modelhas an increased value of the suspension roll damping (i.e.
bx,ref> bx,0). Moreover, the additional actuator influences thebehavior of the other states and inputs, as shown in Figures 1,
2. In particular, one can note an improved tracking of the lateral
velocity (Figure 2), while the yaw rate behavior is similar in
both cases (Figure 1), and is nonzero because of the AFS
saturation (Figure 5). An analysis of the control inputs also
shows higher values of RTV action in the presence of three
actuators.
0 0.5 1 1.5 2 2.5 30
5
10
15
20
25
30
35
(a)
Fig. 1. (a) z,ref(dashdot), zwith 3 actuators (solid), zwith2 actuators (dashed) [deg/s] vs time [s]
0 0.5 1 1.5 2 2.5 31.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0.2(a)
Fig. 2. (a)vy,ref (dashdot), vy with 3 actuators (solid), vy with2 actuators (dashed) [m/s] vs time [s]
0 0.5 1 1.5 2 2.5 30
0.01
0.02
0.03
0.04
0.05
0.06(a)
0 0.5 1 1.5 2 2.5 38
6
4
2
0
2
4
6
8
10x 10
3 (b)
Fig. 3. (a) x,ref (dashdot), x with 3 actuators (solid), xwith 2 actuators (dashed) [deg/s] vs time [s]; (b) ex =x,refx with 3 actuators (solid),ex = x,refxwith 2 actuators (dashed) [deg/s] vs time [s]
4.2 Performance of the Saturation Management
In this section we compare the proposed controller Csm, withsaturation management, and a controller Cw, in which the
saturation is treated, as done usually in applications, simply bysaturating the input value, without a change of priority in the
control specifications.
The maneuver considered, shown in Figure 10.d, is a double
step steer of 120 with longitudinal velocity of33 m/s (118km/h). Figures 86 show the evolution of the state variables
with the two controllers. The algorithm of saturation man-
agement gives tracking priority to z , as shown in Figure 6:while the error withCw grows immediately, as consequence ofthe AFS saturation shown in Figure 11.a, the controller Csmsucceeds to reduce the errorez . This improvement is achievedthanks to the RTV compensation. Clearly, this is paid with a
larger error on lateral velocity tracking (see Fig. 7), because ofthe priority scheme. However, with Csm and with the priorityon the yaw rate tracking, the handling of the vehicle has been
-
8/10/2019 Art. Bianchi
6/9
0 0.5 1 1.5 2 2.5 30.05
0
0.05
0.1
0.15
0.2
0.25
0.3(a)
0 0.5 1 1.5 2 2.5 30.08
0.06
0.04
0.02
0
0.02
0.04
0.06
0.08(b)
Fig.4. (a) x,ref(dashdot), xwith 3 actuators (solid), xwith2 actuators (dashed) [m/s] vs time [s]; (b)ex =x,refx with 3 actuators (solid), ex = x,ref x with 2actuators (dashed) [m/s] vs time [s]
improved, as testified by the spatial trajectory (see Fig. 12),
although this is notone of the control goals. In fact, an enhanced
yaw tracking improves the agility and cornering capabilities of
the vehicle. The roll tracking is displayed in Figures 9, 8, with a
performance that can be considered comparable in both cases.
Figure 10.a shows the total force of the front axle, Figure 10.b
the yaw moment, and Figure 10.c the damping coefficient vari-
ation. The control anglecdue to AFS is shown in Figure 10.d,which results to be lower, in absolute value, with Csm thanwith Cw, because most of work in tracking z is due to theadditional RTV action. This does not mean that the front axle
is not exploited at its maximum extent, since the AFS soon
saturates in both cases (see Fig. 11). We also note that the
SAS system saturates very soon after the AFS saturation, in
both cases, meaning that this actuator is not able to compensate
the lack in control action, due to the saturation of the other
actuators.
In general, we can conclude that the proposed algorithm for the
saturation management is able to achieve a better and deeper
0 0.5 1 1.5 2 2.5 30
5000
10000(a)
0 0.5 1 1.5 2 2.5 32000
1000
0
1000
(b)
0 0.5 1 1.5 2 2.5 31
0
1
2x 10
4 (c)
Fig. 5. (a) Fy,fwith 3 actuators (solid), Fy,fwith 2 actuators(dashed) [N] vs time [s]; (b) Mz with 3 actuators (solid),Mzwith 2 actuators (dashed) [Nm] vs time [s]; (c)bxwith3 actuators (solid), bx with 2 actuators (dashed) [Nmrad/s]vs time [s]
utilization of all the actuators in saturation conditions, resulting
in an improved handling and cornering behavior.
CONCLUSIONS
In this paper, we have analyzed a hybrid management of the
control saturations, in the attitude control of a vehicle which
considers also the roll dynamics. A hybrid control of three actu-
ators allowed solving a tracking control problem and managing
critical conditions, when multiple actuator saturations occur.Simulations results show the improvement of the proposed so-
lution with respect to existing controllers and simpler saturation
schemes.
This work suggests a number of future directions of research.
From the theoretical point of view, the hybrid controller poses
problems regarding the stability in presence of discontinuity
of the control action. Additionally, it would be interesting
to extend the simulation setup to more complex simulation
environments (e.g. Carsim), to show the robustness of our
approach in presence of unmodelled dynamics.
REFERENCESM. Guiggiani,Dinamica del Veicolo, Second Edition (in Ital-
ian), CittaStudi Edizioni, Torino, Italy, 2007.
H. Pacejka,Tyre and Vehicle Dynamics, Elsevier Butterworth
Heinemann, 2005.
G. Burgio, and P. Zegelaar, Integrated Vehicle Control Using
Steering and Brakes, International Journal of Control, Vol.
79, No. 2, pp. 162169, 2006.
J. Ackermann, J. Guldner, R. Steinhausner, and V. Utkin, Linear
and nonlinear design for robust automatic steering, IEEE
Transactions on Control System Technology, Vol. 3, No. 1,
pp. 132143, 1995.
R. A. Williams, Automotive active suspensions Part I: Basicprinciples, IMechE Proc., Part D, vol. 211, pp. 415426,
1997.
-
8/10/2019 Art. Bianchi
7/9
0 1 2 3 4 5 650
40
30
20
10
0
10
20
30
40(a)
0 1 2 3 4 5 610
8
6
4
2
0
2
4
6
8(b)
Fig. 6. (a)z,ref (dashdot), z with the controller Csm(solid),z with the controller Cw (dashed) [deg/s vs s]; (b) ez =z,ref zwith the controller Csm(solid),ez =z,refz with the controller Cw (dashed) [deg/s vs s]
S. M. Savaresi, and C. Spelta, A Single-Sensor Control Strategy
for Semi-Active Suspensions, IEEE Transaction on Control
System Technology, Vol. 17, No. 1, pp 143-152, 2009.
C. Spelta, F. Previdi, S. M. Savaresi, P. Bolzern, M. Cutini, C.
Bisaglia, A novel Control strategy for Semi-Active suspen-sions with variable damping and stiffness,Proceedings of the
2010 American Control Conference, Baltimore, MD, USA,
pp. 4582-4587, 2010.
M. Canale, M. Milanese, and C. Novara, SemiActive Suspen-
sion Control Using Fast ModelPredictive Techniques, IEEE
Transaction on Control System Technology, Vol. 14, No. 6,
pp. 10341046, 2006.
S. C. Baslamisli, I. Polat, and I. E. Kose, Gain Scheduled
Active Steering Control Based on a Parametric Bicycle
Model, IEEE Intelligent Vehicles Symposium, pp. 1168
1173, 2007.
A. Borri,Integrated Vehicle Control using Active Front Steer-ing and Active Differential, Master thesis, University of
LAquila, 2007.
0 1 2 3 4 5 65
4
3
2
1
0
1
2
3
4(a)
0 1 2 3 4 5 61.5
1
0.5
0
0.5
1
1.5
2
2.5(b)
Fig. 7. (a)vy,ref (dashdot),vy with the controller Csm(solid),vy with the controller Cw (dashed) [m/s vs s]; (b) evy =vy,refvywith the controller Csm(solid), evy =vy,refvywith the controller Cw(dashed) [m/s vs s]
0 1 2 3 4 5 60.1
0.08
0.06
0.04
0.02
0
0.02
0.04
0.06
0.08(a)
Fig. 8.x,ref(dashdot), xwith the controller Csm(solid), xwith the controller Cw(dashed) [deg/s vs s]
-
8/10/2019 Art. Bianchi
8/9
0 1 2 3 4 5 60.6
0.5
0.4
0.3
0.2
0.1
0
0.1
0.2
0.3(a)
Fig. 9.x,ref (dashdot), xwith the controller Csm(solid),x
with the controller Cw (dashed) [m/s vs s]
D. Bianchi,Study and Improvements oriented to robustness of
Integrated Control of Vehicle with Active Front Steering and
Active Differential, Master thesis, University of LAquila,
2008.
D. Bianchi, A. Borri, G. Burgio, M. D. Di Benedetto, and S. Di
Gennaro, Adaptive Integrated Vehicle Control using Active
Front Steering and Rear Torque Vectoring, Proceedings of
the48th IEEE Conference on Decision and Control, Shang-hai, China, 2009.
D. Bianchi, A. Borri, G. Burgio, M. D. Di Benedetto, and S.
Di Gennaro, Adaptive Integrated Vehicle Control using Ac-tive Front Steering and Rear Torque Vectoring,International
Journal of Vehicle Autonomous Systems, 2010 (in press).
E. Esmailzadeh, A. Goodarzi, and G. R. Vossoughi, Optimal
Yaw Moment Control Law for Improved Vehicle Handling,
Mechatronics, Vol. 13, No. 7, pp. 659675, 2003.
R. Karbalaei, A. Ghaffari, R. Kazemi, and S. H. Tabatabaei,
Design of an Integrated AFS/DYC based on fuzzy logic
control, IEEE International Conference on Volume Vehicular
Electronics and Safety, pp. 16, 2007.
S. Malan, M. Taragna, P. Borodani, and L. Gortan, Robust
performance design for a car steering device, Proceedings
of the33rd IEEE Conference on Decision and Control, Lake
Buena Vista, FL, USA, pp. 474479, 1994.R. Marino, and P. Tomei,Nonlinear control design: geometric,
adaptive and robust, Prentice Hall International, Hertford-
shire (UK), 1996.
R. Rajamani, Vehicle Dynamics and Control, Mechanical En-
gineering Series, SpringerVerlag, 2006.
S. Sastry, and A. Isidori, Adaptive Control of Linearizable
Systems,IEEE Transactions on Automatic Control, Vol. 34,
No. 11, pp. 11231131, 1989.
P. Setlur, J. R. Wagner, D. M. Dawson, and D. Braganza,
A Trajectory Tracking SteerbyWire Control System for
Ground Vehicles,IEEE Transactions on Vehicular Technol-
ogy, Vol. 55, No. 1, pp. 7685, 2006.J. J. E. Slotine, and W. Li, Applied Nonlinear Control, Prentice
Hall, Englewood Cliffs, New Jersey, 1991.
0 1 2 3 4 5 61
0.5
0
0.5
1x 10
4 (a)
0 1 2 3 4 5 61
0.5
0
0.5
1x 10
4 (b)
0 1 2 3 4 5 62
0
2
4x 10
4 (c)
0 1 2 3 4 5 615
10
5
0
5
10
15(d)
Fig. 10. (a)Fy,f with the controller Csm (solid),Fy,fwith thecontroller Cw(dashed) [N vs s]; (b)Mzwith the controllerCsm(solid), Mzwith the controller Cw(dashed) [Nm vs s];(c) bxwith the controller Csm(solid), bxwith the controllerCw(dashed) [N m rad/s vs s]; (d) cwith the controller Csm(solid),c with the controller Cw (dashed),d (dashdot)[deg vs s]
A. Y. Ungoren, H. Peng, and H. E. Tseng, A Study on Lateral
Speed Estimation Methods,International Journal of Vehicle
Autonomous Systems, Vol. 2, No. 1/2, pp. 126144, 2006.
Z. Lin, A. A. Stoorvogel, and A. Saberi, Output regulation for
linear systems subject to input saturation, Automatica, Vol.
32, No. 1, pp. 29-47, 1996.
T. H. Yang, and E. Polak, Moving horizon control of nonlinear
systems with input saturation, disturbances and plant un-
certainty, International Journal of Control, 58(4), 875903,
1993.
-
8/10/2019 Art. Bianchi
9/9
0 1 2 3 4 5 6
0
0.5
1
(a)
0 1 2 3 4 5 6
0
0.5
1
(b)
0 1 2 3 4 5 6
0
0.5
1
(c)
Fig. 11. Actuator saturation (1) or out of saturation (0) vstime [s], in the case of the controller Csm(solid) or withwith the controller Cw (stars): (a)Fy,f; (b)Mz; (c)bx
0 20 40 60 80 100 120 140 160 180 2000
5
10
15
20
25
30
35
Trajectory x (m)
Trajectoryy(m)
Fig. 12. Trajectory in the plane: reference (dashdot), trajectorywith the controller Csm (solid), trajectory with the con-troller Cw(dashed)