Sei sulla pagina 1di 5

Robust Control of UAVs using H

Control
Paradigm
Muhammad Zeeshan Babar, Syed Ussama Ali, M. Zamurad Shah, Raza Samar, Aamer Iqbal Bhatti
Dept. of Electronic Engineering, Muhammad Ali Jinnah University, Islamabad, Pakistan
zeeshan.babar@jinnah.edu.pk, ussama.ali@gmail.com, zamurad@gmail.com, raza.samar@gmail.com, aib@jinnah.edu.pk
AbstractIn recent past, the use of Unmanned Aerial Vehicles
(UAVs) has been progressively regnant and signicant in several
applications related to civil as well as military purposes. Due to
their numerous benets, the control and dynamic modeling of
UAVs are becoming the attractive eld of research. Considering
all these facts, this paper has present the conventional inner-outer
loop strategy with some proposed modications and a robust
controller is designed for the inner control loop. Due to uncertain
nature of aircraft, Robust H Loop Shaping Design Procedure
(LSDP) is used to synthesis the controller. The designed inner
loop controller is applied to the system and the simulation results
indicate the good performance and robustness margins of the
designed controller against the parameter uncertainties.
I. INTRODUCTION
The Unmanned Aerial Vehicles (UAVs) are growingly reign-
ing in the twenty-rst century. The use of UAVs in the danger-
ous positions and grievous situations has reduced the risk of
human life. The UAVs are designed for precise missions and
maneuvers while an aircraft with pilot is incompatible for it.
Its small size makes it possible to operate in stiff spaces, man
portable and to enjoy the distinctive conguration. The UAVs
can also be used for surveillance as well as reconnaissance
missions in the dangerous areas. Due to these characteristics,
the demand for the UAV development and utilization has seen
unprecedented levels of growth not only in military but also in
civilian applications. These benets motivate many researchers
to carry out useful work towards the control system design
of a UAV. However, the control of UAV is a challenging
work as the UAV is a multi-input multi-output (MIMO), under
actuated, unstable and highly coupled system.
Two paradigms are used for the trajectory tracking of air
vehicles. In the rst strategy, the inner-outer loop separation
principle is used conventionally. In outer loop, guidance is
designed while in the inner loop, the control is designed to
track the command generated by the outer guidance loop. The
guidance loop take the cross track error and heading angle
as input and generate the reference roll angle while in the
inner control loop, the control law generates input commands
to track the reference roll command. In the other approach,
the guidance and control is designed in a single integrated
framework.
The inner-outer loop paradigm, due to its simplicity and
well developed design methods, is conventionally used for
the control of UAVs. The simple proportional derivative (PD)
controller is used by [1]. The results show that it gives good
tracking performance when tracking a straight path but in the
presence of disturbance like wind, the control performance
of PD control demean. The PD control modied with some
nonlinear term is proposed by [2]. This technique improves
the performance in terms of tracking. But no stability proof
of this paradigm is given. Similarly, a nonlinear scheme is
introduced by [3]. This scheme gives good performance than
the PD scheme but in the presence of large cross track error,
the actuator saturation phenomena occurs which may lead
to instability of the system. A sliding mode approach was
introduced by [4]. This technique gives very good results but
some unrealistic assumptions are made in it. There are many
other approaches [5],[6] and [7] that are used in literature for
the trajectory control of UAVs.
Due to uncertain nature of aircraft systems, the robust
trajectory tracking performance in the presence of disturbances
and uncertainties is a challenging task. Over the years, many
traditional control strategies are used for the control of UAVs.
Linear Quadratic Regulator (LQR) optimal control has been
implemented successfully on different air vehicles [8],[9] and
[10]. But the robustness problem in the presence of uncertain-
ties is not addressed. To achieve robust performance and sta-
bilization, robust H

control method given by George Zames


[11] and [12] has been used in literature. This paradigm has
paved its path since the late 1980s to achieve a mature stage
today. This method has very good and desirable characteristics
as it can be implemented to multivariable systems, synthesize
a low order controller and gives the robust stability against
the parameter uncertainties. This method has been intensively
implemented in ight control system for both lateral and
longitudinal dynamics of aircraft [13],[14] and [15].
In this work, H

Loop Shaping Design Procedure (LSDP)


given by K. Glover and D. C. McFarlane [16] and [17] is
used to synthesize the inner loop controller in the two loop
paradigm i.e. Inner-outer loop. The Loop Shaping Design is
a systematic approach to get a controller that guarantees the
robustness and achieves all the performance objectives against
the disturbance and the parameter uncertainties [18]. In inner-
outer loop strategy, it is assumed that the outer guidance
loop generates the desired reference roll angle on the basis of
given cross range and heading angle. While an inner loop H

controller is designed in this work to track the reference roll


command generated by the outer guidance loop. The proposed
scheme is implemented on a 6-DOF simulation model and
the simulation results are presented here which shows the
robustness of the proposed scheme.
The paper structure is as follows: The aircraft model and
its linearization is presented in Section II. In Section III, the
proposed inner loop H

controller is presented. Section IV,


presents the simulation results and the paper work is concluded
in the section V.
II. THE AIRCRAFT MODEL
In dynamic modeling of a UAV, state space realization has
been preferred. In this work, the equation of motions (EOMs)
of UAV have been taken from [19]. The lateral state space
equations given in [19] are represented here.
1) Linear State Space Realization: The general state space
representation is given as:
x(t) = Ax(t) + Bu(t) (1)
y(t) = Cx(t) + Du(t) (2)
In this work, the state vector x(t) and the input vector u(t)
is chosen as:
x(t) =

Roll Rate [rad/sec]


Roll Angle [rad]
Side Slip Angle [rad]
Yaw Rate [rad/sec]

u(t) =

Aileron Command [rad]


Rudder Command [rad]

The state matrix A for the given Lateral system is given as:
A =

Lp
Ixx
0
L

Ixx
Lr
Ixx
1 0 0 0
Yp
mU
g
U
Y

mU
Yr
mU
1
Np
Izz
0
N

+N
t
Izz
Nr
Izz

The input matrix B is given as:


B =

L
a
Ixx
L
r
Ixx
0 0
Y
a
mU
Y
r
mU
N
a
Izz
N
r
Izz

The output matrix C is given as:


C =

0 1 0 0
0 0 0 1

The nal state space representation of the given as:


x =

Lp
Ixx
0
L

Ixx
Lr
Ixx
1 0 0 0
Yp
mU
g
U
Y

mU
Yr
mU
1
Np
Izz
0
N

+N
t
Izz
Nr
Izz

L
a
Ixx
L
r
Ixx
0 0
Y
a
mU
Y
r
mU
N
a
Izz
N
r
Izz

y =

0 1 0 0
0 0 0 1

2) Characteristic Properties and Stability Derivatives: The


UAV considered in this work is a Scaled YAK-54. The stability
derivatives and the other characteristic properties for YAK-
54 are taken from [19][20]. By putting the values of these
stability derivatives and characteristic properties, the state
space representation of the model is given as:
A =

25.6040 0 28.6272 3.2475


1 0 0 0
0.0011 0.2718 0.4707 0.9852
0.6061 0 46.9502 4.5259

B =

482.3682 28.4971
0 0
0 0.3874
4.1029 46.7637

C =

0 1 0 0
0 0 0 1

and
D = 0
III. ROBUST H

CONTROL: THE LOOP SHAPING DESIGN


PROCEDURE (LSDP)
A. Servo Dynamics and Washout Filter
The servo dynamic module are added to both the command
surfaces of the aircraft model to add the dynamic properties
of servo movement. The Servo movements are assumed to be
a rst order transfer function given as:
Servo = H(s) =

H
1
(s) 0
0 H
2
(s)

20
s+20
0
0
20
s+20

During the steady turning ights, a steady yaw rate is


required. A Washout lter is used for this purpose [21]. A
high pass lter is used as a washout lter which only allow the
high frequency content of yaw rate to pass to the controller.
The washout lter in the feedback loop does not move the
equilibrium points of the open loop system. The washout gain
K is found by classical control method i.e. root locus method
for the best damping ratio as shown in Fig. 1.
Root Locus
Real Axis
I
m
a
g
i
n
a
r
y

A
x
i
s
30 25 20 15 10 5 0 5
60
40
20
0
20
40
60
0.44
0.62
0.84
0.05 0.1 0.16 0.23 0.32 0.44
0.62
0.84
10
20
30
40
50
60
10
20
30
40
50
60
System: untitled1
Gain: 0.161
Pole: 10.5 + 10.5i
Damping: 0.707
Overshoot (%): 4.32
Frequency (rad/sec): 14.9
0.05 0.1 0.16 0.23 0.32
Fig. 1. Root Locus Plot of System with Washout Filter
The time constant is chosen in way that zero in the
washout lter traps a pole near the origin and a cut off point
of 1 rad/sec is achieved.
Washout = K W(s) = K
s
s+1
= 0.161
s
s+1
where, =1 and K = 0.161
The complete plant dynamics are now given by:
G = Servo Plant
lat
Washout
B. Proposed Control Scheme
For the trajectory tracking of UAVs, the conventional inner
outer loop paradigm is used as shown in Fig. 2.
Fig. 2. The Conventional Inner-Outer Loop Framework
In this paper, the inner-outer paradigm is addressed in a
different way. The yaw-rate is also feed to the controller after
passing through the washout lter hence making a MIMO
system as shown in the Fig. 3. It is assumed that the outer
guidance loop is generating a reference roll command which
is tracked by the inner control loop. The inner loop controller
is designed by using H

control technique.
Fig. 3. Proposed Inner Outer Loop Framework
C. H

Control Design Method


The general control conguration of H

control design is
introduced by Doyle [15] as shown in Fig. 4.
Fig. 4. The General Control Conguration of H Framework
Where, P and K are generalized plant and generalized
controller respectively. w and z are exogenous input and
output respectively. The sensor output is represented by v
while the control input is represented by u. The main aim
of this conguration is to synthesize a controller that can
minimize the innity-norm of closed loop transfer function
from exogenous input to output, by using the information from
v and manipulating the control input.
D. Loop Shaping Design Procedure
As the aircraft system is sensitive to model uncertainties and
disturbances, a robust control is required to suppress the dis-
turbances and give robustness against the model uncertainties.
In order to have a robust control, H

Loop shaping design


procedure is the best of all as it offer the following benets:
The -iteration is not required.
Simple mathematical tools are available in literature.
It is applicable to SISO systems as well as MIMO
systems.
Synthesize a controller that has lower order than the plant.
The loop shaping design is a two step procedure: In the
rst step, the open loop plant is shaped with the pre and
post weights to achieve the desire performance as shown in
Fig. 5. In the second step, the shaped plant is set for robust
stabilization with the coprime factor uncertainty.
Fig. 5. The Loop Shaping Design Architecture
The rst step in LSDP is very important. The weights should
be selected very carefully due to the dimensions of the system.
The weight selection is explained in detail in [22][23] and [24].
In this work, our aim is to perfectly track the reference roll
command generated by guidance loop. The weight W
111
is on
aileron input channel and is selected as a pure integrator to
increase the gain at low frequencies and give the maximum
weight to aileron, so that controller can generate such a aileron
input which keep the UAV on the desired trajectory. By high
gains at low frequency, we ensure that there is no steady state
error in tracking and also any disturbance in the channel can
be rejected. While the weight W
122
is on the rudder input
channel to ensure the steady state turning. The weight W
2
is
selected to achieve the desired roll off rate at the cross over
frequency and a high roll off rate at high frequencies. After
some iterations and following the method of [22], the selected
weights are:
W
1
=

z
1
W
111
0
0 z
2
W
122

2
s+9
s
0
0
s+7
s+1

W
2
=

c
1
W
211
0
0 c
2
W
222

s+11
s+15
0
0 4
s+10
s+3.5

Where,
z
1
, z
2
, c
1
and c
2
are the constants, used to adjust the gain
of these weights. The Fig. 6 shows the singular value plot of
shaped and un-shaped plant.
10
3
10
2
10
1
10
0
10
1
10
2
10
3
150
100
50
0
50
100
150


Singular Value Plot
Frequency (rad/sec)
S
i
n
g
u
l
a
r

V
a
l
u
e


(
d
B
)
G
Gs
Fig. 6. Singular Value plot of Shaped Plant
The shaped plant model is then used with the normalized
coprime factor uncertainty to synthesis the robust controller
using MATLAB. To verify the performance criteria, a maxi-
mum stability margin given by is estimated which is dened
as:
=
1
min
Where,
min
is dened as the optimal cost. The value
of >0.25 or <4 ensure that the design is successful. In
this work, after synthesizing the resulting margins for robust
stability are: = 2.6589 and = 0.3761. The Fig. 7 shows
the singular value plot of shaped plant and the plant with
controller.
10
3
10
2
10
1
10
0
10
1
10
2
10
3
150
100
50
0
50
100
150


Singular Value Plot
Frequeny (rad/sec)
S
i
n
g
u
l
a
r

V
a
l
u
e


(
d
B
)
Gs
GK
Fig. 7. Singular Value plot of GK and Gs
IV. SIMULATION RESULTS OF REFERENCE ROLL
TRACKING
The linear aircraft model explained in Section II is analyzed
for the roll dynamics of aircraft. It is assumed here that the
outer guidance loop generates 1
o
reference roll command. In
the inner loop, the deigned robust controller controller track
the reference command accurately as shown in Fig. 8.
0 2 4 6 8 10
0
0.2
0.4
0.6
0.8
1
1.2
1.4
Reference Roll Command Tracking
Time[sec]
R
e
f
e
r
e
n
c
e

R
o
l
l

C
o
m
m
a
n
d
[
d
e
g
]


Reference Roll Command Tracking
Fig. 8. Reference Tracking of 1
o
Step Reference Roll Command
The control effort is also shown in Fig. 9. The simulation
result indicate that the control effort remain in limit while
tracking the desired reference trajectory.
0 2 4 6 8 10
0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
Control Effort
Time[sec]
C
o
n
t
r
o
l

I
n
p
u
t

(
d
e
g
)


Control Energy
Fig. 9. Control Effort
A. With Parameter Uncertainty
To check the robustness and performance of the designed
controller, some parameter variations are introduced in the
dynamic model. The lateral derivatives are the essential pa-
rameters in the ight control system. These derivative may
have error due to unmodeled dynamics and modeling un-
certainties. Out of all the parameters, some parameters are
found more sensitive to variations after many iterations. Hence,
20% parametric uncertainty is added to the to the control
derivatives C
n
and C
y
r
, while 30% parameteric uncertainty
is added to C
la
and C
lr
. The simulation results indicate good
robustness of the designed controller against the parameteric
uncertainties.
0 2 4 6 8 10
0
0.2
0.4
0.6
0.8
1
1.2
1.4
Reference Roll Command Tracking
Time[sec]
R
e
f
e
r
e
n
c
e

R
o
l
l

C
o
m
m
a
n
d

[
d
e
g
]


With Parameter Uncertainty
Without Parameter Uncertainty
Fig. 10. Reference Tracking of 1
o
Step Reference Roll Command in presence
of Parametric Uncertainty
0 2 4 6 8 10
0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
Control Effort
Time[sec]
A
i
l
e
r
o
n

(
d
e
g
)


With Parameter Uncertainty
Without Parameter Uncertainty
Fig. 11. Control Effort in presence of Parametric Uncertainty
V. CONCLUSION
A robust controller using H

Loop Shaping Shaping De-


sign Procedure is addressed in this paper which can reject the
disturbance and deal with the parametric uncertainties. The
lateral dynamic model of scaled YAK-54 is discussed. The
inner outer loop strategy is adopted for the trajectory tracking.
The inner loop controller is designed to track the reference
roll command generated by the outer guidance loop. The
simulation results indicate the strong characteristics of H

loop shaping design procedure and give a strong indication


about the robustness of the designed controller. As a future
work, the model is veried for different kind of uncertainties
and disturbance effects with H

Mixed Sensitivity method or


Linear Matrix Inequalities.
ACKNOWLEDGMENT
The authors would like to thank the Control and Signal
Processing Research Group (CASPR), MAJU.
REFERENCES
[1] G. M. Siouris, Missile Guidance and Control Systems. Springer-verlag,
2004.
[2] R. Samar, S. Ahmed, and F. Aftab, Lateral control with improved
performance for uavs, 17th IFAC Symposium on Automatic Control in
Aerospace, Toulouse, France, 2007.
[3] S. Park, J. Deyst, and J. P. How, Performance and lyapunov stability
of a nonlinear path following guidance method, Journal of guidance,
control and dynamics, vol. 30, No. 6, 2007.
[4] M. Z. Shah, R. Samar, and A. I. Bhatti, Lateral control for uavs using
sliding mode technique, 18th IFAC World Congress Milano (Italy),
vol. 3, Aug-Sep 2011.
[5] R. Rysdyk, Unmanned aerial vehicle path following target observation
in wind, Journal of Guidance, Navigation and Control, vol. 29(5),
pp. 10921100, 2006.
[6] M. Niculescu, Lateral track control law for aerosonde uav, 39th AIAA
Aerospace Science Meeting and Exhibit, Reno, NV, January 2001.
[7] R. Samar, S. Ahmed, and M. Nazar, Lateral guidance and control design
for an unmanned aerial vehicle, 17th IFAC World Congress, Seoul,
Korea, 2008.
[8] J. S. Jang and C. J. Tomlin, Autopilot design for the stanford drag-
ony uav: Validation through hardware-in-the-loop simulation, AIAA
Guidance, Navigation, and Control Conference and Exhibit, Montreal,
Canada, 2001.
[9] S. Franko, Lqr based trajectory control of full envelope, autonomous
helicopter, Proceedings of the World Congress on Engineering, London,
UK, vol. Vol. I, July 2009.
[10] V. G. Nair, M. V. Dileep, and V. I. George, Aircraft yaw control system
using lqr and fuzzy logic controller, International Journal of Computer
Applications, vol. Vol. 45 No.9, May 2012.
[11] G. Zames, Feedback and complexity, special plenary lecture adden-
dum, IEEE Conference on Decision and Control, 1976.
[12] G. Zames, Optimal sensitivity and feedback: weighted seminorms,
approximate inverses, and plant invariant schemes, Proc. Allerton
Conference, 1979.
[13] M. Giacomn-Zarzar, R. Ramirez-Mendoza, P. J. Fleming, I. Grifn, and
A. Molina-Cristbal, Robust h controller design for aircraft lateral
dynamics using multi-objective optimization and genetic algorithms,
Proceedings of the 17th World Congress The International Federation
of Automatic Control, Seoul, Korea, July 2008.
[14] R. Adams and S. Banda, Robust ight control design using dynamic
inversion and structured singular value synthesis, IEEE-Transactions
on Control System Technology, vol. 1, pp. 8092, 1993.
[15] J. Doyle, K. Glover, P. Khargonekar, and B. Francis, State-space
solutions to standard h
2
and h control problems, IEEE Transactions
on Automatic Control, vol. 34(8), pp. 831847, August 1989.
[16] K. Glover and D. McFarlane, Robust stabilization of normalized
coprime factor plant descriptions with h bounded uncertainty, IEEE
Transactions on Automatic Control, vol. 34, 1989.
[17] K. Zhou, J. C. Doyle, and K. Glover, Robust and Optimal Control.
Prentice-Hall, 1996.
[18] S. Skogestad and I. Postlethwaite, Multivariable Feedback Control:
Analysis and Design. John Wiley & Sons, INC., 2 ed., October 2005.
[19] H. I. E. Leong, Development of a 6dof nonlinear simulation model
enhanced with ne tuning procedures, Masters thesis, The University
of Kansas, Lawrence, KS, 2008.
[20] B. L. Stevens and F. L. Lewis, Aircraft Control And Simulation. John
Wiley & Sons, INC., 2 ed., 2003.
[21] P. J. P. How, Aircraft stability and control, MIT Lecture Notes in
Aeronautics and Astronautics, 2004.
[22] R. A. Hyde, The Application of Robust Control to VSTOL Aircraft. PhD
thesis, Girton College, Cambridge, 1991.
[23] A. Lanzon, Simultaneous synthesis of weights and controllers in h
loop-shaping, Proceeding of the 40th IEEE Conference on decision and
Control Orlando, Florida USA, Dec 2001.
[24] A. Lanzon, Weight optimization in h loop shaping, Automatica,
vol. 41, 2005.

Potrebbero piacerti anche