International Journal of Control Science and Engineering 2012, 2(6): 181-188

DOI: 10.5923/j.control.20120206.06

A Self-Tuning Fuzzy Logic Controller for Aircraft Roll

Control System
Omur Akyazi1,* , Mehmet Ali Usta2 , Adem Sefa Akpinar2

1
Surmene Abdullah Kanca VHS, Karadeniz Technical University, Trabzon, Turkey
2
Dept. of Electrical and Electronics Engineering, Karadeniz Technical University, Trabzon, Turkey

Abstract In this paper, an aircraft roll control system based on autopilot operating conditions is modeled and simu lated
using Matlab/Simulink. The modeling phase begins with the derivation of required mathematical model to describe the
lateral d irectional motion of an aircraft. Then, Linear Quadratic Regulator (LQR), Fuzzy Logic Controller (FLC) and
Self-Tuning Fu zzy Logic Controller (STFLC) are applied for controlling the roll angle o f the modeled aircraft system.
Simu lation results of ro ll controllers are presented in time do main and the results obtained with STFLC are co mpared with
the results of FLC and LQR. Finally, the performances of roll control systems are analysed in order to decide which control
method gives better performance with respect to the desired roll angle. According to simu lation results, it is shown that
STFLC deliver better performance than FLC and LQR.
Keywords Aircraft Ro ll Control, Linear Quadratic Regulator (LQR), Fu zzy Logic Controller (FLC) and Self-Tuning
Fuzzy Logic Controller (STFLC)

Generally, an aircraft is controlled by three main surfaces.

1. Introduction These are elevator, rudder and ailerons. Pitch control can be
achieved by changing the lift on either a forward or aft
The development of autopilots closely followed the control surface. If a flap is used, the flapped portion of the
successful development of powered man-carry ing airplane tail surface is called an elevator. Yaw control is achieved by
by the Wright brothers[1]. The first auto matic flight deflecting a flap on the vertical tail called the rudder and
controller in the world is designed by the Sperry brothers in roll control can be achieved by deflecting small flaps
1912. The Sperry brothers developed an autopilot that is located outboard toward the wing tips in a d ifferential
sensitive to the movements of an aircraft. When an aircraft manner[1]. These flaps are called ailerons. Elevator, rudder
deviated fro m a part icular flight route, this autopilot and ailerons are depicted in Figure 1.
adjusted the pitch, roll and heading angles of an aircraft.
Then, in 1914, the Sperry brothers demonstrated this
autopilot at the Paris air-show. To demonstrate the
effectiveness of their design, Lawrence Sperry t rimmed his
airplane for straight and level flight and then engaged the
autopilot[1]. Since then, the fast advancement of high
performance military, co mmercial and general aviation
aircraft design has required the develop ment of many
technologies; these are aerodynamics, structures, materials,
propulsion and flight controls[2]. Currently, the aircraft Figure 1.Aerodynamic controls of an aircraft[1].
design relies heavily on automatic control systems to
monitor and control many of the aircraft subsystems[2]. The two ailerons are typically interconnected and both
Therefore, the development of auto matic control systems ailerons usually move in opposition to each other. The
has played an important role in the growth of civil and ailerons are used to bank the aircraft. The banking creates an
military aviation[1]. Modern aircrafts are much mo re unbalanced side force co mponent of the large wing lift force
complex and includes a variety of automatic control system. which causes the aircraft’s flight path to curve[3]. Thus,
when the pilot applies right push force on the stick, as the
* Corresponding author: aileron on the right wing is deflected upward, the aileron on
[email protected] (Omur Akyazi ) the left wing is deflected downward. As a result of this, the
Published online at http://journal.sapub.org/control
lift on the left wing is increased, wh ile the lift on the right
Copyright © 2012 Scientific & Academic Publishing. All Rights Reserved
182 Omur Akyazi et al.: A Self-Tuning Fuzzy Logic Controller for Aircraft Roll Control System

wing is decreased. So, the aircraft performs a rolling motion rrepresent the angular rates components of roll, pitch and
to the right as viewed fro m the rear of the aircraft. yaw axis and the term u, v and w represent the velocity
The rolling motion of an aircraft is controlled by adjusting components of roll, p itch and yaw axis.
the roll angle. In this study, an autopilot is designed to Referring to Figure 2 and Figure 3, the rig id body
control the roll angle of an aircraft. In aircraft modeling equations of motion are obtained fro m Newton’s second
phase, the aerodynamic forces (lift and drug) as well as the law, see[1]. But, a few assumption and approximation need
aircraft’s inert ia are taken into account[4]. The actual model to be considered before obtaining the equations of motion.
is a third order nonlinear system, which is linearized about Assume that the aircraft is in steady-cruise at constant
the operating point[4]. A modern linear quadrature regulator altitude and velocity, thus, the thrust and drag cancel out
(LQR) and intelligent controllers (FLC and STFLC) are and the lift and weight balance out each other. Also, assume
developed for the roll control of the modeled aircraft system. that change in pitch angle does not change the speed of an
Performances of these controllers are analysed with respect aircraft under any circumstance[4]. Under these
to the desired roll angle. Co mparison of these control theory assumptions, the lateral directional motion of an aircraft is
is presented and discussed in terms of performance analysis. well described by the following kinematic and dynamic
differential equations.
•

2. Modeling of A Roll Control System Y + mgCθ Sθ = m(v + ru − pw) (1)

• •
The equations governing the motion of an aircraft are L = I x p − I xz r + qr ( I z − I y ) − I xz pq (2)
very complicated as a set of six nonlinear coupled • •
differential equations. However, under certain assumptions, N = − I xz p + I z r + pq( I y − I x ) + I xz qr (3)
they can be decoupled and linearized into the longitudinal
and lateral equations. Ro ll control is a lateral problem and Equation (1), (2) and (3) are nonlinear and they can be
this work is developed to control the roll angle of an aircraft linearized by using small-disturbance theory. According to
for ro ll control in order to stabilize the system when an small-d isturbance theory, all the variab les in the equation
aircraft performs the ro lling motion. The ro ll control system (1), (2) and (3) are replaced by a reference value plus a
is shown in Figure 2. perturbation or disturbance, as given in equation (4).
u = u 0 + ∆u v = v0 + ∆v w = w0 + ∆w
p = p0 + ∆p q = q0 + ∆q r = r0 + ∆r
δa
Y = Y0 + ∆Y L = L0 + ∆L M = M 0 + ∆M
δ = δ 0 + ∆δ (4)
For convenience, the reference flight condition is
assumed to be symmetric and the propulsive forces are
assumed to remain constant. This implies that,
v0 = p0 = q0 = r0 = φ0 = ψ 0 = 0 (5)
Figure 2.Description of roll control system[1]. After linearization the follo wing equations are obtained,
see[1].
d 
 − Yv ∆v − Y p ∆p + (u 0 − Yr )∆r − ( g cosθ 0 )∆φ = Yδr ∆δ r
 dt 
(6)
d  I d 
− Lv ∆v +  − L p ∆p −  xz + Lr ∆r = Lδa ∆δ a + Lδr ∆δ r
 dt   I x dt 
(7)
I d  d 
− N v ∆v −  xz + N p ∆p +  − N r ∆r = N δa ∆δ a + N δr ∆δ r
Figure 3.Definition of forces, moments and velocity components in a body
 I z dt   dt 
fixed frame[1]. (8)
The lateral d irect ional equations of motion consist of the
In this figure, Yb and Zb represent the aerodynamics force side force, ro lling mo ment and yawing mo ment equations
components,ϕ and δa represent the orientation of aircraft of motion. It is sometimes convenient to use the sideslip
(roll angle) in the earth-axis system and aileron deflection
angle Δβinstead of the side velocityΔv. These two quantities
angle respectively. The forces, mo ments and velocity
are related to each other in the follo wing way;
components in the body fixed frame of an aircraft system
∆v ∆v (9)
are shown in Figure 3 where L, M and N represent the ∆β ≈ tan −1 =
u0 u0
aerodynamic mo ment components, the term p, q and
 International Journal of Control Science and Engineering 2012, 2(6): 181-188 183

Using this relationship and if the product of inertia Ixz =0, system. So, the rudder deflection g iven in equation (10) is
the lateral equations of motion can be rearranged and not used.
reduced into the state space form in the following manner.  •  − 0.254 0 − 1 0.183 ∆β   0 
Yp ∆ β•   (11)
 •   Yβ  Y  g cos θ 0   Yδr   ∆ p  − 15.969 − 8.395 2.19 0   ∆p  − 28.916
∆ β•   u − 1 − r   ∆β   0  •= [∆δ a ]
u0  u 0  u   ∆p   u0  (10) +
∆ p   0  ∆δ  ∆ r   4.549 − 0.349 − 0.76 0   ∆r   − 0.224 
Lδr   a 
0

 •  =  Lβ Lp Lr 0    +  Lδa
 ∆r  ∆δ  •   0 1 0
  
0   ∆φ   0 

 ∆ r  N Np Nr 0     N δa N δr   r   ∆φ 
 •  β ∆φ  
 ∆ φ   0 1 0 0     0 0  Transfer function fro m aileron deflect ion angle to roll
For this system, the input will be the aileron deflection angle is given by the follo wing equation.
angle and the output will be the roll angle. In this study, the ∆φ ( s) − 28.92s 2 − 29.81s − 140.8 (12)
= 4
data from General Aviation Airplane: NA VIONa [1] is used ∆δ a ( s) s + 9.409s 3 + 14.02s 2 + 48.5s + 0.3979
in system analysis and modeling. The lateral direct ional
derivatives stability parameters for this airplane are given
Table I. 3. Design Process of The Proposed
Table 1.The lateral directional derivatives stability parameters
Controller
The Dynamic Pressure Ԛ and t Fuzzy Logic Controller (FLC), Self-Tun ing Fuzzy Logic
Controller (STFLC) and Linear Quadratic Regulator (LQR)
−
Q = 36.8 Ib / ft 2 QS c = 38596 ft ⋅ Ib are proposed for the roll control system and in this section;
General these controllers are described in detail.
Aviation QS = 6771 Ib c 2u 0 = 0.016 s
Airplane: 3.1. Linear Quadratic Regul ator (LQR)
NAVIONa Components
Yawing Rolling During last decades, a new approach to control system
Y-Force
Moment Moment design has evolved. This approach is co mmonly called
Derivatives
Derivatives Derivatives modern control theory. Linear Quadratic Regulator (LQR) is
Pitching
Yv=-0.254 Nv=0.025 Lv=-0.091 a method in modern control theory and it is an alternative and
Velocities
Side Slip
very powerful method for flight control system designing.
Yβ =-44.665 Nβ =4.549 Lβ =-15.969 The method is based on the manipulation of the equations of
Angle
motion in state space form and makes full use of the
Rolling Rate Yp =0 Np =-0.349 Lp =-8.395
appropriate computational tools in the analytical process[5].
Yawing Rate Yr=0 Nr=-0.76 Lr=2.19 LQR control system for the lateral direct ional control of an
aircraft is shown in Figure 4.
Rudder
Deflection
Yδr=12.433 Nδr=-4.613 Lδr=23.09 The state and output matrix equations describing the
Aileron lateral d irectional equations of motion can be written as the
Yδa=0 Nδa=-0.224 Lδa=-28.916
Deflection following equation.
•
Before obtaining transfer function, let’s plug in nu merical x(t ) = Ax(t ) + Bu (t )
values given Table I by using equation (10). This work y (t ) = Cx (t ) + Du (t ) (13)
presents the roll control schemes for roll angle of an aircraft
Δβ
Δδa + u(t) Lateral Aircraft Δp
N
Dynamics Δr
− Δφ

Kβ
+
+
Kp

+ Kr
+

Kφ

Feedback Gain Matrix

Figure 4. Full-state feedback controller with reference input for the roll control system
184 Omur Akyazi et al.: A Self-Tuning Fuzzy Logic Controller for Aircraft Roll Control System

And that all of the four states x are available for the mechanis m and defu zzification. A block diagram o f a fu zzy
controller. The feedback gain is a mat rix K of the optimal control system is shown in Figure 6.
control vector. In Figure 6, the values of error (e(k)) and its change (Δe(k))
K = [K β K p K r Kϕ ] occurring during the operation of the system form the crisp
u (t ) = − K ⋅ x(t ) + ∆δ a ⋅ N (14) inputs of the system. These two inputs defined as in (16) and
(17).
So as to minimize the perfo rmance index,
∞
e( k ) = r ( k ) − y ( k ) (16)
( )
J = ∫ x T Qx + u T Ru ⋅dt (15) ∆e(k ) = e(k ) − e(k − 1) (17)
0 r(k), y(k) and k are exp ressed as the reference input, the
Where Q is state-cost matrix and R is performance index actual output of the system and the sampling step
matrix. For this study, R=1 and Q=CT xC where C is the respectively. These crisp inputs e(k) and Δe(k) are converted
matrix fro m state equation (13) and CT is the matrix to fuzzy membership value on the fu zzy subsets. There are
transpose of C. For designing LQR controller, the value of three main fu zzy subsets defined as negative (N), zero (Z)
the feedback gain matrix, K, must be determined. The and positive (P). Depending on these subsets the number of
following block is shown how to determine the values of K. rules can be derived.
These fuzzy membership values are used in the rule base
A Modern Control Systems in order to execute the related rules so that an output can be
B Design Package generated. A rule base consists of a data table which includes
MATLAB K informat ion related to the system. As an examp le, if a fu zzy
Q
logic controller with nine rules is desired to realize, these
R [K]=lqr(A,B,Q,R)
rules can be defined in Tab le II.
Figure 5.Determine the values of matrix K. Table 2.Rules for the fuzzy logic controller

K=[0.5284, -0.5349, -0.0917 -8.6567] values are obtained e

N Z P
by using method is depicted as Figure 5 as the weighting ∆e
factor equals 75. To obtain the desired output in other words N N N Z
to reduce steady-state error, one must use a feed-forward
scaling factor called N. Because, the full-state feedback Z N Z P
system does not compare the output to the reference, it
compares all states multip lied by the feedback gain matrix to P Z P P
the reference. These are shown in Figure 4. So, the reference
An inference mechanism emu lates the expert’s decision
must be scaled by scaling factor N. The scaling factor N is
making in interpret ing and applying knowledge about how
obtained from Mat lab function that is a designer-defined
best to control the plant. Adefuzzification interface converts
function in m-file code. In this case, N=-8.6603 is
the conclusions of the inference mechanis m into the crisp
determined.
inputs for the process. A general overlooked v iew of the FLC
3.2. Fuzzy Logic Controller (FLC) is given in Figure 7 where the processes fro m inputs e and Δe
to output Δu are shown. The input data blocks to represent
In most research literature, a fu zzy controller system is fuzzy membership functions for the error e, error change ∆e
commonly defined as a system that emu lates a human expert. and the controlled output change ∆u are shown in Figure 7.
In this case, the knowledge of the human operator wou ld be The user is able to edit and change the parameters of the
put in the form of a set of fu zzy linguistic rules. These rules membership functions on this stage without going into the
would produce an appro ximate decision in the same manner detail of the FLC.Figure 7T he input and output of the FLC.
a human would do. The fu zzy controller is co mposed of four
elements. These are fuzzificat ion, rule base, inference

e(k) Rule Base

Defuzzification
Fuzzification

Δu(k)

Δe(k) Inference

Figure 6.The basic structure of fuzzy logic based controller

 International Journal of Control Science and Engineering 2012, 2(6): 181-188 185

Figure 7.The input and output of the FLC

Figure 8.The structure of Self-T uning FLC

3.3. Self-Tuning Fuzzy Logic Controller (STFLC) difference between system output and reference signal.
STFLC is developed to improve the controller Structure of self-tuning FLC is shown in Figure 8.
performance by tuning the range values of fuzzy subsets of
error and its change used in direct fuzzy controller. The 4. Simulation and Results
symbols of on-line changing fuzzy gains are G1 and G2
respectively for error (e) and change of error (Δe). In order to An aircraft roll control system is simu lated using LQR,
adjust the gain parameters G1 and G2, two d ifferent fu zzy FLC and Self-Tuning FLC in order to present and discuss
logic controllers are used[8]. The inputs of gain-adjusting simu lation results. Simu link model of the system with these
FLCs are system output and error signal which is the controllers is shown in Figure 9.
186 Omur Akyazi et al.: A Self-Tuning Fuzzy Logic Controller for Aircraft Roll Control System

Figure 9.Simulink model of the system with three proposed controllers

For all simu lations, the reference value is selected as 0.15 radian wh ich is equal to 8.625 degrees. Firstly, the various
membership functions of FLC are examined and the best membership function for this system is determined. Figure 10
shows the comparison of membership functions.

Figure 10.The system response with various membership functions

It is observed that the triangle membership function gives examined to understand which fu zzy rule table gives better
the best response as compared to others. After determining response. FLC with nine rules gives better response than
the membership function, the various rule tables of FLC is others. It is shown in Figure 11.
 International Journal of Control Science and Engineering 2012, 2(6): 181-188 187

Figure 11. The system response with various fuzzy rule tables

Obtained fu zzy ru le table and membership function which g ives the best response for this system is used with both FLC
and Self-Tuning FLC. Then the system responses of LQR, FLC and Self-Tuning FLC are p lotted on the same graph for a
better comparison. The system responses with these controllers are shown in Figure 12.

Figure 12.The performance comparison of proposed controllers

It is observed that STFL controller gives faster response system. LQR, FLC and Self-Tuning FLC are successfully
as compared to FLC and LQR in terms of rising time. But, designed and presented for this system. As a result, among
the results clearly demonstrate that LQR controller is these controllers, STFLC gives the best performance in
occurred overshoot more than FLC and STFLC. terms of rising time, settling time, steady-state error and
percent overshoot. According to the results from
simu lation and analysis, STFLC has good and acceptable
5. Conclusions performances.
In this paper, the model of an aircraft roll control system
that is helpful in developing the control strategy for an actual
aircraft system was designed for Matlab/Simulink
environment and control methods were proposed for this REFERENCES
188 Omur Akyazi et al.: A Self-Tuning Fuzzy Logic Controller for Aircraft Roll Control System

[1] R. C. Nelson, 1998, Flight Stability and Automatic Control, [5] M ıchael V. Cook, 2007, Flight Dynamics Prıncıples, Elsevıer,
M cGraw Hill, Second Edition. Second Edition.

[2] Lucio R. Riberio and Neusa M aria F. Oliveira, “UAV [6] I. H. Altas and A. M . Sharaf, “A Generalized Direct Approach
Autopilot Controllers Test Platform Using M atlab/Simulink for Designing Fuzzy Logic Controllers in M atlab/Simulink
and X-Plane”, 40th ASEE/IEEE Frontiers in Education GUI Environment”, Accepted for publication in International
Conference, October 27-30, 2010, Washington, DC. Journal of Information Technology and Intelligent
Computing, Int. J. IT&IC no.4 vol.1.
[3] M . A. Usta, O. Akyazi and A. S. Akpinar, “Aircraft Roll
Control System Using LQR and Fuzzy Logic Controller”, [7] Joao P. Hespanha, April 1,2007, Umdergraduate Lecture
International Symposium on INnovations in Intelligent Notes on LQG/LQR Controller Design.
SysTems and Applications (INISTA 2011), pp. 223-227, June
15-18, 2011, Istanbul, Turkey. [8] S. Ramesh and A. Krishnan, “A Self –Tuning Fuzzy Logic
Controller for a Frequency Stabilization in a Parallel AC –
[4] Nurbaiti Wahid and M ohdFua’adRahmat, “Pitch Control DC Two Area Interconnected Power System”, European
System Using LQR and Fuzzy Controller”, 2010 IEEE Journal of Scientific Research ISSN 1450-216X Vol.51 No.1
Symposium on Industrial Electronics and Applications (2011), pp.6-17, © Euro Journals Publishing, Inc. 2011
(ISIEA 2010), October 3-5, 2010, Penang, M alaysia