Welcome to Scribd. Sign in or start your free trial to enjoy unlimited e-books, audiobooks & documents.Find out more
Download
Standard view
Full view
of .
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
AIRCRAFT LANDING CONTROL IN WIND SHEAR CONDITION

AIRCRAFT LANDING CONTROL IN WIND SHEAR CONDITION

Ratings: (0)|Views: 34|Likes:
Published by Ruchit Pathak
AIRCRAFT LANDING CONTROL IN WIND SHEAR CONDITION
AIRCRAFT LANDING CONTROL IN WIND SHEAR CONDITION

More info:

Published by: Ruchit Pathak on Feb 03, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

08/10/2013

pdf

text

original

 
 
AIRCRAFT LANDING CONTROL IN WIND SHEAR CONDITION
CHIA-LIN LEE, JIH-GAU JUANG
 
Department of Communications, Navigation and Control Engineering, National Taiwan Ocean University, Keelung, TaiwanE-MAIL: jgjuang@mail.ntou.edu.tw
Abstract:
Most aircraft accidents occurred during final approach orlanding. This study proposes cerebellar model articulationcontroller (CMAC) to improve the performance of automaticlanding system (ALS). The atmospheric disturbances affect notonly flying qualities of an airplane but also flight safety. If theflight conditions are beyond the preset envelope, the automaticlanding system (ALS) is disabled and the pilot takes over. Aninexperienced pilot may not be able to guide the aircraft to asafe landing at the airport when wind shear is encountered. Anadaptive type-2 fuzzy CMAC (FCMAC) is applied to PIDcontrol to construct intelligent landing system which can guidethe aircraft to a safe landing in severe wind shear environment.
Keywords:
Automatic landing system; PID control; Fuzzy CMAC;Wind shear
1.
 
Introduction
Atmospheric disturbance can shake any airplane, nomatter how big it is. One of the most dangerous disturbanceis wind shear, a large change in wind speed or direction over a short distance. Wind shear is caused by microbursts, whichare winds that blast down from showers or thunderstorms andhave caused several airline crashes over the years. Windshear occurs at all altitudes from the ground to the top of theatmosphere and it can be horizontal or vertical. Weather forecasts try to alert pilots to turbulence, but meteorologistsoften can give only very general forecasts. Pinning downexact locations of the worst turbulence is difficult. On March1, 2008, at Hamburg airport, a Lufthansa Airbus A320 tried toland in crosswind conditions which exceeded the limit for theaircraft and made the left wing touch ground. The pilots then performed a go around and successfully saved the aircraftfrom crashing. According to a survey of the NationalTransportation Safety Board (NTSB) [1], 22.6 percent of aircraft accidents in the years of 1989 to 1999 were weather related. Most aircraft accidents occurred in final approach or landing. Another NTSB report, between 1994 and 2003, therewere 19562 aircraft accidents. Weather was a contributingfactor in 4159 of these accidents and involved 4167 aircraft.Of the 4159 weather-related accidents, 2726 were due towind conditions. In addition, a single accident may involvemultiple weather conditions. According to the statistics of Flight International 10-16, January 2006 issue [2], there were23 accidents/incidents affected by weather, causing total 324(34 crew and 290 passengers) fatalities. The average accidentfatality caused by weather is 14 people. It was apparent thatmost of cases were in the landing phase.Since 1965, most aircraft have had ALS installed.According to the Federal Aviation Administration (FAA)regulations [3], environmental conditions considered thedetermination of dispersion limits as being: headwinds up to25 knots, tailwinds up to 10 knots, crosswinds up to 15 knots,moderate turbulence, and wind shear of 8 knots per 100 feetfrom 200 feet to touchdown. If the flight conditions are beyond the preset envelope, the ALS is disabled and the pilottakes over. An inexperienced pilot may not be able to guidethe aircraft to a safe landing at the airport. The goal of this paper is to show that the proposed intelligent ALS can relievehuman operators and guide the aircraft to a safe landing in asevere turbulence environment. Most of the improvements inthe ALS system have been on the guidance instruments, suchas GNSS Integrity Beacons, Global Positioning System,Microwave Landing System, and Automatic Land PositionSensor [4]-[7]. By using improvement calculation methodsand high accuracy instruments, these systems provide moreaccurate flight data to the ALS to make the landing smoother.However, these researches do not include weather factorssuch as wind turbulences. Recently, intelligent concepts suchas neural networks, fuzzy system, genetic algorithm, andhybrid systems have applied to flight control to increase theflight controller’s adaptively to different environments[8]-[11]. This paper proposes an intelligent aircraft automaticlanding control that uses type-2 FCMAC [12]-[14] toimprove the performance of conventional ALS. Adaptivelearning rates are obtained by the use of the Lyapunovstability theory. The convergence of the proposed controlscheme is guaranteed.
2.
 
System description
At the aircraft landing phase, the pilot descends from the
11802011 IEEE978-1-4577-0308-9/11/$26.00 © 
Proceedings of the 2011 International Conference on Machine Learning and Cybernetics, Guilin, 10-13 July, 2011
 
 
cruise altitude to an altitude of approximately 1200 feetabove the ground. The pilot then positions the aircraft so thatthe aircraft is on a heading towards the runway centerline.When the aircraft approaches the outer airport marker, whichis about 4 nautical miles from the runway, the glide pathsignal is intercepted, as shown in Fig. 1. As the airplanedescends along the glide path, its pitch, attitude, and speedmust be controlled. The descent rate is about 10 ft/sec and the pitch angle is between -5 to +5 degrees. Finally, as theairplane descends 20 to 70 feet above the ground, the glide path control system is disengaged and a flare maneuver isexecuted. The vertical descent rate is decreased to 2ft/sec sothat the landing gear may be able to dissipate the energy of the impact at landing. The pitch angle of the airplane is thenadjusted, between 0 to 5 degrees for most aircraft, whichallows a soft touchdown on the runway surface. A simplifiedmodel of a commercial aircraft that moves only in thelongitudinal and vertical plane is used in the simulations for implementation ease
[10]
.
0 ft50 fttouchdown1200 ftglide pathflare pathRunway positionAltitude
Figure 1. Glide path and flare path
There are several wind shear models [15]. The most usedone is given as follow()
 x
W k A x
=

()()*
h
hW k B xh
=

(1)The distribution of the horizontal wind versus the horizontaldistance is given by
1

a x
 ()
 A x
=
12
x ab a
+
 
b xa
<<
(2)
1

b x
>
and the distribution of the vertical wind versus the horizontaldistance is given by
 0

a x
 
()
 B x
=
 x ac a

c xa

3)

b xb c

b xc
0

b x
where
300
a
=

4300
b
=
,2
a bc
+=
,
*
1000
h
=
.
 x
is thehorizontal position (ft) relative to ground of the aircraft.
a, b,
and
c
are various horizontal distances (ft) measured from theinitial position.
h
is the vertical altitude (ft) of the aircraft.
h*
 is the reference altitude (ft).
is the wind shear intensity(ft/sec). Fig. 2 shows a wind shear profile with a wind speedof 30 ft/sec.
 
0510152025303540-30-20-100102030Time (sec.)
        f       t        /      s      e      c
Wind Shear velocity components: Horizontal (Solid) & Vertical (Dashed)
 
Figure 2. Wind shear profile
3.
 
Controller design
PID controllers are the most used controller inengineering industry. One of the applications is the aircraftlanding system. Controller inputs consist of altitude andaltitude rate commands along with aircraft altitude andaltitude rate. The pitch command
c
θ 
is obtained from the PIDcontroller. Then, the pitch autopilot is controlled by pitchcommand. In order to enable aircraft to land more steadywhen an aircraft arrives to the flare path, a constant pitchangle will be added to the controller. In general, the PIDcontroller is simple and effective but there are somedrawbacks such as apparent overshoot and sensitive toexternal noise and disturbance. When wind disturbance isencountered the PID controller may not be able to guide theaircraft to land safely. With CMAC compensator the proposed controller can overcome these disadvantages. Ituses a traditional PID controller to stabilize the system andtrain the CMAC to provide precise control. The gains of PIDcontroller are adjusted based on experiences, what it providesare tolerable solutions, not desired solutions. The CMAC caneffectively meliorate these conditions. The overall controlscheme is described in Fig. 3, in which the control signal
isthe sum of the PID controller output and the type-2 fuzzy
1181
Proceedings of the 2011 International Conference on Machine Learning and Cybernetics, Guilin, 10-13 July, 2011
 
 
CMAC output. The inputs for the CMAC and PID controller are: altitude, altitude command, altitude rate, and altitude ratecommand. The PID controller provides tolerable solutions. Ineach time step
, the CMAC involves a recall process and alearning process. In the recall process, it uses the desiredsystem output of the next time step and the actual systemoutput as the address to generate the control signal
CMAC 
. Inthe learning process, the control signal of the pitch autopilot,
, is treated as a desired output. It is used to modify theweights of CMAC stored at location which is addressed bythe actual system output and the system output of the nexttime step.
 
The output of the CMAC is the compensation for  pitch command. When the wind shear is too strong, the ALScan not control the aircraft to land safely. Here we use atype-2 fuzzy CMAC control scheme to improve the ability of wind shear resistance of the ALS. CMAC is a type of artificial neural network proposed in the literatures
[16]
. TheCMAC represents one kind of associative memory technique.In the addressing technique, each input space (state variable)is quantized and the output space is divided into discretestates. A quantized input vector specifies a discrete state andis used to generate addresses for retrieving information frommemory for this state.
Figure 3. The type 2 CMAC control scheme
The type-2 fuzzy theorem is utilized into CMACstructure in order to promote more accurate resolution thanconventional fuzzy CMAC. The mapping procedure of type-2fuzzy CMAC is similar as type-1 fuzzy CMAC. Each phaseof mapping is described as follows. The
 X 
is an
n
-dimensional input space. Type-2 fuzzy CMAC uses theinterval type-2 fuzzification method of the fuzzy theorem asits addressing scheme. After the input vector to the intervaltype-2 fuzzy set is being fuzzified, the input state values aretransformed to upper firing strength and lower firing strength,which is based on corresponding interval type-2 membershipfunctions. We choose the product inference method as thet-norm operator. The
 j
th
rule’s upper firing strength
 j
c
andlower firing strength firing strength
 j
c
in type-2 fuzzyCMAC could be computed as:)()(......*)(*)()(
121
21
i jnin j j j  j
 xc xc xc xc xc
in
Π
=
==
(4))()(......*)(*)()(
121
21
i jnin j j j j
 xc xc xc xc xc
in
Π
=
==
(5)The type-reduced set of the type-2 fuzzy CMAC using thecenter of sets type reduction :
¦¦³ ³ ³ ³ 
==
==
n j j jn j jcccwww www ccc
wcc y y y
 N  N  N 
11],[],[],[],[ cos
111111
............],[ (6)
 
It is an interval type-1 set determined by its left and right end points
 y
and
 y
, which can be written as follows [13]:
¦¦¦¦¦¦
+==+====
++==
 N  R j j R j j N  R j j j R j j j N  j j N  j j j
ccwcwc cwc y
111111
(7)
¦¦¦¦¦¦
+==+====
++==
 N  L j j L j j N  L j j j L j j j N  j j N  j j j
ccwcwc cwc y
111111
(8)
w
and
w
are the corresponding weights of 
c
and
c
,respectively. L and R can be obtained as follows:
Step 1.
Assume that the pre-computed
 j
w
are arranged inascending order, i.e.,
 N 
www
...
21
 
Step 2.
Compute
 y
 by initially setting 2/)(
j j j
ccc
+=
 for 
 N  j
,....2,1
=
and let
y y
=
 
Step 3.
Find)11(
 R R
such that
1
+
R R
w yw
 
Step 4.
Compute
 y
with
 j j
cc
=
for 
 R j
and
 j j
cc
=
 for 
 R j
>
and let
y y
=
 
Step 5.
If 
y y
then go to step 6. If 
y y
=
then stopand set
y y
 
Step 6.
Set
y y
=
and return to Step 3.The procedure for computing
 y
is very similar to theone just given for 
 y
. In Step 3, find )11(
 L L
suchthat
1
+
L L
w yw
. Additionally, in Step 2 compute
 y
 initially setting 2/)(
j j j
ccc
+=
for 
 N  j
,....2,1
=
and inStep 4 compute
 y
with
 j j
cc
=
for 
 L j
and
 j j
cc
=
 for 
L j
>
. The defuzzified output is simply the average of 
 y
 
1182
Proceedings of the 2011 International Conference on Machine Learning and Cybernetics, Guilin, 10-13 July, 2011

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->