view full paper

24TH INTERNATIONAL CONGRESS OF THE AERONAUTICAL SCIENCES
AUTOMATIC TAKE-OFF AND LANDING SYSTEM
FOR FIXED WING SMALL AIRCRAFT
Takashi Nagayama*, Naoshi Akamine*, Akihiro Yamane*
*Fuji Heavy Industries Ltd , Aerospace Company
Keywords: Unmanned Aerial Vehicle, Automatic Flight Control, Take-off & Landing
Abstract
Fuji Heavy Industries Ltd., Aerospace Company
has developed an experimental aircraft for
automatic takeoff and landing. An automatic takeoff and landing system, based on flight control
technologies from various UAVs, was added on to
a powered glider, and we successfully
substantiated by several times of flight tests. From
this results, we confirmed that this system can be
applied to UAVs, and also to manned aircrafts for
future automatic transportation system. This
report describes the outline of our experiments of
automatic take-off and landing.
1 General Introduction
In recent years, UAV(Unmanned Aerial
Vehicle) applications are spreading to various
fields such as agricultural spraying, surveillance
and observation, and more complicate missions.
To simplify and to rationalize its operation, it is
wanted for UAVs to have the ability of take-off
and landing on runway, like conventional
aircraft. Fuji Heavy Industries Ltd.(FHI) have
been promoting the research and development
of automatic take-off and landing system for
various UAVs. At the High Speed Flight
Demonstration (HSFD), an experimental
program of JAPAN Aerospace Exploration
Agency (JAXA), we established the automatic
take-off and landing technology of a high speed
fixed wing UAV with a jet engine. And, for
rotary winged UAVs, we succeeded automatic
take-off and landing experiments with RPH 2,
an unmanned helicopter for civil uses such as
observation and agricultural spraying.
Based on such autonomous flight control
technologies for various UAVs, we started
FABOT (Fuji Aerial roBOT) experiment program,
which aimed to demonstrate automatic take off
and landing by a low-speed, low wing loading and
small-size aircraft. In this report describes the
overview of the FABOT system.
2 System Components
2.1 Vehicle Specifications
For experiment system, we decided to add an
Automatic Flight Control System (AFCS) on an
existing aircraft to reduce the costs and term. For
the platform, we selected a powered glider HK36TTC115 Super Dimona, production of Diamond
Aircraft, which has low wing loading, and
ordinary tricycle system.
The aircraft outlook is shown in Fig.1 and the
vehicle's major specifications shown in Table 1.
Fig.1 Outlook of the Experimental Aircraft
1
TAKASHI NAGAYAMA, NAOSHI AKAMINE, AKIHIRO YAMANE
GPS receiver, a laptop PC, and a wireless modem.
Thus, the on-board system is truly compact to be
boarded on a small aircraft, while the ground
equipment is portable enough to realize easy
operation.
The flight computer processes the input from the
navigation system, according to fully automatic
guidance and flight control algorithm from take
off to landing, then outputs control commands to
actuator controller, which drives the control
surfaces and throttle. Actuators are installed under
the co-pilot seat.(see Fig.4)
In this experiment, the AFCS controls only
elevator, aileron, rudder, throttle. Other effecters
such as foot brake, spoiler, and other isolated
operation such as starting engine, are pilot
controlled, for the purpose of simplifying the
system.
For safety, the system has a disengage device in
the linkage, which enables to switch the authority
from autopilot to human pilot in case of
unexpected trouble. The action to disengage is
only to grip the lever attached to the control stick.
It enables quick disengage in critical situations,
especially in low altitude.
Table 1 Vehicle Specifications
[m]
7.28
Length
[m]
16.33
Span
[m]
1.78
Height
[kg]
770
Maximum Weight
[m2]
15.30
Wing Area
73.5
[kw]
Maximum Power
2.2 Automatic Flight Control System(AFCS)
Fig.2 shows the block diagram of the Automatic
Flight Control System (AFCS). The remarkable
characteristics of this system are that ; It has
sufficiently precise, but low-cost Navigation
system integrated from various sensors, such as
GPS, IMU, air data sensor, etc. Each sensor data
is processed in the flight computer and provided to
guidance law as hybrid navigation data.
As the position sensor, we adopted an RTK-GPS
system, which makes it possible to detect position
and ground speed with enough accuracy for takeoff and landing. The differential information is
send to the vehicle from the portable ground
equipment shown in Fig.3, which consists of a
Trigger
SW
GPS Anttena
RTK-GPS
(Onboard)
IMU
Air Data
Sensor
Magnet
Meter
ON-BOARD SYSETM
Lattitude,
Longitude,
Groundspeed
Elevator
Actuator
Actuator
commands
Actuator
Controller
Attitude,
Rate
Pressure
altitude,
Airspeed
Flight
Computer
Actuator
feedback
Ailron
Actuator
Rudder
Actuator
Elevator
Disengege Device
Wireless
Modem
Disengage
Lever
Potentiometer
Ailron
Potentiometer
Rudder
Potentiometer
Throttle
Actuator
Magnetic
heading
Engine
Potentiometer
Control Data
Surface Data
GPS Differential
Data
Sensor Data
Data
Logger
GPS Anttena
Wireless
Modem
RTK-GPS
(Ground)
Monitor
PC
GROUND SYSETM
Fig.2 System Block Diagram of AFCS
2
AUTOMATIC TAKE-OFF AND LANDING SYSTEM FOR FIXED WING SMALL AIRCRAFT
3 Features of Automatic Take-off and Landing
Furthermore, we designed it possible to disengage
without the lever action, by pushing or pulling the
control stick with a certain strength.
Fig.5 shows the outlook of the control stick
equipped with the disengage lever, and the engage
indicators on the consol panel.
The AFCS makes the aircraft fly automatically
according to pre-programmed flight plans. In this
experiment, we set the waypoints as a track
pattern around the airport, shown in Fig.6.
3.1 Automatic Take-off
Uplink Antenna
When the pilot turns on the trigger switch on the
console panel, the AFCS starts the take-off phase,
shown in Fig.7.
After starting take-off phase, it controls the
throttle to hold maximum power, and each
effecter moves to keep the taxing attitude of the
vehicle. When reached to the rotation speed,
elevator control pulls up to make airborne, which
Wireless
Modem
GPS Receiver
GPS
Antenna
Fig.3 Portable Ground Equipment
Engage Indicators (LEDs)
Disengage Lever
Navigation Sensors
& Flight Computer
Actuators
Fig.4 Installation of AFCS
Fig.5 Control Stick & Console Panel
AIMING AIR SPEED：55kt
START POINT
of
GLIDE SLOPE
Cruise Phase
Way-Point # 3
400m
CRUISE-PHASE
END POINT
of
INITIAL CLIMB
40
E:2
G
mA
L
D
ITU
LTSTART POINT
A
NG
of
MI
FINAL FLARE
AI
ALT:7mAGL
RUNWAY
400m
Takese
Take-off Pha
Phase
19
00
m
400m
T
Way-Point # 1
400m
3.5deg
N
LA
E
AS
PH
G
N
DI
210
240m
Way-Point # 2
Way-Point # 4
E
AS
PH
FF
m
O
0
E290
AK
0m
Landing Phase
Fig.6 Experiment Flight Pattern
3
TAKASHI NAGAYAMA, NAOSHI AKAMINE, AKIHIRO YAMANE
phase, shown in Fig.8.
In the first step of the landing phase, it is
programmed to keep 3.5 degree of flight path
angle and 60 kts of approach speed. When it pass
certain flare altitude, it starts flare mode. Elevator
starts pull up to the flare pitch angle, and throttle
decelerate to the idle speed, then touch down.
After touch down, it transits to taxi mode by
reduction of the airspeed
Course tracing is achieved by aileron control in
the approach, and it switches to the rudder control
after starting flare.
is the rotation mode, and next, passing certain
altitude, it transits to climb mode.
In the climb mode, the throttle control maintains
max power, while the elevator keeps the target
airspeed. The take-off phase is completed after
reaching the first waypoint, and then the AFCS
starts the cruise phase.
Before rotation, course tracing is achieved by
rudder control so that the aileron control is used to
hold the bank angle horizontally. After airborne,
course tracing is switched to the aileron control.
3.2 Automatic Landing
When completed all the cruise waypoints
scheduled in the flight plan, it starts the landing
3. CLIMB
1. TAKE-OFF ROLL
2. ROTATION
ELEVATOR ： Pull Up until
ELEVATOR ： Hold the
the Rotation Pitch Angle
Take-Off Pitch Angle
AILERON
：
Keep
the Attitule Horizontal
AILERON ： Keep the Attitule Horizontal
RUDDER　： Keep the course
RUDDER　： Keep the Course
THROTTLE ： Hold Max Power
THROTTLE ： Accelerate from Idle
to Max Power
Accelerate on the Center of the Runway
ELEVATOR ： Hold the Airspeed
AILERON ： Keep the Course
RUDDER　： Trin and Yaw-Damper
THROTTLE ： Hold Max Power
Reach the target Altitude
and Transit to Cruising
Pull Up and Airborn
Fig.7 Automatic Take-Off
1. GLIDE SLOPE
ELEVATOR ： Hold the Flight Path Angle
AILERON ： Keep the Course
RUDDER　： Trin and Yaw-Damper
THROTTLE ： Hold the Airspeed
2.
Glide on the scheduled
slope to the runway
FLARE ALT.
(7m)
FLARE
ELEVATOR ： Pull Up until
the Flare Pitch Angle
3. LANDING ROLL
AILERON ： Keep the course
ELEVATOR ： Pull Up until
RUDDER　： Trin and Yaw-Damper
the Taxing Pitch Angle
THROTTLE ： Decelerate until the Idle Speed
AILERON ： Keep the Attitule Horizontal
RUDDER　： Keep the Course
THROTTLE ： Keep Idling
Flight Path Angle (3.5deg)
Start to Pull Up at Flare Alt.
Keep the Center of the Runway
Fig.8 Automatic Landing
4
AUTOMATIC TAKE-OFF AND LANDING SYSTEM FOR FIXED WING SMALL AIRCRAFT
4 Results of Flight Tests
3000
2000
Cruise Phase Started
250
200
150
100
50
0
Target Alt 240(m)
10
0
-10
10
5
0
-5
-10
15
10
5
0
-5
-10
40
1000
EAS
(m/s)
Take-off Direction
-1000
RWY 25
0
07
Position North-South [m]
Altitude
(m)
Pitch Attitude
(deg)
Fig.10 is the time histories in the automatic takeoff phase. From the beginning of taxing to the end
of climbing, the course error was less than 1 meter,
which is accurate enough for take-off from a 20meter wide runway.
After airborne, the climb rate and the airspeed was
Course Error
(m)
4.1 Take-off Results
Take-off Started
Rate of Climb
(m/s)
An example of flight path in flight tests is shown
in Fig.9.
It shows that the aircraft traced the scheduled
flight pattern precisely.
30
20
10
Flight Path
Scheduled Pattern
460
※ This air-data sensor can't measure
airspeed below 17 m/s.
480
500
520
540
560
580
Altitude [m]
Time [sec]
300
Fig.10 Result of Test Flight
- Automatic Take-off
Target Alt 240(m)
200
Take-off
& Climb
100
Landing Approach
0
-4000
-3000
-2000
-1000
0
1000
2000
3000
Position East-West [m]
Fig.9 Result of Test Flight - Flight Path
4000
kept approximately same to the targeted values.
And, after reaching the target cruise altitude, 240
meters, the transition to the cruise phase was
achieved smoothly.
4.1 Landing Results
Fig.11 Automatic Landing
Fig.11 is the scene of automatic landing. The time
histories in the landing phase is shown in Fig.12.
The course error was about 2 to 3 meters, also
enough accuracy for landing to a 20-meter wide
runway.
After started the flare mode, the rate of decent
reduced smoothly, and just before touch down,
the rate of decent was 0.3 m/s, which means very
soft landing.
As is shown above, we confirmed the ability of
our automatic take-off and landing system. The
AFCS could keep the vehicle so stable that the
test pilot could hold his hands up during the
landing approach, shown in Fig.11 and Fig.13.
5
TAKASHI NAGAYAMA, NAOSHI AKAMINE, AKIHIRO YAMANE
Pitch Attitude
(deg)
Rate of Climb
(m/s)
Course Error
(m)
Altitude
(m)
Flare Started
5 Conclusion
Touch Down
100
80
60
40
20
0
We integrated an automatic flight control
system to an existing small aircraft with
minimum modification, and the automatic takeoff and landing experiment has completed
successfully for the first time in Japan.
This is very important result which leads to the
expansion of the freedom of UAV operations.
On the other hand, this result we achieved will
be applicable to future small aircraft, which will
be needed for the next generation aircraft
transportation system. On the basis of the
success of FABOT experiment, we are going to
respond to as many requests from various fields.
10
0
-10
4
2
0
-2
-4
10
5
0
-5
-10
EAS
(m/s)
40
30
20
10
980
1000
1020
1040
1060
Time [sec]
Fig.10 Result of Test Flight - Automatic Landing
The Test Pilot is holding his hands up
to demonstrate that the AFCS works good.
Fig.13 A Scene Just Before Touch Down
6