UAVS Project - (BEng Aerospace Dissertation) by ibnfirdous

VIEWS: 5,662 PAGES: 50

More Info
									            C      O V E                N T R                Y
        U N I V E R S I T Y

          Faculty of Engineering and Computing
              Department of Aerospace Engineering
                      BEng Aerospace Technology

                        388SYS Individual Project

                         UAVS Project 2009

                 Author: UBAIER AHMAD BHAT
                Supervisor: DR. S M HARGRAVE

Submitted in partial fulfilment of the requirements for the Degree of Bachelor of
                              Aerospace Technology

                               2008 - 2009
“I have not failed. I've just found ten
    thousand ways that won't work.”

              --Thomas Alva Edison

The work described in this report is the result of my own investigations. All sections
of the text and results that have been obtained from other work are fully referenced.
The dissertation and the original work related to it have not been submitted to any
institute for an academic award. I understand that cheating and plagiarism constitute
a breach of University Regulations and will be dealt with accordingly.

                                    Signed:                    Date:

   Bhat, U. 2009. UAV Project 2009, BEng Aerospace Technology
   Dissertation. Faculty of Engineering and Computing, Coventry

   Unmanned Arial Vehicles have seen an unprecedented growth in recent years
   in both military as well as civilian application domain. This has increased the
   interest and research in unmanned technology at academic level. Coventry
   University as also invested in technology related to development of UAVs like
   an autopilot and flight simulators. This report details the work done on
   implementation of an autopilot system on UAV and simulation of the aircraft.

   The approach used for this project was to get familiar with the autopilot
   system (Micropilot MP2028g) and install it onto an aircraft. The project also
   involved an attempt to simulate and analyse of the pitch dynamics of the
   aircraft model acquired.

   The autopilot was implemented on a test aircraft and was tested on ground
   but a flight test did not take place. Different packages for simulations and
   problems related to them were also investigated.

   Keywords: unmanned aerial vehicle, UAV, autopilots, flight simulation,


Declaration                                                        i
Abstract                                                           ii
Table of Contents                                                  iii
Acknowledgements and Dedication                                    v

1. INTRODUCTION                                                    1

     1.1   Aims and Objectives of the Project                      1
     1.2   Outline of the Report                                   1
     1.3   Modifications to project aim.                           2

2. LITERATURE REVIEW                                               3

     2.1   Unmanned Arial Vehicles                                 3
           2.1.1 Introduction                                      3
           2.1.2 Types of UAVs                                     3
           2.1.3 Components of a UAV System                        5
           2.1.4 Applications                                      6

     2.2   Stability and Automation of UAVs                        8

     2.3   Dynamics of Flight                                 10
           2.3.1 Introduction                                 10
           2.3.2 Six Degrees of Freedom (6DOF)                10
           2.3.3 Pitch dynamics and Longitudinal stability    10

     2.4   Modelling and Simulations                          12
           2.4.1 Introduction                                 12
           2.4.2 Modelling and Simulation Techniques          12
           2.4.3 Applications                                 12
           2.4.4 Problems                                     13

     2.5   System Identification                              14

3. UAV PLATFORM                                                15

     3.1   INTRODUCTION                                        15
     3.2   TEST PLANE                                          15
     3.3   2009 UAV PLATFORM                                   18

4. THE AUTOPILOT                                               20

     4.1   Introduction                                        20
     4.2   Components                                          20
     4.3   Installation                                        22

      4.4    Working and Configuration            23
      4.5    Problems                             23

5. FLIGHT SIMULATOR                               25

      5.1    Introduction                         25
      5.2    Merlin Flight Simulator              25
      5.3    Flight Simulation Modals             28
      5.4    Flight Tests                         28

6. System Identification using AeroSim            33

7. Conclusion                                     39

8. Further Work and Recommendations               40

9. References                                     41

10. Bibliography                                  42

11. Appendixes

A     Initial Documentation and Interim Report        I

B     Test Results                                IX


I would like to thank Dr. Stephen Hargrave for supervising the project and helping
me throughout the course. I would also like to thank Mr. Basini, Mr. Fisher, Dr.
Lambert and Mr. Thorley for their help for the project and the course.

I would also like to thank fellow students Mohammad Bahrami, Arnaud Dupin,
Thomas Gobeaut, Douglas Mackenzie and Kelvin Nelson.

I would like to thank Allah for giving be strength and patience throughout the project
and the course. I would also like to thank my family for their financial and emotional

I would also like to dedicate this project to those innocent children who have been
killed in UAV strikes worldwide and pray that the unmanned technology is not used
for any inhumane activities.


1.1 Aims and Objectives of the Report

The report presents the details of the work done on implementation of MicroPilot
MP2028g autopilot on the UAV platforms developed in the university. The report also
presents the results of research into the modelling and simulation of the UAV. The
main objectives set for the project were as follows:

   1. Familiarisation and implementation of the MicroPilot 2028
   2. Familiarisation with Merlin Flight Simulator and modelling and simulation of
      the UAV.
   3. System identification and generation of linearised model of the pitch dynamics
      of the aircraft.

1.2 Outline of the Report

Chapter 2: Literature Review:
This chapter background information about UAVs and their applications, Automation
of aircrafts, modelling and simulation of systems.

Chapter 3: UAV platform
This chapter gives details about the UAV platforms used for the project.

Chapter 4: Autopilot
In this chapter details about the MicroPilot MP2028g along with some information
related to installations and setup.

Chapter 5: Flight Simulation
This chapter contains detailed information about Merlin Flight simulators and gives
some details modelling and simulation of the aircraft.

Chapter 6: System Identification
This chapter contains some details about AeroSim Blockset.

Chapter 7: Conclusion

Chapter 8: Further Work and Recommendations



Appendix A:
This Appendix contains the initial documentation and the interim report of the project.

Appendix B
This Appendix contains the flight test and simulation results.


1.3 Modifications to the project aims

Initially it was planned to use Merlin Flight Simulator for simulation and modelling
purpose but due to some reason, which are discussed later in the report, this was
changed to AeroSim Blockset. This had a major effect on the aims that the project
had initially set to in regards to simulations.


2.1 Unmanned Arial Vehicles

2.1.1 Introduction

An Unmanned Arial Vehicles (or UAV), also known as a drone, is an aircraft without
a human operator on board (MSN Encarta c. 2008). UAVs are different that guided
missile or cruise missiles in that the air vehicle is designed to come back and be re-
used. These are also different from normal remote controlled aircrafts in that they
have a high technology control systems on board which means can operate out of
line of sight and at altitudes where person on the ground cannot readily see them.
These air vehicles still need a pilot or an operator who rather than being seated in
the aircraft itself is located in a control centre normally referred to as a Ground
Control Station. (UAVSA c. 2008).

2.1.2 Types of UAVs Classification by Aircraft Configuration

In general an aircraft is any flying machine/vehicle in all possible configuration be it
fixed-wing, rotary-wing, a balloon or an airship and potentially any aircraft can be
converted into an UAV with the right technology and control systems installed on
board (Beggs 2009). Therefore one of the ways classifying different types of UAVs is
by type of aircraft configuration.

a) Fixed wing UAV

These are the most common types
of UAVs and are currently being
used      for    wide     range    of
applications. These can be of
different sizes like the big Northrop
Grumman Global Hawk with a
wing span of 39.9m (Northrop
Grumman 2007a) to the small
backpack Casper form Becker
Avionics with a wing span of just
2.5 m (Becker Avionics 2006).
Some examples of different size
and range fixed wing UAVs can be
seen in figure 2.1a

One of the reasons for the
success of fixed wing UAVs is that
these are quite stable and require      Figure 2.1a Examples of some fixed wing UAVs (BBC
simpler   systems     to   control                         News 2009)
compared to other configurations.


b) Rotary-wing UAVs

There has been a lot of development in rotary-wing UAVs in recent years. The
manoeuvrability and the ability to maintain the aircraft in hovering of a Rotary wing
UAVs or a helicopter UAV present several advantages for UAV applications.
However, rotary-wing aircrafts are very unstable and therefore more difficult to
control and require the application of reliable control laws. (Remub et al. 2007:111)

    Northrop Grumman MQ-8B Fire Scout is an           A-160 Hummingbird by Boeing/Frontier
            advanced rotary-wing UAV                             (US OSD 2005)
            (Northrop Grumman 2007)
                           Figure 2.1b Examples of Rotary-wing UAVs.

c) Airship UAVs

A number of unmanned airship projects, both free-flying and tethered (aerostats),
have been initiated to provide synergistic capabilities to those provided by unmanned
aircraft, most notably extended persistence. There appears to be potential for
synergy between airships and UAS that enhance capability or reduce cost in several
mission applications including force protection, signals intelligence collection,
communications relay and navigation enhancement. An airship UAV’s most
significant challenge appears to be limited mobility. (US OSD 2005:32)

Example of some Airship UAVs can be seen in Figure 2.1c

Advanced Airship Flying Laboratory by American   Tethered Aerostat Radar System (TARS) by ILC
              Blimp Corporation                                      Dover
                          Figure 2.1c Types of UAVs (US OSD 2005)

CHAPTER 2: LITERATURE REVIEW Classification by Autonomy

Based on the autonomy of the flight control system the UAVs can be categorised into
three main levels:

Level 1: no autonomy on-board the aircraft. This will be more like a remote controlled
aircraft (similar to a hobby plane). However different systems on-board the aircraft
like sensors, cameras, GPS receivers and transmitters can be used to assist the on
ground operator to fly the air craft out of sight.
Level 2: semi-autonomous aircraft with autopilot on-board to keep the aircraft stable.
Difficult tasks like landing and takeoff will still be carried out remotely by a human
Level 3: fully-autonomous aircraft with onboard flight management system that can
manage the whole set operation (including takeoff and landing) once the mission is

These levels of autonomy are only related to the basic flying capabilities of the UAV
like take off, landing, cruse, some navigation and not to any other autonomous
function that a UAV might be required like tracking objects, object recognition,
collision avoidance etc. Autonomy of the aircraft has been discussed in detail in the
section 2.2.

2.1.3 Components of a UAV System

A UAV system can be divided into three main components. These will be the UAV
platform, the ground station and the payload (UAVSA c. 2008). These have been
discussed individually below.

                          Fig. 2.1c Components of a UAV system. UAV Platform

The UAV Platform itself will consist of the following components/systems:

a) The airframe

As mentioned earlier the airframe of a UAV can be of any configuration i.e. fixed-
wing, rotary-wing etc. Like any conventional aircraft the airframe must have the
required aerodynamic properties, be light weight and must be able to sustain the
specified loads for the different application.


b) The propulsion system

The type of propulsion system used within a UAV will depend on the type of aircraft
and task requirements.

c) Flight control System.

The flight control system (FCS) is the most important system within the UAV. The
system is connected to different types of sensors (gyroscopes, pressure sensors),
navigations systems and avoidance systems. Complexity of the FCS again depends
on the UAV application. Payload

The UAV may be required to carry different types of payloads. One of the most
common payloads is a camera or an optical sensing system. This is usually used by
the operator on the ground station to see were the UAV is going or look at particular
targets. Other payloads can be infrared cameras, radiation detectors, air sampling
systems etc. Military UAVs may have munitions, radars, scanners etc. Ground Control System or Station

At the ground control station the operator controls or monitors the UAV. The ground
station will therefore have avionics flight display, navigation systems, Position
Mapping system, system health monitoring. The Ground Station will also have some
kind of manual control devices like joysticks to control the aircraft. (UAVSA c. 2008)

The Ground Control station will also have system to receive data from the UAV like
video, photographs, etc depending on the onboard systems

2.1.4 Applications

Unmanned Arial Vehicles (UAVs) have seen unprecedented levels of growth in
military and civilian applications domains. UAV applications are usually referred to as
3D or D3, ‘dull, dirty and/or dangerous’. The primary applications identified and used
to date all involve putting the UAV and its payload in environments where the pilot in
a manned operation might be significantly at risk of losing his life or dying of
boredom. Currently the UAV applications are mostly defence related and main
investments are driven by future military scenarios. Today the civilian UAV market is
small compared to the military one. Significant civil markets for UAVs are still to
emerge, with only limited niche applications being currently available. However, in
the next 10-15 years, the expectations for the market growth of civil and commercial
aerial robotics are very high. (Ollero and Manza 2007:3)

There is also a cost justification for using UAVs. They can have longer operational
duration: they can require less maintenance: they can be cheaper on fuel to operate:
they can be operated remotely and sometimes autonomously carrying out the
mission with the minimum of human intervention and supervision: and they can be
deployed in a number of different terrains and are not always dependent on prepared


runways. Some argue that the use of UAVs in the future will be a more responsible
approach to certain airspace operations from an environmental, ecological and
human risk perspective (UAVSA c. 2008).

Below are some of the examples of UAV applications taken from UAVS website.

      Aerial Policeman and Crowd Monitoring
      Aerial Reconnaissance
      Aerial Traffic and Security Watch
      Air to Air Missiles, Air to Ground Missiles, Anti-Tank Missiles
      Battlefield Management
      Crop Dusting, Crop Management
      Disaster damage estimation, Disaster effects management
      Fire Fighting
      Fishery Protection, Forestry
      Geophysical surveys
      Guided Shells
      Life raft Deployment
      Litter on beaches and in parks
      Maritime and Mountain Search and Rescue
      Mineral exploration
      Oil and Gas Exploration and Production
      Oil and gas pipeline
      Pollution Control and Air Sampling
      Search and Rescue
      Telecoms relay and signal coverage survey
      Waterways and shipping
      Wide Area Munitions Deployments


2.2 Stability and Automation of UAVs

Need for taking control out of the human hands and putting it under some sort of
automatic control has been there since a long time in aviation history. The reason
being that of itself, the path of any aircraft is never stable and aircrafts only neutral
stability in heading. Without control, aircraft tend to fly in a constant turn. In order to
fly a straight and level course continuously-controlling corrections must be made
(McLean 1990). For the human pilot this can be very exhausting and therefore there
is a need for some sort of automatic flight control system (AFCS). In an Unmanned
the need for having some sort of stability devices and AFCS is therefore quite
obvious because an operator on ground will not be able to provide the necessary
control correction will be very difficult.

As mentioned in the previous section UAV can be dived into different categories
based on autonomy they have. This section will have a bit more detail into how
stability and autonomy of UAVs.

a) Stability of un-autonomous UAVs.

An un-autonomous UAV can simply be called a remote control aircraft or a hobby
plane. Some people will argue that such an aircraft cannot be categorised as a UAV
as mentioned section 2.1. The stability of such an aircraft will depend more on the
aerodynamic stability of the airframe. Such stability can be achieved by using
conventional methods like using a dihedral wing for roll stability and positioning the
centre of mass on front of aerodynamic centre for longitudinal stability (Kermode
2006). The ground operators will thus be responsible for manoeuvres and provide
corrections from time to time. The problem in such an aircraft will be that the
increase in aerodynamic stability will mean decrease in manoeuvrability.

b) Stability of semi-autonomous UAVs

In a semi-autonomous UAV the roll of operator is limited to manoeuvres lonely.
Flying the aircraft will be more like playing a video game where the operator can
move around using a joystick and does not have to worry about stabilising or
providing any control corrections. An onboard computer or an autopilot will to this
job. Apart from the aerodynamic stability methods mentioned in previous section
following methods/devices

       Aerodynamic Stability Augmentation System (SAS): This system makes
       use of the aerodynamic control surfaces like elevators, ailerons and rudders
       to provide improved handling qualities (McLean 1990:10). For example for the
       pitch stability will be achieved by applying elevator deflections proportional to
       the pitch rate which will provide additional damping. Example of such an
       autopilot system will be in Chapter 5. This type of systems will require
       gyroscopes to record the changes and provide feedback to the autopilot.

       Non-Aerodynamic Stability Augmentation Systems: Non-aerodynamic
       augmentation devices are used mostly for small and miniature UAVs. Such
       devices are used because adverse meteorological conditions generally affect
       smaller aircraft more strongly than they affect larger aircraft and there is no


                  known way to create controlled aerodynamic forces sufficient to counteract
                  the uncontrollable meteorological forces on miniature aircraft. Such systems
                  can use devices like telescopic shafts with a mass pod or counter weight at
                  the end. The telescoping shaft and mass pod are stowed in the rear of the
                  aircraft. When deployed, they extend below the aircraft. (Langley Research
                  Centre 2005) A similar example of such a system can be of the long rod or
                  shaft that a funambulist uses while tightrope walking to balance himself.

          c) Fully autonomous UAVs

          A fully autonomous UAV will have AFCS or autopilot that completely replaces the
          human operator in the control of the aircraft. Such AFCS will be capable of take off,
          landing, following a navigation path and depending on the specification able to
          perform non-aggressive and aggressive manoeuvres. The components and sensors
          with fully autonomous UAV will include gyroscopes, pressure sensors, GPS
          navigations system, collision avoidance system and electronic compass. The aircraft
          will uses similar methods of stability as mentioned earlier.

          Table 2.1 shows a list of commercially available systems autopilot system for UAVs.

                          Table 2.1 List of Commercially Available Autopilots (UAS Research 2008)
Company                          Autopilot Models Used on                                       website
A Level Aerosystems (Russia)     Autopilot           A Level Aerosystems, Russia     

                                                     RQ-2A/B Pioneer (Pioneer Inc., USA)
                                 MIAG                BQM-74 (Northrop Grumman, USA)
BAE Systems (USA)                                                                    
                                 NSU                 CL327 (Bombardier, Canada)
                                                     Mirach100/5E (Galileo Avionica, Italy)
Curtiss Wright -                 Vista Controls
                                                     Global Hawk (Northrop Grumman, USA)
Vista Controls Corp (USA)        IMMC
                                 Autopilot I         Cabure, Yagua (Nostromo, Argentina)
Mavionics (Germany)                                                                  
                                 Autopilot II        Yarara (Nostromo, Argentina)

                                                     Scarab (SCR, Spain)
                                                     X-Vision (SCR, Spain)
                                                     Boomerang (BlueBird Aero, Israel)
MicroPilot (Canada)              MP2028                                              
                                                     Blueye (BlueBird Aero, Israel)
                                                     ALBA (INTA, Spain)
                                                     Casper (Becker Avionics, Israel)
Rockwell Collins, Inc (USA)      G-311               Sky-X (Aliena Aeronautica, Italy)
Selex Sensors & AS (UK)          Selex FCS           Phoenix (BAE Systems, UK)       
                                                     Scorpio (EADS, France)
                                                     Copter I (SurveyCopter, France)
                                                     APID (CybAero, Sweden)
                                 wePILOT 1000
                                                     DRAC (EADS, France)
weControl (Switzerland)          wePILOT 2000                                        
                                                     DVF 2000 (SurveyCopter, France)
                                                     Skeldar (Saab, Sweden)
                                                     Tracker (EADS, France)
                                                     RMAX (Yamaha, Japan)


2.3 Dynamics of Flight

2.3.1 Introduction

In the previous section it was said that the reason why Autonomous Flight Control
Systems were used in aircrafts was to keep the aircraft stable and under control.
This section is gives some background information about to flight dynamics as it is
directly related to control an stability of a aircraft and thus links up with the
discussion about the autonomous flight controls

2.3.2 Six Degrees of Freedom (6DOF)

An aircraft has got six degrees of freedom in other words it can move in three
directions along three axes (up/down, forward, backward, lift, right) and rotate in
three directions around the three axis (pitch, roll and yaw).Figure 2.3a shows the
6DOF in a body axis system and also shows the conventional nomenclature for
different forces, velocities and moments.

              Figure 2.3a Six degrees of freedom of an aircraft in body axis system
                                       (McLean 1990:17)

Stability or the lack of it is a property of an equilibrium state. The equilibrium is stable
if, when the body is slightly disturbed in any of its degrees of freedom, it returns
ultimately to its initial state. An aircraft which may be stable with respect to one
degree of freedom but unstable with respect to another (Etkin 1996). Therefore
based on the inherent stability of the aircraft and AFCS does not to control all the
degrees of freedom. And for this reason autopilots used on an aircrafts can be one-
axis, two-axis or three axes.

2.3.3 Pitch dynamics and Longitudinal Stability

One of the aims set for the project was to look into pitch control of the UAV and
therefore it is important to give some brief background to pitch dynamics. The pitch
of the aircraft is controlled by the elevators. Any deflection on the elevator will
change the pitch of the aircraft. The aircraft usually oscillates at this point before


settling on the new pitch angle. The change in pitch can also be induced by
activation of high lift devices like flaps, movement of centre of gravity or external
forces and turbulence. There are types of oscillation around the longitudinal axis.

a) Short Period Pitching Oscillations

      The short period pitching oscillations is a damped oscillation about the pitch
      axis. Whenever the aircraft is disturbed from its pitch equilibrium the mode is
      excited ad manifested itself as an oscillation in which the principal variables
      are pitch rate and angle of attack. Typically the frequency of the mode is 0.5
      to 2 Hz. In manned aircraft, this is in the region of a human pilot natural
      frequency, therefore it is essential that the short period mode should be well
      damped, otherwise severe handling problems can arise (NFTC 2007:22). For
      computer controlling a UAV will usually be capable to handle a higher
      frequency but it still needs to be damped as it will damage the onboard
      systems or payload. It is the short period oscillation which recovers the angle
      of attack to its trim value, and this happens sufficiently quickly that the speed
      remains substantially constant.

b) Long Period Pitch Oscillations

      The long period pitch oscillations, also known as phugoid oscillations, is the
      most commonly seen as a lightly damped low frequency oscillation in speed,
      which couples with height. Whenever the aircraft is disturbed from trim speed,
      the mode manifests itself as a sinusoidal oscillation in which the principal
      variables are pitch attitude and speed changes. A significant feature of the
      mode is the angle of attack remains substantially constant throughout. Thus
      the change in pitch attitude gives a corresponding change in flight path angle,
      and hence the height excursions. The period is long, typically in the band of
      40 to 100 seconds, and the damping mainly due to the rate of change of drag
      of the aircraft whit speed change. I is possible to deduce that the period of the
      mode is determined most entirely by the datum speed, and a good
      approximation is that the period is seconds is equal to 0.25 of the true
      airspeed in knots (NFTC 2007:23). Since the phugoid mode has such a low
      frequency, it poses an undemanding task, and an AFCS can control the mode
      even when it is unstable. The phugoid manifests itself as a trimming problem,
      which, although not regarded as hazardous when poorly damped however it
      will hinder the UAV applications like taking images, sensing or tracking a


2.4 Modelling and Simulation

2.4.1 Introduction

Modelling and simulations are one of the most important parts of any control system
design and integration. This section gives some background information about
modelling and simulation and how it can be used for UAV systems.

2.4.2 Modelling and Simulation techniques

In order to understand the behaviour of a system, mathematical models are needed.
These models are equations which describe the relationship between the inputs and
output of a system. The basis for any mathematical model is provided by the
fundamental physical laws that govern the behaviour of that system (Bolton
2003:185). For an aircraft the basis of the model will be the relationship between
different Newton’s Laws and Aerodynamic laws that define the motion of an aircraft.

As said in the previous section an aircraft has got six degrees of freedom. This
makes the mathematical model of the aircraft very complex and will take a lot of time
and effort to put all the relationships together. It is therefore important to set of
conditions and assumptions which will define the boundary of the system. For
example with the mathematical modelling of the aircraft a simpler model can be
made by considering only three degrees of freedom or even just two degrees of
freedom. This would make the equations much simpler.

Another important thing to know in modelling is that the equations will include
different constant values or coefficients which are set for the physical system. These
coefficients are needed to be determined for the model to represent the true system.
For example with a simple spring mass model it is important to know the spring
coefficient and the mass for the model to truly represent the system. For an aircraft
these can be different aerodynamic coefficient like lift coefficient, drag coefficient etc.

These coefficients may be determined by doing measurements and calculations on
the actual system, by experimentation or even by simulations that used a different
set of data. For an aircraft this might mean doing wind tunnel tests or CFD tests.

Simulation is the process in which a range of inputs are put through the
mathematical model and the outputs would show the behaviour of the system. Once
the model has been verified it can be used for designing the controller and for
predicting the actual model.

Modelling and simulations are usually done with aid of computer packages. Table
2.4a lists some of the modelling tools that are available for aerospace applications.

2.3.3 Applications

As mentioned in the beginning modelling and simulations are a very important part of
control system designs. These can also be used to investigation behaviour of
systems for optimisation or proving a new concept without risking a real system.


                 Table 2. List of Flight Modelling and Simulation tools
Package                          Description
Merlin Flight Simulator          Commercially available flight simulator that enables
                                 the user to configure their own aircraft models for
X-plane                          This is a flight simulator package which enables
                                 users to configure their own aircrafts. This program
                                 can run on and PC
JSB Sim                          This is another flight simulator package with user
                                 definable aircrafts. It’s an open source software and
                                 therefore will enable to the user to make changes
                                 and even change the working of the simulations
YS Sim                           This is another flight simulator package with user
                                 definable aircrafts. It’s an open source software and
                                 therefore will enable to the user to make changes
                                 and even change the working of the simulations
Matlab                           This is a very powerful mathematical computational
                                 package which is widely used for modelling and
                                 control related applications
Simulink                         Part of Matlab this package enables modelling a
                                 system with the help of GUI blocks
AeroSim Blockset                 This package working using Simulink and is
                                 designed for aerospace applications and modelling.
Aerospace Blockset               This package working using Simulink and is
                                 designed for aerospace applications.

2.3.4 Problems with modelling

A model is defined by assumptions about the real system and therefore can never be
hundred percent accurate. These assumptions can also include the different
coefficients used for the modelling and without the right values the model will be way
off form the real system. In order to get more accurate results a more complex model
will be required. The problems in doing so will be that a complex model will require
more computation and might cause delay when used in control systems.


2.5 System Identification and control

2.5.1 Introduction

System Identification is the a method of developing a model from a real system in
which tests are used to determine the response from the system to some input, e.g.
a step input and then finding a model that fits the response (Bolton 2003:240).

2.5.2 System Identification tools

Different software packages can be used for system identification some of these are
listed in Table 2.5a
               Table 2a. List of System Identification Software packages
Package                           Description
Matlab                            A very powerful computational tool with may system
                                  identification functions
Simulink                          Part of Matlab this package enables modelling a
                                  system with the help of GUI blocks
System Identification Toolbox This is a Matlab/Simulink model used for system
SIDPAC                            This is a Simulink block set used specifically for
                                  aircraft system identification.

These tools can be used to analysis the outputs for the system on step-response
plots, bode plots and root locus plots.


3.1 Introduction

There were two UAV platforms that were planned to be used for this project. One
was built as a test plane for 2008 UAV built projects and the second one was to be
built for the 2009 UAV projects. The second one was to be constructed with the help
of other students working on other aspects of UAV project.

3.2 Test Plane

The test plane is a high wing Clipped Wing Taylorcraft ARF model aircraft. The
aircraft had been chosen for its stability and enough space for the payload (Castle
2008:37). The test plane had been flown by the 2008 UAV Project students using
remote control and therefore aim was to use the test plane for the testing of autopilot
and initial modelling for 2009 UAV projects as well.

                   Figure 3.2a Top Gun Clipped Wing Taylorcraft from Hanger-
                                      9 (Hanger-9 2008)

3.2.1 The Airframe

The dimensions of the airframe are given in the table below.

                           Table: Speciation of the Test Plane
Wing span                                    1.07 m
Wing Area                                    0.428 m2
Length                                       1.485 m
Weight                                       6 kg approx

3.2.2 Propulsion Unit and Speed Controller

The aircraft uses an ‘AXi 5330/18 Gold Line’ electric motor with radial mounted
propeller holder (see figure 3.2b). The motor uses two FlightPower EON28 11.1 LiPo
batteries. The output is controlled through a Jeti ADVANCE 90 plus controller.


                             Figure 3.2b (Model Motors 2006)

                                      Table 3.2a
                               AXi 5330/18 Gold Line
                      Specification (Model Motors 2006)

              No. of cells                                     max 32
                                                       max. 10s Li-Poly

              RPM/V                                            259 RMP/V

              Max. efficiency                                        90%

              Max. efficiency current                 25 - 60 A (>85%)

              No load current / 10 V                                  2A

              Current capacity                                  75 A/60 s

              Internal Resistance                                  32 m

              Dimensions (Diameter x Length)                   63 x 64 mm

              Shaft diameter                                        8 mm

              Weight with cables                                    652 g


                  Table 3.2a
   Jeti ADVANCE 90 plus controller Specifications
                  (Jeti 2005)
     Dimensions:              65 x 55 x 17 mm
        Weight:                   75/90 g
  Continuous current:               90 A
   Accu NiXX / LiXX:            14-32 / 5-10
        Voltage                  15 – 42 V
                                                        Figure: 3.2c Jeti Advance 90 plus speed

3.2.3 Control system

                                       PCM9X                    RS10DS.
The aircraft is manually controlled by PCM ii Transmitter via RS10DS The aircraft
has got four servos which control the ailerons, elevators and rudder.

           JR PCM9X ii Transmitter                                 JR RS10DS Receiver
                                     Figure: Manual control unit

The aircraft is autonomously controlled by Micro ilot MP2028g autopilot. Details
about the autopilot are given in Chapter 4.

3.2.4 Power supplies

The aircraft has got separate power supply autopilot, servos and the motor.

3.2.5 Work done on test plane

The test plane had been assemble by the 2008 Project students but there were
some of the tasks required to make the aircraft airworthy

   •                                                                systems.
       Installation of remote control receiver, autopilot and other systems
   •   Measurement of the aircraft for modelling and simulation.


3.3 UAV 2009 Platform

3.3.1 The Airframe

Figure 3.3a below show the dimensions for the aircraft. The airframe was designed
by other students working on the UAV project. The aircraft was still under
construction at the time of this report being written.

                  Figure 3.3a Pheonix 09 UAV designed by M. E .Bahrami


                       Figure 3.3b UAV fuselage under construction.

3.3.2 Propulsion Unit and Speed Controller

Same as the test plane. See section 3.2.2

3.3.3 Control system

Same as the test plane. See section 3.2.3

3.3.4 Power Supply

Same as the test plane. See section 3.2.4

3.3.5 Other components

An extra servo will be used to unlock and drop the flaps.


4.1 Introduction

The autopilot used for the UAV is a MicroPilot MP2028g. The autopilot is capable of
flying the UAV fully autonomously. The autopilot uses gyroscope, air pressure and
GPS as senses for inertia, speed, altitude and position to control the aircraft. During
the project the autopilot been tested on ground but a flight test could not be

4.2 Components

The MicroPilot MP2028g autopilot system consists of the following components.

                       Table 4a: List of components for autopilot system
Component                                    Description

MP2028g CORE

                                                    This is the main autopilot board and all the other
                                                    components are connected to it.

                    Figure 4.2a (MicroPilot 2004)

MP2028g AGL

                                                    Above Ground Level (AGL) is an ultrasonic altimeter
                                                    that provides altitude information to an altitude of
                                                    about 4.88m above the ground. It is required for
                                                    autonomous takeoff and landing.

                    Figure 4.2a (MicroPilot 2004)


MP-Servo board

                                                  The Servo board distributes the servo signal to each

                  Figure 4.2a (MicroPilot 2004)


                                                  This is the GPS antenna which connects to the
                                                  CORE board. The antenna is mounted on a 10cm x
                                                  10cm copper board.

                  Figure 4.2a (MicroPilot 2004)

Radio Modems and MP-COM board

                                                  The radio modem connects to the ground station
                                                  computer and communicates with the autopilot via
                                                  the COM board.

                                   Figure 4.2a

Faraday Cage

                                                  The Faraday cage had been designed by 2008 UAV
                                                  Project Students to shield the CORE board from any
                                                  external interference especially from the motor.

                                  Figure 4.2a


Software                                 The software is used to communicate with the
                                         autopilot from the ground station.

                                         MicroPilot Horizon:
                                         This software provides a GUI for configuring the
                                         autopilot and setting up flight path.

                                         Microsoft HyperTerminal:
                                         The HyperTerminal can be used to communicate
                                         with the autopilot. It uses a commend line
                                         interference but will enable to configure more set
                                         and acquire flight data.

4.3 Installation

The autopilot components were installed on the control board. In this section only the
important points to be remembered while installing the MP2028g is mentioned. Full
details of installing MP2028g can be found MicroPilot Autopilot Manual and
Installation DVD.

a) Control Board

All the autopilot components are installed on a single board. The layout of the control
board for the test aircraft is shown in figure 4.3a

                     Figure 4.3a: Autopilot Board for Test Plane.

b) Servo Board


The servo board was connected to the servos for normal setting. This means that
only servos for ailerons, elevators, rudder and throttle were connected. Figure 4.3b
below shows which pins are supposed to be connected to which servo.

                              Figure 4.3b: Servo Board

4.4 Configuration

The autopilot was configured using the MicroPilot Horizon software. It was
configured for Normal settings and under a fake GPS lock.

The autopilot was then tested on ground for basic servo responses.

4.5 Problems with the autopilot

There were some issues with configuration the autopilot

a) Communication Problem

The autopilot should be able to communicate with the ground station either by using
MicroPilot Horizon or by using Microsoft HyperTerminal. The autopilot was able to
communicate with the Horizon software properly most of the time but it did freeze the
computer while configuring the Rudder setting couple of times. This behaviour
however only happened while using the personal laptop and did not occur when
universities laptop was used. This might have happened have happened due to the
former being too slow.

The main communication problem occurred when HyperTerminal was used to
communicate with the autopilot. The HyperTerminal only communicated once and for
other attempts showed nonsense values on the screen as shown in figure 4.5.

The communication using HyperTerminal is very important to get the flight data for
any further investigation. Most importantly the data is very important for analysis for
selection of proper gains.


  Figure 4.5a Screenshots of the HyperTerminal data. Top: The data on the HyperTerminal screen
  when a proper connection was estabilished. Above: The HyperTerminal data showing nonsense

b) Battery problem:

The autopilot seems to drain the batteries in a very short period of time and this
delayed configuring it. However with the new batteries this did not seem to be a

The autopilot had some issues with the AGL unit in the 2008 UAV project. However
the AGL seem to respond properly during the ground tests showing the right
distance from ground. It might be possible that the interference from the motor would
have affected the AGL unit in the past.


5.1 Introduction

The UAV was planned to be simulated using Merlin Flight Simulator present at
Coventry University’s Aerospace Lab. The objective behind doing so was to
understand the stability of the aircraft so that the proper gains for the MicroPilot
MP2028g can be determined without risking the UAV in a real flight test. In order to
do this it was important to understand the working of Merlin Simulator1 and its own
Arthur Autopilot.

5.2 Merlin Flight Simulator

5.2.1 Simulator Choice

There were two flight simulators available for the simulations, Merlin’s MP520 and
MP521. Some of the options available on the two can be seen in Table 5.2a.

           Table 5.2a Merlin Flight Simulators present at Coventry University.

                                       Merlin MP520                Merlin MP521
                                                                     Excalibur I
   Aircraft Config. Software             Excalibur I
                                                                  Excalibur II (beta)
   Aircraft Systems
   Autopilot                                                            Arthur
   Flight Controls Systems
   Motion base actuators                   Electric                   Hydraulic

Because no autopilot was present on MP520 it was initially decided to use MP521
for the simulations. However MP520 could still be used if the autopilot was not
needed for the particular flight and if the model was not for Excalibur II format.

5.2.2 Aircraft Configuration Software

There were two aircraft configuration software available with MP521, the Excalibur I
(Ex-I) and Excalibur II (Ex-II). Excalibur I is a stable version and also works on
MP520. However the aircraft model designed using Excalibur I is very simple and it
does not provide options for advance design like multiple section wing, horizontal
and vertical panels.

MP521 had beta version of Excalibur II. This version does provide advance
configuration. During the initial stage of the project Ex-II was used to configure


aircraft models but due to some problem the models did not work or were unstable.
These issues were reported to Merlin. Another problem with Ex-II was that it could
not be installed on any other computer which meant that the aircraft needed to be
configured in lab only. Therefore it was decided to use Ex-II only and keep the
models simple.

5.2.3 Autopilot

The MP521 simulator has software autopilot installed on it called Arthur. Arthur is a
4-axis (roll, pitch, yaw and altitude) autopilot which means it can be used for fixed
wing as well as rotary wing aircrafts. The cockpit controller deflections (from. pitch
and roll stick, Yaw pedals and throttle lever) are mixed with autopilot servo deflection
to provide closed-loop control. (Merlin Products 2007) Control Modes

Arthur provides control using five modes. Figure 5.2a shows the Control Panel inside
the simulator cockpit to control the autopilot. It also shows the backend for Arthur
which shows the current status of the autopilot and can also be used to set gains for
the different modes

1.) Stability Augmentation System (SAS) Modes

       The stability augmentation seeks to provide handling qualities in roll, pitch and
       yaw by providing additional damping. For example for the pitch stability this is
       achieved by applying a control surface deflection proportional to the pitch rate.

       It is important to engage the SAS Modes before any other autopilot modes
       can be used.

2.) Autopilot (AP) Modes

       The AP modes maintain a particular flight parameter. For example for the
       Altitude (ATT) mode the autopilot will maintain the aircraft at a specific pitch
       altitude. ATT mode will either maintain the current altitude when the mode is
       engaged or will move the altitude set in command mode.

       There are three modes available under AP mode:
       a) Altitude (ATT) mode
       b) Wing Leveller (WNG LVL) mode
       c) Turn Co-Coordinator (TC) mode


    Figure 5.2a Merlin Flight simulator Autopilot: The Flight Control Panel and Arthur backend.

3) Command (CMD) Modes

      The CMD modes are used to set the demands for AP mode. For example the
      Altitude hold mode will be used to set the required altitude and vertical speed.
      The ATT autopilot mode will then try to maintain this altitude.

      There are three modes available under CMD mode
      a) Heading Hold (HDG) Mode
      b) Altitude Hold (ALT) Mode
      c) Speed Hold (IAS) Mode

      These modes can be engaged only if the related AP modes are engaged first.

4) Navigation (NAV) Mode

      Arthur includes two navigation modes for tracking VOR beacons and
      instrument landing systems. These modes were not required for this project.


Control Mixing

It is important that the aircraft model has got External
Controls enabled on the system settings of Excalibur
in order for autopilot to engage (see figure 5.2b).

5.3 Flight Simulation Models

In order to understand the working of the flight
                                 pre-built stable model
simulator and Arthur autopilot a pre
of Cessna172 was used. The plan was to modify this
model convert it into the Test Plane model and UAV            Figure 5.2b The external Controls
                                                              settings must be checked on the
model.                                                        Systems settings on the Excalibur
                                                                editor for autopilot to function.
One of the issues with the configuring the models
were the simulator could not run small models and therefore the models were sca
up. Another issue was that Excalibur only as combustion engines as power plant
which means that the total mass of the aircraft will reduce once the consumption of
fuel. One way to overcome this was to use a value of zero for SFC.

5.4 Test Flights

The aim of the flight testes was to look into the longitudinal stability of the aircraft.
This meant analysis of the natural short-period oscillations and long-   -period (or
Phugoid) oscillations damping of the aircraft and how the autopilot will assist in
further damping.

The tests were done to find the open loop response and close loop response i.e.
when the autopilot was engaged.

5.4.1 Open Loop and Close Loop Response Test

In order to get the open loop response of the aircraft different types of signals were
manually inputted using the control stick.

The closed loop response flight tests were similar to the open loop test the only
difference being that the autopilot was switched on for the tests.

One of the main problems encountered during the tests was that a perfect and
repeatable input could not be inputted manually. Another problem for these tests
was that as these were done quite earlier stage in the project getting used to the
flight simulator took some time.

5.4.2 Comparison test

These testers were done in the last stages of the project using Merlin Simulator in
order to see how the different modes available within the autopilot will affect the
stability if flight. For this three flights were conducted. The Cessna 172 model was


a) Autopilot off: For the first test took place in autopilot off. The flight test procedure
is given below

Time (min)          Action                                     Comment
00:00               Fly from 609.6m (2000ft) with IAS
                    129.64 km/h (70 kt)
02:15               Pull stick back for 45 sec                 Aircraft stalls
03:00               Release stick                              Wait for aircraft to settle
11:30               Stop.                                      Save as fly1

b) Autopilot on SAS Pitch mode. For this test the autopilot was left on in stability
augmentation mode

Time (min)          Action                                     Comment
00:00               Fly from 609.6m (2000ft) with IAS
                    129.64 km/h (70 kt) SAS mode
02:30               Pull stick back for 45 sec                 Aircraft stalls
03:15               Release stick                              Wait for aircraft to settle
11:45               Stop.                                      Save as fly2

c) Autopilot on in ATT mode. For this test the autopilot was set in ATT mode. The
flight test procedure is given below

Time (min)          Action                                     Comment
00:00               Fly from 609.6m (2000ft) with IAS
                    129.64 km/h (70 kt) ATT mode
02:15               Pull stick back for 45 sec
03:00               Release stick                              Wait for aircraft settle
11:30               Stop.                                      Save as fly3

The results for these tests are discussed in next section.

5.5 Results and Analysis

5.5.1 Open Loop and Close Loop Response Analysis

As mentioned section 5.4 these tests were done in the earlier stage of the project.
The full list of test and results can be found in Appendix B.

There was a major problem with these tests because proper input signals could not
be generated and because of the unfamiliarity with the M521 keeping the aircraft in a
straight path was found to be very difficult.


                                                                                                      Elv Angle
                                                                                                      Pitch Angle
                                                            X: 43
                                                            Y: 20.9
                                                                                X: 61
                                                                                Y: 17.15



                                                X: 33
                                                Y: -9.31



                                         X: 28
                                         Y: -28.61

                 0      10       20      30            40             50   60              70   80   90             100

                     Figure : Pitch response for a impulse elevator deflection

5.5.2 Comparison test analysis.

Figure 5.5.2a, 5.5.2b and 5.5.2c show the results from the comparison test in No
                                                               .     differences
Autopilot Mode, SAS Pitch Mode and AP ATT Mode respectively. The differ
between the three are quite clear from these plots.

                             Figure 5.5.2a Results for the flight without Autopilot.


                    Figure 5.5.2b: Aircraft flight in SAS Autopilot mode.

                   Figure 5.5.2c: Aircraft flight in ATT Autopilot mode.


Without the autopilot the elevator deflection looks almost like a square signal which
is maintained at a constant value till the control stick is released. Comparing this with
the SAS Pitch mode plot, it is possible to see deflections even when the control stick
has been held in set position. This is even more visible in the third test when the ATT
mode was engaged on the autopilot. Another difference between the three is the
amplitude and frequency of oscillations in altitude plot of the three tests. The values
for these are given in table 5.5a

                             Table 5.5a : Results analysis of comparison flight

                                                                       Control Stick Pulled

                                                                    No. of                Frequency
                                                          Amp1   oscillations       t1     ω (rad/s)        T (sec)
                                                Fly 1    50.5367 3.020134          14.9    0.42169            45
                                                Fly 2     80.8    2.472527         18.2    0.34523            45
                                                Fly 3     48.7    3.020134         14.9    0.42169            45

                                                                      Control Stick Released
                                                                      No. of              Frequency
                                                          Amp2      oscillations    t2     ω (rad/s)        Ts (sec)
                                                Fly 1     48.6           11        21.1    0.297781          232.1
                                                Fly 2     49.6           10         21     0.299199           210
                                                Fly 3      4.3            5        22.6    0.278017           113

       Figure 5.5.2d . Cmparison Test analyis

5.6 Conclusion

There wasn’t much analysis done one the data from the flight simulations because
most of these test were done on C172 model in order to get used to the simulator
and for the actual UAVs the data would be quite different. The results were however
recorded and some examples can be seen in the Appendix B. This was done at an
early stage of the project and it was found that the input signals were making the
aircraft unstable. It was also for that the data was difficult to analyse as the results
were not the perfect step responses which were shown on text books. At that point it
was thought that may be a different package could be used to simul  simulate to have more
control over the simulation.


6.1 Introduction

In order to get more controlled modelling and simulation different options for system
identification tool were researched. Based on the research done U-Dynamics
AeroSim Block set seemed to be a good option to use for both simulation and
system identification.

6.2 Reasons for choosing AeroSim Blockset

The AeroSim seemed to be a very attractive package for modelling and simulation
part of the project for the following reasons:

   •   AeroSim offered a wide range of complete aircraft models for Simple up to
       6DOF Aircraft models (U-dynamics 2008)
   •   AeroSim could be used on any PC with Matlab/Simulink installed on it. (U-
       dynamics 2008). This was very attractive as it meant the data can be
       analysed straight after the simulation was done.
   •   AeroSim offered Visual output to FlightGear Flight Simulator and Microsoft
       Flight to simulator. (U-dynamics 2008) The block also included a Cessna 172
       model for FlightGear demo. This seemed to be a good feature as same
       model had been tested on Merlin Flight Simulator.
   •   From the research done and looking into different journal articles and
       conference papers it was found that AeroSim blockset was used for many
       similar projects.
   •   The blockset was free for Academic and Educational purposes. (U-dynamics

Based on the above features offered by AeroSim blockset decision to move the
simulations from MP521 to AeroSim for system identification was taken.

6.2 Aircraft model configuration problem with AeroSim

This was the main problem with using the blockset. The aircraft configuration model
for AeroSim was very complex compared to the models used previously. It required
the many aerodynamic coefficients which were not used for any of the previous
models and therefore were not know. For example the Lift coefficient as a sum of
many other coefficients as shown in the equation below.

To determine the coefficients required much deeper understanding of aircraft
aerodynamics. The other problem was that to derive these coefficients manually
would take a lot of time and would therefore shift the whole aim of the project.


The block model for the Cessna 172 was completed even the aircraft configuration
was not complete. Figure 6.2a and Figure 6.2b

      Figure 6.2a Simulink model designed for C172 using AeroSim Blockset.

                    Figure 6.2b Response for running the Simulink model


6.3 DDATCOM software package

Research was done to find some kind of computational tool to derive the coefficient.
A software package called DDATCOM was found to be used in some projects that
               e               coefficients.
could give these aerodynamic coefficients. Data compendium (DATCOM) had been
developed by United States Air Force (USAF) and contained a large collection of
information containing of classical aerodynamics and analysis and experimental
data. (Jung and Tsiotras 2007) DDATCOM or DigitalDATCOM is a computer
database of this information and uses it on to the aircraft model to determine the
aerodynamic coefficients.

6.3.1 DATCOM modelling

DATCOM used a different set of perimeter for aircraft configuration. In particular it
required coordinates for the shape of the aircraft. In order to do so measurements
were taken from C172 photograph. This can be seen in figure 6.3a.

     Figure 6.3a: Measurement taken from Cessna 172 photograph (EuroControl, c. 2009) for

After taken the measurements data was put into the DDATCOM aircraft configuration
file. Based on this data DDATCOM produced a 3d model of the aircraft as seen in
figure 6.3b

Despite the 3d model looking alright there were some problems within the
configuration which caused error and the results for the model could not be accrued.
The screen shot of the output file are shown in figure 6.3c


                       Figure 6.3b: DATCom model for Cessna 172

                         Figure 6.3c Output file from DDATCOM

It wasn’t possible to resolve these errors in the due time. DDATCOM did plot some
graphs which can be found in Appendix B.


6.4 Flight Gear simulation problem

The AeroSim blockset contained an example of a C172 model that would simulate in
Flight gear package, see figure 6.4a. The problem was that the C172 model was
actually in flight gear and therefore could not be used for any other analysis. Also
problems were encountered while trying to use this example. Example can be seen
in 6.4b.

                 Figure 6.4a C172 model simulation in Flight Gear

           Figure 6.4b Error while trying to run Flight gear with AeroSim.


6.5 Conclusion

As can be seen from the previous sections the experience with using the AeroSim
did not go very well and no system identification analysis was done unfortunately.
The models and configuration files can be found in the CD attached with this report.


The aim of this project was to implement and understand the working of MP2028 and
to model and simulate the aircraft and the autopilot. Even though the project did not
go as planned the following can be listed as the achievements or the positive points
of the project.

   •   Better understanding of aircraft control systems.
   •   Familiarisation and implementation of the MP2028g autopilot
   •   Better understanding of working of Merlin Flight simulator
   •   Knowledge of different packages available for simulation.

However there are some things that the project did not achieve.

   •   The simulations did not work as planned specially with AeroSim Blockset.
       This also ended up in lack of analysis.
   •   A linearised model of the pitch dynamics was not designed.
   •   A real flight test did not take place and therefore no real data could be
       acquired for further analysis.

The move to use AeroSim Blockset made the simulation side of the project more
complicated. Also not getting to fly the test plane according to plan was also a de-
motivation. If flown the data from the aircraft would have been good for comparison
for simulations and broadened the analysis. There was a very useful knowledge
gained during the project though and therefore if the project was run again the use of
AeroSim Blockset will still be considered but this time he models would be designed
specifically for the blockset rather than adapting or converting the models for other


There is a lot of potential work for continuing and improving all aspects of this

The autopilot has not been used in a real flight till now and therefore the full potential
of the autopilot has not been tested till date. Therefore there is still some work that
can be done in this area.

As seen in the report the major problems were related to simulations. It would be
recommended to install Matlab/Simulink on the Merlin Simulators. This can be used
for more controlled simulations and analysis of aircraft models. It might be possible
to integrate the models with AeroSim Blockset.

All the open source and freeware software like DDATCOM, Flight Gear, and
AeroSim has been included in the project and would help further work related to a
similar project.


BBC News (2009) Robot planes take to the skies [online] available from
    <> [12 March 2009]
Becker Avionics (2006) Casper 250 Manual [online] Becker Avionics. Available
    0Manual.pdf> [28 Oct 2008]
Beggs, B. (2009) ‘BAE Systems: Tornado to Typhoon & UAV’. Royal Aeronautical
    Society Lecture delivered on 21 January 2009 at Coventry University Techno
Castle, R. (2008) UAV Built Project 2008. Unpublished dissertation. Coventry
Etkin, B. & Reid, L. D. (1996) 3rd edn. Dynamics of Flight. USA: John Wiley &
    Sons, Inc.
EuroControl (c. 2009) Aircraft Performance Database V2.0 [online] available from
    <> [31 Jan 2009]
Hanger9 (2009) Clipped Wing Taylorcraft ARF (HAN1850): Hangar 9 [online]
    available from <http://www.hangar-> [March 2009]
Jeti (2005) Jeti Model ADVANCE 90 plus [online] available from
    <> [Dec 2008]
Jung, D. ; Tsiotras, P. (2007) 'Modelling and Hardware-in-the-Loop Simulation for
    Small Unmanned Arial Vehicle.' Proceedings from AIAA Infotech@Aerospace
    2007 Conference held May 7-10 2007 in California. AIAA 2007-2768, Available
    from <> [03 Feb
Kermode, A.C. (2006) 11th edn. ed. by Barnard, R.H. & Philpott, D.R. Mechanics of
    Flight. Harlow: Pearson Education Limited
McLean, D. (1990) ed. by Grimble, M.J. Automatic Flight Control Systems. UK:
    Prentice Hall International Ltd.
Merlin Products (2007) Arthur Autopilot User Manual, UK: Merlin Products.
MicroPilot (2004) MicroPilot Installation Video [DVD] Canada: MicroPilot
Model Motors (2006) AXI 5330/18 GOLD LINE [online] available from
    GOLD> [12 Dec 2008]
MSN Encarta (2008) Unmanned Aerial Vehicle [online] available from
    ml> [27 Oct 2008]
Northrop Grumman (2007a) RQ-4 Block 20 Global Hawk [online] available from
    <> [14 March 2009]
Northrop Grumman (2007b) MQ-8B Army Fire Scout [online] available from
    > [14 March 2009]
Ollero, A; Manza I (2007) 'Introduction' In Multiple Heterogeneous Unmanned Arial
    Vehicles. ed. by Ollero, A. & Maza, I. Germany: Springer-Verlag: 1
Raptis, I. A. ; Valavanis, K. P. (2007) 'Airplane Basic Equations of Motion and
    Open-Loop Dynamics.' In Advances in Unmanned Arial Vehicles. ed. by
    Valavanis, K. P. The Netherlands: Springer: 49


Remub, V. ; Deeg, C. ; Musial, M. ; Hommel, G. ; Cuesta, F. ; Ollero, A. (2007)
   'Autonomous Helicopters.' In Multiple Heterogeneous Unmanned Arial Vehicles.
   ed. by Ollero, A. & Maza, I. Germany: Springer-Verlag: 111
UAS Research (2008) ‘UAS Yearbook 2008/2009’ [online] USA: University of North
   Dakota. Available from <> [11 Nov 2008]
UAVSA or Unmanned Aerial Vehicle Systems Association (2008) Unmanned
   Aerial Vehicle Systems Association What is a UAV? [online] available from
   <> [27 Oct 2008]
US OSD (2005) ‘Unmanned Aircraft Systems Roadmap 2005-2030’ [online] USA:
   Office of Secretary of Defence. Available from
   <> [03 Mar 2009]


Beggs, B. (2009) ‘BAE Systems: Tornado to Typhoon & UAV’. Royal Aeronautical
    Society Lecture delivered on 21 January 2009 at Coventry University Techno
Etkin, B. & Reid, L. D. (1996) 3rd edn. Dynamics of Flight. USA: John Wiley &
    Sons, Inc.
Garnell, P. & East, D.J. (1977) Guided Weapon Control Systems. UK: Pergamon
Jeti (2005) Jeti Model ADVANCE 90 plus [online] available from
    <> [Dec 2008]
Jung, D. ; Tsiotras, P. (2007) ‘Modelling and Hardware-in-the-Loop Simulation for
    Small Unmanned Arial Vehicle.' Proceedings from AIAA Infotech@Aerospace
    2007 Conference held May 7-10 2007 in California. AIAA 2007-2768, Available
    from <> [03 Feb
Kermode, A.C. (2006) 11th edn. ed. by Barnard, R.H. & Philpott, D.R. Mechanics of
    Flight. Harlow: Pearson Education Limited
Manai, M. ; Desbiens, A. (2005) 'Identification of a UAV and Design of a Hardware-
    in-the-Loop System for Nonlinear Control Purpose' Proceedings from AIAA
    Guidance, Navigation, and Control Conference and Exhibit held Aug 5-18 2007
    in California. AIAA-2005-6483, Available from
    pdf > [19 Feb 2009]
McLean, D. (1990) ed. by Grimble, M.J. Automatic Flight Control Systems. UK:
    Prentice Hall International Ltd.
MSN Encarta (2008) Unmanned Aerial Vehicle [online] available from
    ml> [27 Oct 2008]
NFTC or National Flying Laboratory Centre (2007) Flight Laboratory Course for
    Coventry University. Unpublished booklet. UK: Cranfield University.
Ogata, K. (2002) 4th edn. Modern Control Engineering. USA: Prentice-Hall, Inc.
Raptis, I. A. ; Valavanis, K. P. (2007) 'Airplane Basic Equations of Motion and
    Open-Loop Dynamics.' In Advances in Unmanned Arial Vehicles. ed. by
    Valavanis, K. P. The Netherlands: Springer: 49
Remub, V. ; Deeg, C. ; Musial, M. ; Hommel, G. ; Cuesta, F. ; Ollero, A. (2007)
    'Autonomous Helicopters.' In Multiple Heterogeneous Unmanned Arial Vehicles.
    ed. by Ollero, A. & Maza, I. Germany: Springer-Verlag: 111
UAS Research (2008) ‘UAS Yearbook 2008/2009’ [online] USA: University of North
    Dakota. Available from <> [11 Nov 2008]
US OSD (2005) ‘Unmanned Aircraft Systems Roadmap 2005-2030’ [online] USA:
    Office of Secretary of Defence. Available from
    <> [03 Mar 2009]


To top