Development of a miniature low cost UAV helicopter autopilot platform and formation flight control of UAV teams 1

Its unlimited potential for various practical implementationsmotivates our NUS UAV research team to carry out a comprehensive study and exploration.One of our goals is to develop a minia

FORMATION FLIGHT CONTROL OF UAV TEAMS

YUN BEN(B.Eng and M.Eng, Harbin Institute of Technology, China)

A THESIS SUBMITTEDFOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF ELECTRICAL & COMPUTER ENGINEERING

NATIONAL UNIVERSITY OF SINGAPORE

2010

Unmanned aerial vehicles (UAVs) have become a popular research topic worldwide recently.Their important role and their great potential are being explored continuously Amongvarious UAVs, small UAV helicopter has special attractiveness to the academic circle, due

to its small size, unique flight capacities, outstanding maneuverability, and low cost days with the fast development in manufacturing skill and martial science, constructing asmall-scale UAV helicopter which is upgraded from advanced hobby-purpose helicopter be-comes feasible and affordable Its unlimited potential for various practical implementationsmotivates our NUS UAV research team to carry out a comprehensive study and exploration.One of our goals is to develop a miniature UAV helicopter platform and construct severalUAVs as test beds towards implementing multiple UAV team formation flight With limitedresearch budget, low-cost miniature UAV helicopter was chosen as our research platform.Developing an autonomous miniature UAV helicopter is a challenge for several reasons.Firstly, helicopters are inherently unstable and capable of exhibiting high acceleration rates.They are highly sensitive to control inputs and require high frequency feedback with mini-mum delay for stability Secondly, the dynamics of helicopters are unstable, multivariable,highly coupled, which is thus hard for modeling Thirdly, helicopters have limited on-boardpower and payload capacity Flight control systems must be compact, efficient, and lightweight for effective on-board integration As the saying goes “the lower the price, the worsequality of the goods”, smaller and lower cost sensors also mean lower performance, whichmakes the things even harder Finally, helicopters are extremely dangerous

Nowa-Our research on miniature UAV helicopters started from the year 2003 The overalldesign and control procedure for the miniature UAV is similar as the small-scale UAVhelicopters, which includes: (1) hardware construction, (2) software system development,(3) dynamic modeling, and (4) control law design and implementation Besides these, with

a low-cost micro-size Inertial Measurement Unit (IMU) and a miniature UAV helicopter, weneed thoroughly calibrate sensors, optimize our Global Position System (GPS) informationand also self-develop attitude determination algorithm In spite of the difficulties, one ofthe advantages of choosing a miniature UAV helicopter as the platform lies in that theexperiment with a miniature helicopter is simple: it is easy to find a small area for testing

i

and thus can greatly reduce our previous research time This is the reason why we are able

to extend our research in the formation flight

The fundamental requirement for carrying our UAV research is having reliable miniatureUAV helicopters in hand In the first year, we have constructed and tested the miniatureUAV helicopter One compact with special design , for constructing a miniature UAV heli-copter platform with minimum payload and lowest cost has been summarized and proposed.After the autopilot system hardware was integrated, software was developed to ensure all ofthe hardware components work properly Several software necessaries such as sensor datacollection and filtering, control signal collection, and data storage were implemented on it.Since a dynamic model is needed before designing the controller, based on the classic mod-eling method, we first developed a linear multi-input multi-output (MIMO) model throughtime-domain and frequency-domain system identification To avoid flight crash due to tem-porary GPS signal block or failure, a special robust cascaded modeling and control methodwas proposed The autopilot system performs very well in hovering and low speed flying inreal flight experiment

With the reliable UAV helicopter platform at hand, we then further extended our search in the formation flight field We present a full scheme for the formation flight ofmultiple UAV helicopters Moreover, we adopt the leader-follower pattern to maintain afixed geometrical formation while navigating the UAVs following certain trajectories Morespecifically, the leader is commanded to fly on some predefined trajectories, and each fol-lower is controlled to maintain its position in formation using the measurement of its inertialposition and the information of the leader’s position and velocity, obtained through a wire-less Compact Flash (CF) card More specifications are made for multiple UAV formationflight In order to avoid possible collisions of UAV helicopters in the actual formation flighttest, a collision avoidance scheme based on some predefined alert zones and protected zones

re-is employed Simulations and real flight experimental results have verified that our design

is effective

To conclude this work, we summarize the research achievements and contribution, thenprovide the meaningful research directions for future

“The important thing in life is to have a great aim, and the determination to attain it”.Johan Wolfgang von Goethe, (German Poet and dramatist).

It is beyond doubt that the work in this thesis can not be completed without the support,advice, and encouragement of others, teachers, colleagues, friends, and family members Inthis acknowledgement, I wish to formally acknowledge their support and pay special thanks

to the role they have played during the course of this project

First and foremost, I would like to gratefully and sincerely thank Prof Ben M Chen forhis guidance, understanding, and patience in all aspects of this research during my studies

at National University of Singapore

I also owe a debt of gratitude to Dr Kai-Yew Lum and Prof T H Lee for their tance and counsel, and for the generous time and invaluable insights that they shared with

assis-me in reviewing my work, as well as for having served as my academic advisor throughoutthis endeavor

The members of the UAV group have contributed immensely to my personal and fessional time at NUS The group has been a source of friendships as well as good adviceand collaboration I have had the pleasure to work with all of them Special thanks to Mr.Alvin Cai, who has explored the avionics in his final year project and Dr Guowei Cai whohas explored the MIMO system modeling I am especially grateful for them: Dr KemaoPeng, Dr Miaobo Dong, Mr Feng Lin and Mr Xiangxu Dong

pro-To my family, for their reassuring confidence in the inevitable conclusion of this work,

in particular I would like to thank my mother for her unwavering faith in me, my fatherwho I deeply admire, and my brothers for their strong support, of which I’m truly grateful.Finally I would like to thank the two most important people in my life, my wife anddaughter You mean more to me than life itself Thank you for your love, support, encour-agement, and most importantly, laughter Thank you

iii

Summary i

1.1 General Overview 1

1.2 Project Background 3

1.3 Technical Background 5

1.3.1 Coordinate Frames Definition 6

1.3.2 Attitude Definition 8

1.3.3 Actuator Definitions 9

1.4 Goals and Objectives 10

1.5 Outline of this Thesis 11

iv

2 Systematic Construction of a Miniature UAV Platform 13

2.1 The UAV System Hardware 14

2.1.1 Bare Helicopter: TREX 450 14

2.1.2 Avionic System 16

2.1.3 Servo Control Module 23

2.1.4 Other Issues Related to the UAV Design 28

2.2 Software Development 29

2.2.1 Onboard Software 30

2.2.2 The Ground Station: a Monitoring Program 33

2.3 Ground and Flight Test Evaluation 35

2.3.1 EMI Test 35

2.3.2 Vibration Test 35

2.3.3 Manual Flight Test 36

2.4 conclusion 40

3 Robust GPS Enhancement and Attitude Determination 41 3.1 Sensors Calibration 42

3.1.1 Magnetometer Hard and Soft Iron Calibration 43

3.2 Typical Filtering Algorithms 45

3.2.1 Complimentary Filtering Algorithm 45

3.2.2 Kalman Filtering Algorithm 48

3.2.3 Extended Kalman filtering Algorithm 50

3.2.4 Discrete-time H∞ Filtering 52

3.2.5 Comparison of the Filtering Algorithms 54

3.3 Atttitude Determination 54

3.3.1 Determination of Aircraft Attitude 55

3.3.2 H∞ Filtering for Euler-angles Determination 58

3.3.3 Experimental Results 62

3.4 GPS Signal Enhancement 64

3.4.1 Position Signals 65

3.4.2 H∞ Filtering for GPS Position Signal Enhancement 69

3.4.3 Position Determination Experimental Results 69

3.5 Conclusion 72

4 Dynamic Modeling and Control of the Miniature UAV Helicopter 73

4.1 Data Collection and Preprocessing 74

4.1.1 Select the Input Signals 74

4.1.2 Collect Flight Test Data 76

4.1.3 Preprocessing of the Raw Dataset 80

4.2 Helicopter Aerodynamics Model Structure 81

4.2.1 6 Degree of Freedom (DOF) Rigid-body Dynamics 82

4.2.2 Coupled Rotor Flapping Dynamics 82

4.2.3 Yaw Rate Gyro Dynamics 83

4.3 MIMO Modeling and Control Method 84

4.3.1 Modeling of BabyLion UAV Helicopter 84

4.3.2 Control Law Design for BabyLion UAV Helicopter 91

4.4 Cascaded Modeling and Control Method 94

4.4.1 Cascaded Modeling 95

4.4.2 Control Law Design for BabyLion UAV Helicopter 101

4.5 Simulation and Implementation Results 104

4.5.1 MIMO Control Method 104

4.5.2 Cascaded Control Implementation Results 107

4.5.3 Comparison of the Two Control Methods 113

4.6 Conclusion 118

5 Formation Control Modeling, Control Law Design and Implementation 120 5.1 Model of a Leader-Follower Formation Flight 121

5.2 Control of Formation Flight 126

5.2.1 Dynamic Inversion Control Law for Outer-loop Control of the Follower UAV 127

5.2.2 Leader-Follower Formation Flight Input Constraint 128

5.2.3 Multiple UAV Formation Flight 129

5.3 Collision Avoidance 135

5.3.1 Collision Avoidance for Two UAV Case 136

5.3.2 Multiple UAV Group Collision Avoidance 142

5.3.3 Obstacle Avoidance 142

5.4 Formation Flight Test Result 145

5.4.1 Formation Flight Experiment Setup 146

5.4.2 Test Result 148

5.5 conclusion 151

6 Conclusions 1536.1 Contributions 1546.2 Future Work 155

1.1 The UAV helicopter, BabyLion 4

1.2 UAV helicopter family in NUS 4

1.3 Reference frames 7

1.4 The formation frame 8

1.5 Definition of normal Euler angles 9

2.1 Basic helicopter - TREX 450 15

2.2 The avionics system 16

2.3 Functional system architecture [31] 17

2.4 MNAV100CA 19

2.5 Stargate development platform (processor board and daughter card) 20

2.6 AmbiCom wireless 802.11 card 22

2.7 Auto mode switch on RC transmitter 25

2.8 PPM signal and its decoded output 25

2.9 MNAV mode switching function 26

2.10 A typical receiver circuit diagram with PPM decoding 26

2.11 A typical receiver print circuit board (PCB) with PPM decoding 27

2.12 UAV avionics 29

2.13 Soft model architecture [31] 31

2.14 Scheduling time line 32

2.15 Ground station GUI with NUS campus map 34

2.16 UAV in flight 34

2.17 Manual hovering flight test 40

3.1 Calibration example: accelerometer offset, scale factor compensation 42

viii

3.2 Effects of soft/ hard-iron distortions 44

3.3 Magnetic calibration readings 44

3.4 Rate gyro performance Solid: the true attitud; Dash: integrated output of the rate gyro 46

3.5 Accelerometer performance solid: the true attitude; Dash: the accelerometer estimation 46

3.6 Attitude angle estimation using accelerometer 47

3.7 Comparison of Euler angles estimated by H∞ filter (solid line) and those estimated by NAV420 (dashed line) 63

3.8 The measured body-frame y-axis position and velocity during a hovering flight from GPS 66

3.9 The position integrated by velocity (bold line) and that measured by GPS receiver (thin line) 67

3.10 Comparison of the x-y position estimated by H∞ filter (bold line) and that measured by NAV100CA (thin line) 70

3.11 Comparison of the position estimated by H∞filter (solid line) and that mea-sured by NAV100CA (dashed line) 71

4.1 Typical frequency sweep input signal 76

4.2 Typical doublet input signal 77

4.3 Input signals in the yaw channel perturbation experiment 78

4.4 Position outputs in the yaw channel perturbation experiment 78

4.5 Velocity outputs in the yaw channel perturbation experiment 79

4.6 Angular rates in the yaw channel perturbation experiment 79

4.7 Euler angles in the yaw channel perturbation experiment 80

4.8 An illustration for tip-path-plane 83

4.9 Verification of the identified model 90

4.10 General flight control scheme for UAVs 91

4.11 General flight control scheme for UAVs 98

4.12 Verification of the identified model 100

4.13 Outputs of hovering flight simulation 105

4.14 Control inputs of hovering flight simulation 106

4.15 Outputs of automatic hovering flight test 106

4.16 Control inputs of automatic hovering flight 107

4.17 Automatic hovering flight test with cascaded control 112

4.18 UAV in flight 113

4.19 Automatic circle path tracking flight test 117

5.1 TREX450-XL Babylion formation fleet 121

5.2 Leader-follower UAV formation geometry 122

5.3 The controller structure for the follower UAV 128

5.4 The lateral flight limitation of UAV 129

5.5 Horizontal level plane formation geometry for 3 UAVs 130

5.6 Multiple UAV group communication, global/local information is distributed by leader UAV 134

5.7 Formation basic shape: line 135

5.8 Formation transformation 135

5.9 Multiple UAV group maneuvers of formation, rejoin, reconfiguration, splitting and deformation 136

5.10 Alert zone and protected zone surrounding a UAV 137

5.11 The relative configuration in horizontal view, showing the protected zone and alert zone 138

5.12 Multiple UAV group with an intruder 142

5.13 UAV and an obstacle 143

5.14 UAV avoids an obstacle 145

5.15 UAV group avoids obstacles 146

5.16 Design overview of data flow among helicopters in a “leader-chaser” flight experiment 148

5.17 UAVs in formation flight, (Screen capture from ground video camera) 149

5.18 Flight result, (a) Horizontal plane Red – virtual leader’s trajectory; Blue – follower trajectory; Cyan – follower reference 150

5.19 Flight result, (b) Longitude and Latitude Dashed line – follower trajectory; Solid line – follower reference 150

5.20 XY plot of flight data for leader-follower formation flight 151

5.21 Tracking separation between the leader and follower UAVs 152

2.1 Specifications of TREX 450 helicopter compared with Raptor 90 helicopter 15

2.2 Specifications of MNAV100CA compared with NAV420CA 18

2.3 Specifications of wireless card: data rate and range 22

2.4 Power utilisation 23

4.1 Trim values for the tested flight conditions 81

4.2 State and input variables with the physical meanings 87

4.3 Estimated parameters through system identification, MIMO modeling method 88 4.4 Estimated parameters through first principle modeling approach 89 4.5 Estimated parameters through system identification, cascaded modeling method 99

xi

Latin variables

A State matrix in linearized model structure

B Input matrix in linearized model structure

Bbody→L The transformation matrix from the UAVs body frame to the leaders wind frame

BL→body The transformation matrix from the leaders wind frame to the UAVs body frame

Bb2n Velocity transformation matrix from body frame to NED frame

Bn2b Velocity transformation matrix from NED frame to body frame

BW L Velocity transformation matrix from follower’s body frame to leader’s body frame

bx bias in the GPS velocity information

C Output matrix in linearized model structure

f Forward error in the formation frame

fc Forward clearance in the formation frame

g The acceleration of gravity

h Heave error in the formation frame

hc Heave clearance in the formation frame

I Moment of inertia matrix

Ixx Rolling moment of inertia

Iyy Pitching moment of inertia

Izz Yawing moment of inertia

is Main shaft tilting angle

KI Integral gains of the embedded controller

KP Proportional gains of the embedded controller

Ka Proportional gain of amplifier circuit

Kcol Proportional gain of the main blade’s collective pitch change to collective pitch servo input

Ksb Contribution from stabilzer bar flapping to main blade’s cyclic pitch

Kped Proportional gain of the tail blade’s collective pitch change to tail rotor servo deflection

l Lateral error in the formation frame

lc Lateral clearance in the formation frame

Mb Moment vector

Mbx Body frame rolling moment component

xii

Mby Body frame pitching moment component

Mbz Body frame yawing moment component

Mhf Pitching moment generated by horizontal fin

Mmr Pitching moment generated by main rotor

Mu Pitching speed moment derivative

Mv Pitching speed moment derivative

m Mass of helicopter

mx magnetometer output in x-axis

my magnetometer output in y-axis

mz magnetometer output in z-axis

Oe The origin point of NED frame

OABC The origin point of body frame

Of The origin point of formation axis frame

Pb Position vector in body frame

Pn Position vector in NED frame

pxb Body frame x-axis position

pxn NED frame x-axis position

pyb Body frame y-axis position

pyn NED frame y-axis position

pzb Body frame z-axis position

pzn NED frame z-axis position

p∗

x The true position with reference to GPS position

p Body frame rolling angular velocity

q Body frame pitching angular velocity

R Main blade radius

r Body frame yawing angular velocity

rsb Stabilizer bar inner radius

T Main rotor thrust force

u Body frame x-axis velocity

ua Body frame x-axis velocity relative to the airmass

un NED frame x-axis velocity

uwind Body frame x-axis wind velocity

Va Velocity vector relative to the airmass

Vb Velocity vector in body frame

Vn Velocity vector in NED frame

Vtrim Trimmed flight speed in steady state

Vwind Velocity vector of wind

v Body frame y-axis velocity

va Body frame y-axis velocity relative to the airmass

vn NED frame y-axis velocity

vwind Body frame y-axis wind velocity

w Body frame z-axis velocity

wa Body frame z-axis velocity relative to the airmass

wn NED frame z-axis velocity

wwind Body frame z-axis wind velocity

Xa Body frame x-axis rotor spring derivative

XABC Body frame x-axis spring derivative

Xe NED frame x-axis rotor spring derivative

Xf formation frame x-axis spring derivative

XN ED NED frame x-axis spring derivative

Xsf magnetometer soft calibration x-axis ratio

Xof f magnetometer soft calibration x-axis bias

Xu Body frame x-axis speed derivative

YABC Body frame y-axis spring derivative

Yb Body frame y-axis rotor spring derivative

Ye Body frame y-axis rotor spring derivative

Yf formation frame y-axis spring derivative

YN ED NED frame y-axis spring derivative

Ysf magnetometer soft calibration y-axis ratio

Yof f magnetometer soft calibration y-axis bias

Yv Body frame y-axis speed derivative

Ytr Tail rotor thrust force

y Output vector in linearized model structure

ZABC Body frame z-axis spring derivative

Zcol Heave direction control derivative

Ze NED frame z-axis spring derivative

Zf formation frame z-axis spring derivative

ZN ED NED frame z-axis spring derivative

Zw Body frame z-axis speed derivative

Greek variables

φ Rolling angle in NED frame

θ Pitching angle in NED frame

θcol Collective pitch angle of main blade

θped Collective pitch angle of tail blade

ψ Yawing angle in NED frame

γsb Stabilizer bar rotor time constant

τ Main rotor time constant

Ω Main rotor rotating speed governed by engine governor

Ωb Angular velocity vector in body frame

Ωn Angular velocity vector in NED frame

δlat Aileron servo input

δlon Elevator servo input

δcol Collective pitch servo input

δped Rudder servo input

¯

δped Tail rotor servo (rudder servo) deflection

µ Advance ratio

Abbreviations

AGC Automatic Gain Control

AHRS Attitude and Heading Reference Systems

EKF Extended Kalman Filter

EMI Electromagnetic Interference

GUI Graphical User Interface

GPS Global Positioning System

IDENT Time-domain Identification Toolkit Integrated in MATLAB

IF Intermediate-Frequency

IMU Inertial Measurement Unit

INS Inertial navigation system

I/O Input Output

JTAG Joint Test Action Group

LQG Linear Quadratic Gaussian

LQR Linear Quadratic Regulator

Li-Po Lithium-Polymer

MAVs Miniature Aerial Vehicle

MEMS Micro-Electro-Mechanical System

MIMO Multi-Input Multi-Output

NED North-East-Down

NUS National University of Singapore

PCB Printed Circuit Board

PCMCIA Personal Computer Memory Card International Association

PID Proportional-Integral-Derivative

POSIX Portable Operating System Interface for UNIX

PPM Pulse-position Modulation

RC Radio-Controlled

RF Radio Frequency

RPM Rotations Per Minute

RPV Remotely Piloted Vehicle

RTOS Real-time Operating System

SISO Single-Input Single-Output

TCP Transmission Control Protocol

TPP Tip-Path-Plane

UAV Unmanned Aerial Vehicle

UDP User Datagram Protocol

USB Universal Serial Bus

2D Two-Dimensional

Unmanned aerial vehicles, or commonly known as UAVs, are autonomous flying vehiclesequipped with sensing devices and possibly weapons The UAVs have many potential mili-tary and civil applications They can avoid the human risk inherent to human-piloted aerialvehicles, particularly in missions in hostile environments, for example, reconnaissance overhostile territories, or attacking battle damage of enemy targets (see [18]) After World War

II, some countries began to develop their UAVs for military purposes With the rapid opment of high technology nowadays, UAVs have become very feasible lately due to recentadvances in sensor technologies, data processing hardware, and propulsion systems In thelast two decades, UAVs have attracted a significant interest More and more effort has beenput into this area

devel-Today there is a wide range of UAVs including rotor helicopters and fix-wing aircraftsused all over the world which are designed for different levels of performance depending

on their applications UAVs range in size from the man-portable (mini- and micro-UAVs),which are powered by electricity and usually weigh less than 2 kg, small-scale UAV which arepowered by oil or gasoline and typically weighs 10-50 kg, to full-size aircraft (large size UAV).Among various UAVs, the large UAVs are used mostly for military purposes In contrast, the

1

smaller, tactical UAVs are being developed to support tactical units to perform very shortrange “over the hill” and “around the corner” reconnaissance, and assist in protection Whileeach mission requires a different profile and capabilities, the man portable Miniature AerialVehicles (MAVs) are designed to provide reasonably good performance at an affordable price.Miniature UAV helicopter also has special attractiveness to the academic circle because oftheir smaller size, expendable lower cost, outstanding maneuverability, ease of uses, andwide range of capabilities.

The smaller size UAV helicopters are commonly upgraded from radio-controlled hobbyhelicopters by assembling an avionic system They are one of the best platforms applied

as typical plants for academic research as they can be docilely manipulated in a manualmode and can also be easily switched to an automatic mode Many research groups haveconstructed their own small scale UAV helicopters for their research purposes (see [27]).Success has been achieved in many research areas such as modeling and identification (see[43]), control techniques implementation (see [33]), aerial image processing (see [52], [42]),

to name a few However, due to the highly dynamic behavior of helicopter, limited payloadcapability and lack of highly accurate sensors, miniature UAVs are still rare

Designing a miniature UAV helicopter is a challenging job, there are many aspects thatneed to be thoroughly considered with the special constraints, such as:

1 hardware components selection which is limited by the budget and also the payloadcapability;

2 modeling and control of a nonlinear Multiple input Multiple Output(MIMO) copter system which is the core issues of the research purpose;

heli-3 software design which is usually programmed in an embedded system to collectsensor data, send sensor information to the ground station, calculate command signal, andissue the command to the actuators The ground station receives remote data from the on-board system for monitoring purposes Other software capability includes sensor informationprocessing such as sensor filtering and optimization

While a single autonomous UAV can be very useful in performing various tasks, multipleUAVs operating as a team to accomplish a given task cooperatively may offer even greateradvantages in certain applications, such as in target search and detection in a large area ofcoverage With a mini-UAV platform being built, it is natural for us to expand our researchdomain to the formation flight of multiple UAVs, as formation flight is one of the interestingtopics of further research in the UAV field [29].

In the next sections of this chapter, a general project background on the work completed

by our NUS UAV-research team in the last five years will be addressed in Section 1.2

In Section 1.3, a brief technical background of the designing a UAV helicopter will beintroduced The objective and goal of the project is presented in Section 1.4 and an outline

of the remaining chapters will also be listed in Section 1.5

The research on small-scale UAV helicopters in National University of Singapore (NUS) hasstarted from year 2004 During the last five years, our NUS UAV research team has success-fully constructed multiple small-scale UAV helicopters, developed efficient UAV softwaresystems, identified the high-fidelity linear and nonlinear dynamic models, and implementedlinear and nonlinear control laws to realize fully autonomous flight [10]

Since 2004, we have first developed a small-scale UAV helicopter, called HeLion, which

is upgraded from fuel-powered Raptor 90 helicopter (with 1.4 m rotor span), by equippingwith custom designed onboard computer system [10] with a comprehensive onboard andground station software system [22] The bare helicopter weighs 4.85 kg and approximately

13 kg when mounted with the avionics system An accurate aerodynamics model has beenderived for flight control law design and high performance automatic flight has been realized.The same avionics system is also successfully applied for a relative bigger size UAV, calledHenglion, which is upgraded from a Copterworks helicopter with the weight of 20kg Mean-while, based on these achievements, we extended our research interest to develop mini-size

Figure 1.1: The UAV helicopter, BabyLion.

Figure 1.2: UAV helicopter family in NUS

UAV based on smaller helicopters An ultra low cost mini hobby RC helicopter, namedTREX 450 with 0.7 m rotor span, 700g weight is chosen for this purpose Following the pre-vious version of UAVs, we named this mini UAV as “Babylion”, which is shown in Fig 1.1.The collection of the UAVs developed by our NUS UAV group is shown in Fig 1.2

The major difference of BabyLion to Helion is that the size is scaled down by a

math-ematical scale in both weight and cost Because of the smaller scale, the helicopter is nowbattery-powered and very much lighter The full flight time of BabyLion is about 8 minutes.

In the point of view of practical implementation, mini UAV helicopter has some unique vantages compared with the larger counterparts, including: 1) more agility; 2) portable byone person, much easier and faster to assemble and move; 3) less noise; 4) lower cost (about1/10 of the cost of the larger size UAV); 5) less dangerous during flight, and 6) more suitablefor forest, urban searching, and indoor flight Besides this, BabyLion may also be an idealtest bed for further research and technologies like vision based control and formation flying

ad-or even swarming

However due to its smaller size, extremely strict payload, and more sensitive namics, constructing a mini UAV helicopter is much more difficult and challenging Thedisadvantages includes: 1) rapid dynamics with helicopter, which makes it hard for model-ing; 2) low-cost small-size sensors with poorer performance; 3) actuator with relative lowerresolution; 4) working close to the GPS noise level such as the velocity is around: 0.5m/swhile the GPS velocity resolution is only 0.1-0.3 m/s, and position error is around 1.5 m,while the GPS resolution is about 3 m As such, most of research groups are still keen

aerody-on the Raptor 90 size (with 1.6 m rotor span, 5kg payload) or larger scale UAV helicopterplatforms

Fortunately during the construction of HeLion UAV helicopter we have obtained a lot

of experience, which enables us to conquer various problems and successfully build up aminiature UAV helicopter platform The detailed hardware integration, sensor optimizationand control design theory will be discussed in Chapter 2 to 4

In this section, we introduce some technical background, including cordinate frames, attitudedefinition and actuator definition

1.3.1 Coordinate Frames Definition

Three axis systems are utilized in both the modeling and the control of the control system.These are the Earth axis system, the helicopter body axis systems and the formation axissystem

Earth axis system (North East Down (NED) frame)

1 The Earth is assumed to be flat

2 An arbitrary point on the Earths surface is defined as the origin (Oe)

3 A right handed orthogonal system of axes (Oe, Xe, Ye, Ze) is defined at the origin.Where Xe points due North, Ye due East and Ze vertically down

The NED frame is shown in Fig 1.3

Aircraft body axis systems

Each aircraft has its own right handed orthogonal axis system (OABC, XABC, YABC,

ZABC) mapped onto it

1 The origin (OABC) is located at the center of gravity of the aircraft and is constrained

to move with it

2 XABC is defined to point out through the aircrafts nose

3 YABC out through the right hand wing

4 ZABC out through the bottom of the aircraft

The axis system is constrained to move with the aircraft as it rolled, pitched, and yawed.The body frame is also shown in Fig 1.3

Formation axis system

The right handed orthogonal formation axis system (Of, Xf, Yf, Zf) is used to specifythe offset of the target formation point from the lead aircraft

1 The axis systems origin (Of) is located with the origin of the lead aircraft‘s bodyaxis system

Figure 1.3: Reference frames.

Figure 1.4: The formation frame.

2 Xf is defined to be parallel to the Earth‘s surface and to point in the compass headingdirection of the lead aircraft

3 Yf completes the right hand orthogonal set

4 Zf is defined to point vertically down towards the Earth

The X − Y plane of the formation axis system therefore always remains parallel to thesurface of the Earth, with the Xf axis aligned to the compass heading of the lead helicopter.This is important because if the formation axis system rolled and pitched with the leadhelicopter in the same way as the body axis system, in low level flight with a reasonablyspaced formation, the trail aircraft could be forced into the ground by the rolling or pitchingmotion of the lead [39] The Formation frame is shown in Fig 1.4

1.3.2 Attitude Definition

A set of the so-called normal Euler angles are commonly used to describe the orientation

of an aircraft There are three Euler angles, which include the heading or yaw angle (ψ),pitch angle (θ), and roll angle (φ) These angles are referenced to the local horizontal plane

Figure 1.5: Definition of normal Euler angles.

which is perpendicular to the earth’s gravitational vector as depicted in Fig 1.5 Heading

is defined as the angle in the local horizontal plane measured clockwise from a true north(earth’s polar axis) direction Pitch is defined as the angle between the aircraft’s longitudinalaxis and the local horizontal plane (positive for nose up) Roll is defined as the angle aboutthe longitudinal axis between the local horizontal plane and the actual flight orientation(positive for right wing down) (see [13, 14]) We note that in Fig 1.5, (XNED,YNED,ZNED)and (XABC,YABC,ZABC) represent the north-east-downward coordinate and the aircraftbody coordinate respectively

1.3.3 Actuator Definitions

The Rotor Velocity is given by the angular velocity of the rotor multiplied by its radius.Normally a helicopter has a fixed angular velocity of the main and tail rotor Lift thrust isthen varied by increasing or decreasing the collective pitch The main rotor is powered by

a DC motor

The Collective pitch is the pitch of the main rotor blades which make the helicopterperform a ’lift-down’ motion On a helicopter, lift thrust is normally controlled by pitchingall the main rotor blades equally The amount of lift generated is determined by the pitchangle of each rotor blade as it moves through the air The lift thrust is controlled by directlyvarying the angular velocity of the main rotor [25]

The Cyclic pitch is the pitch of the main rotor blades which make the helicopter perform

a roll or pitch motion The lift thrust vector can be tilted, such that it is not parallel withthe main rotor shaft axis This is done by tilting the swash plate to change the pitch angle

of the rotor blade The lift generated by the rotor blade changes and the helicopter will tilttowards the side which is experiencing lesser amount of lift Two inputs are necessary tocontrol cyclic pitch namely lateral pitch and longitudinal pitch, both of which are controlled

by one servo respectively By combining these two inputs it is possible to tilt the lift thrustvector towards any direction

Lateral pitch is the pitch of the main rotor blades which makes the helicopter perform

a roll motion

Longitudinal pitch is the pitch of the main rotor blades which makes the helicopter flyforward or backwards, i.e., pitch motion

Lateral plane is a plane parallel to the YABC− ZABC plane

Longitudinal plane is a plane parallel to the XABC− ZABC plane

The goal of this project is first to build a miniature UAV platform The appropriate partswhich range from the IMU, on-board computer to power systems have to be chosen andintegrated On board software and ground station monitoring program should be imple-mented

The next is to develop an attitude determination algorithm using the IMU inertial puts such as accelerometer, gyroscope, magnetometer, etc, as low-cost IMU usually does notprovide attitude reference GPS position enhancement algorithm should also be developedfor the raw GPS position resolution is 3m only, which is insufficient for a short distanceflight less than 20 m

out-Another goal is to make system identification for the assembled UAV so that a goodcontrol algorithm can be designed for autonomous flight A simple controller will be designedusing the Linear Quadratic Regulator (LQR) method and tested in an autonomous hoverand trajectory tracking.

Finally, the last objective is to develop formation flight basic theory, design formationflight control, and implement formation experiment based on a leader-follower mode

1.5 Outline of this Thesis

The remaining content of this thesis is divided into five chapters All of the key works for oursmall-scale UAV family are presented in detail, including: (1) platform design and construc-tion, software system introduction; (2) attitude determination and navigation enhancement;(3) dynamic modeling, flight control law design, and implementation; (4) formation flighttheory and control; (5) conclusion and future research

Chapter 2 presents the development of the hardware and software for the miniature UAVhelicopter platform The hardware construction for the flight control system involves theintegration of the processor, sensor, wireless communication card, and power supply battery.The software development for the project introduces the software running in the embeddedonboard computer system to collect this data, implement flight control laws, drive actuators,communicating with the ground station, and as well as a ground supporting system thatmonitors the UAV flight status and also issues command to the UAV helicopter

Chapter 3 presents an integration scheme of a low-cost inertial attitude and positionreference system for a mini unmanned helicopter by utilizing the robust H∞ filtering tech-nique A comparison of the common filter algorithms is first introduced Next, the H∞

filtering algorithm is proposed and the model for the H∞ filtering algorithm is deduced.The application results are presented at last

Chapter 4 focuses on the dynamic modeling and control for our miniature UAV copter We first introduce a MIMO modeling and control method, then propose a robustcascaded modeling and control method The structure determination, parameter identifica-tion, and model validation, are also introduced The corresponding linear control law design

heli-is then given Implementation experiment in real flight heli-is also shown

In Chapter 5, we focus on the formation flight control law design and implementation

We adopt the leader-follower pattern to maintain a fixed geometrical formation while igating the UAVs following certain trajectories More specifications are made for multipleUAV formation flight A collision avoidance scheme based on some predefined alert zonesand protected zones is also proposed Simulations and real flight experimental results arefinally shown in detail

nav-Finally, Chapter 6 presents the conclusion of this project The contribution of this work

is presented, and the focus of future research is discussed

Systematic Construction of a

Miniature UAV Platform

Generally speaking, a basic UAV autopilot system should include the following components

or software programs:

a) sensors to measure the attitudes of the aircraft;

b) an embedded computer system;

c) communications system to provide wireless communication with the ground station;d) power supply system;

e) software running in the computer system to collect this data, implement flight controllaws, drive actuators, and communicate with ground system An automatic flight controlalgorithm based on the flight control law is programmed to calculate the actuators’ inputs;f) A ground supporting system which provides wireless communication, schedule flightcourses and collect in-flight data

In this chapter, we present a comprehensive UAV design and integration method ing building the autopilot system hardware with all necessary components from a) to d) inthe list above, as well as the software running in the embedded system and the ground stationnotebook as e) and f) in the list above The autopilot will be mounted on a radio-controlledhelicopter that is commercially available in the market

includ-13

2.1 The UAV System Hardware

There is often a trade-off for the design of the avionics hardware and the design requirements

An optimal design solution is only the compromise solution The major issues for thedesigning a mini-UAV include (1) bare helicopter performance; (2) hardware componentsselection and integration; (3) onboard system layout design; (4) power consumption design;(5) electro-magnetic-interference (EMI) shielding; (6) anti-vibration Although some small-scale UAV helicopter platforms have been successfully built up based on RC helicopters, nouniform, time-saving, and effective design methodology has been clearly documented in theliterature

In this section, the proposed UAV autopilot will be described in detail In general, itincludes four key steps: (1) bare helicopter selection; (2) hardware component selection; (3)integration to the RC helicopter system via a servo control module; and (4) other relatedissues, such as the supporting structure and physical design, etc

2.1.1 Bare Helicopter: TREX 450

The choice of aircraft is primarily based on available model helicopters in the Remote Control(RC) helicopter market We already have a Rapter 90 helicopter which is powered by gasengine However, the operating cost for it is too high and the consumable accessories arealso expensive We choose the TREX 450 helicopter which is powered by electrical engine(Fig 2.1 ) Five onboard digital servo actuators are used to drive the helicopter Morespecifically, the aileron, elevator, and collective pitch servos are in charge of tilting the swashplate to realize the rolling motion, pitching motion, and to change the main rotor’s collectivepitch angle The throttle servo, cooperated with a hobby-purpose Rotations Per Minute(RPM) governor, is used to control the engine power One high-speed servo, associated by alow cost yaw rate gyro, is employed to control the yaw motion Like the other brands of RChobby helicopters, TREX 450 is equipped with the stabilizer bar, which acts as a damper

Table 2.1: Specifications of TREX 450 helicopter compared with Raptor 90 helicopter

Specifications TREX 450 Raptor 90Full Length of Fuselage 685 mm 1410 mmFull Width of Fuselage 162 mm 190 mmFull Height of Fuselage 228 mm 476 mmMain Rotor Diameter 680 mm 1605 mmTail Rotor Diameter 150 mm 260 mm

Maximum Flight Time 12 minutes 12 minutes

Figure 2.1: Basic helicopter - TREX 450

to reduce the over-sensitive aerodynamic forces caused by the ultra small size of helicopterand facilitate manual control The reason to choose TREX 450 helicopter to replace Raptor

90 is mainly because of its lighter weight and compatible performance Moreover, the flightexperiment operation cost is relatively low The replacements for it are by far the easiest

to buy at the least expense which is about 1/10 of the cost for the Rapter 90 The mainfeatures of the TREX 450 helicopter compared with Raptor 90 are listed in Table 2.1:From the above table, we can see the payload for the TREX 450 is only 0.8 kg, whichrestricts our hardware selection for the avionics system

Figure 2.2: The avionics system.

2.1.2 Avionic System

This section briefly explains the characteristics of each component comprising the designedUAV system The avionics system comprising of the MNAV100CA, Stargate Processorboard and Wireless CF Card is shown in Fig 2.2 This system is mounted under themain rotor near the centre of gravity and fastened to the landing gear The power systemcomprises of two Lithium Polymer batteries with 11.1V and 7.4V respectively The avionics

is powered by the 7.4 V battery after a regulator of 5 V

Functional system architecture is shown in Fig 2.3

Figure 2.3: Functional system architecture [31].

two resides in the NED frame It is noted that the three-axis Euler angles are notnecessary since they can be estimated by using an Extended Kalman filter (EKF) asreported in [31] and H∞ filtering reported in [60]

2 The measuring ranges of the axis acceleration, axis angular rate and axis magnetic are set as ±2 g, ±150◦

three-and ±0.7 Gauss, respectively, in accordancewith the specifications of the commonly used commercial products The selectedthreshold values are reasonable since we do not intend to cover the extreme or acrobaticflight conditions As a result, the acceleration, angular rate, and magnetic sensors

do not change dramatically during flight tests Based on this setting, we need also

to carefully perform an anti-vibration design to avoid the measurement saturationcaused by various vibration sources associated with the UAV This is to be addressed

in Section 2.1.4

3 On the basis of meeting all of above mentioned requirements, the size, weight andpower consumption of the adopted INS/GPS should be minimized

Table 2.2: Specifications of MNAV100CA compared with NAV420CA

A compact INS/GPS, namely, MNAV100CA, shown in Fig 2.4 along with the virtual

counterpart, is selected for BabyLion The MNAV100CA [20], is a calibrated digital sensor

and servo control system designed for use in RC vehicles The MNAV100CA utilizes an

ATmega128L microcontroller to manage two 8-channel 16-bit A/D (analog to digital)

con-verters, GPS receiver, 9 servo interfaces, also the PPM (Pulse-position modulation) input

interface and sensor calibration in the internal EEPROM The advantage of this is that

because this product is specifically made for miniature robotic platforms The two key

re-quirements: navigation and servo control are satisfied in one package thus saving space and

weight compared to having a separate GPS aided inertial system and servo control board

which are normally used by the bigger UAV systems (for example, the Helion UAV)

The key specifications of this sensor are listed in Table 2.2, which clearly shows that

all of the requirements are satisfied Furthermore, by using MNAV100CA, the weight and

power consumption of the INS/GPS sensor is greatly reduced, as compared to those of

the fully integrated INS/GPS, NAV420CA, installed on HeLion It is to be verified in

Section 2.3 by various flight tests that the compact INS/GPS yields the similar level of

working performance as the expensive NAV420 adopted in HeLion

Figure 2.4: MNAV100CA.

From the Table 2.4, we can see that the MNAV100CA meets the requirements except

it does not provide attitude information This means we have to develop attitude mation algorithm Meanwhile, compared with NAV420CA, the MNAV100CA has a lowerperformance but much smaller size and lighter in weight

esti-The MNAV plugs directly through a 51 pin serial connector to the main processor board.The MNAV is from the same family of products as the processor board hence there is no needfor separate serial communications or data acquisition board for the two to communicate.This link allows the Stargate to give commands to the MNAV and the MNAV to send sensordata to the Stargate for processing The Stargate also provides DC (Direct Current) power

to the MNAV through this link This direct interface reduces the connection requirements

of wires which can carry vibration to the MNAV

The onboard sensor package in Fig 2.4 includes accelerometers, angular rate sensors,magnetometers, static and dynamic pressure sensors, and a GPS sensor The servo controlmodule includes both RC servo hardware and an RC receiver PPM interface Up to 9 RCservos can be directly connected to the MNAV with the option to use a separate powersource

Figure 2.5: Stargate development platform (processor board and daughter card).Stargate Processor Board

In our Helion UAV, we use a PC-104 ATHENA running at 600 MHz with the size of 96 mm

× 90 mm × 10 mm, and 200g weight However, due to the size and weight budget, we have

to choose a lighter embedded computer, named Stargate, as the on board processor.The Stargate [21], shown in Fig 2.5, is a single board computer with enhanced com-munications and sensor signal processing capabilities The Stargate uses Intel’s 400MHzX-scale processor and runs an embedded Tiny Linux as its operating system (OS) TheStargate is connected to the MNAV via the standard 51-pin connector and takes digitalsensor outputs from the MNAV

The Stargate main processor comes with a compact flash and PCMCIA (Personal puter Memory Card International Association) connection However, with the Stargatedaughter card which can be easily attached to the main processor via a standard 51 pinconnection, additional interfaces like Ethernet, Serial, JTAG (Joint Test Action Group) andUSB (Universal Serial Bus) Connectors will be available The avionics system does notrequire these connections whilst flying, and hence the daughter card has not been included

Com-in the avionics system for the purpose to save space and weight

A real-time software architecture is implemented on the Stargate and hosts multipletasks which are called, scheduled, and finished within a predictable time limit, using thePOSIX (Portable Operating System Interface for UNIX) thread standard, synchronizationprimitives, and Round-Robin (RR) scheduling In order for the system to remain real timeand deadlines to be kept, the application software can only run at a maximum rate of 50Hz.

A set of software is created by Jung Soon Jang from Crossbow technologies for fix-wing UAVand is open source [21] More codes can be easily added for various autonomous platformsincluding UAV helicopter

Although inferior to PC104 ATHENA Computer, the Stargate performance is still ficient for this project and is chosen for its convenience and ease of use In order to provide

suf-a short to medium communicsuf-ation rsuf-ange with the ground stsuf-ation, the Stsuf-argsuf-ate uses suf-a highgain wireless 802.11b card which is plugged into the CF Card slot The card chosen forthis purpose is the Ambicom Wireless CF Card as the Stargate is preloaded with the sup-ported driver, and has good communication range (up to 500m at 1Mbps) with low powerrequirements

For data logging, the Stargate has 32MB worth of flash memory of which only 10MB isused by the Linux software This local memory is used for all flight data

Wireless Communication Card

To provide a short-to-medium range and high speed remote access link to the Stargate, theadvanced Stargate kit ships with the AmbiCom Wave2Net wireless 802.11b card This card

is plugged into the available CF slot using the PCMCIA adaptor module

The Stargate is preconfigured to recognize this card and to automatically load therequired device drivers However, some additional configuration is required to allow thiscard to join a network using the ad-hoc or access point based methods There is a filenamed “wireless.opts” located in the ”/etc/pcmcia” folder that allows users to specify the

Figure 2.6: AmbiCom wireless 802.11 card.

Table 2.3: Specifications of wireless card: data rate and range

Range Data Rate

Table 2.4: Power utilisationComponents Voltage (V) Current (A) Power consumptionMNAV 3.7 to 16 VDC 0.16 A at 5 VDC ≤ 0.8W at5V DC

by Crossbow that the applied voltage does not exceed 6V as it may overheat Hence theallowable voltage range is between 3.7V and 5V A 5V regulator is used to transfer 7.4V to5V as the power supply for the Stargate With a 5V regulator, a 7.4V battery with 1200mAhcapacity would last a maximum of 120 minutes Both batteries are attached together andslung into the battery carriage compartment of the TEX450

The power that the avionics system consumes can be quantified in Table 2.4

Prior experimentation as well as online reports state that at the end of a Li-Po batterylife, the voltage of the battery will drop suddenly If it falls below a certain point, theMNAV firmware can malfunction Furthermore, the PPM functionality is implemented viasoftware and if the MNAV fails, the pilot would not be able to regain manual control This

is potentially dangerous for both human bystanders and the UAV hardware so batteriesshould be charged before each flight and not be used for periods exceeding 100 minutes

2.1.3 Servo Control Module

With the sensors collecting the attitude, velocity, and position information of the helicopterand an embedded processer working as the brain, the final part of the flight control is theactuation of the servomotors installed on the helicopter Radio-controlled helicopters aretypically equipped with two pair servomotors that actuate the main rotor swash plate, thetail collective pitch servo, and the engine throttle Usually the engine control and the tail