• No results found

Quad Copter

N/A
N/A
Protected

Academic year: 2021

Share "Quad Copter"

Copied!
53
0
0

Loading.... (view fulltext now)

Full text

(1)

Quadrot

Quadrot

or UAV

or UAV

August 30

August 30

2009

2009

 This paper describes a four rotor ver

 This paper describes a four rotor vertical-take-off-and- tical-take-off-and-landing unmanned aerial vehicle, commonly known as a landing unmanned aerial vehicle, commonly known as a Quadrotor, designed for objective completion in an urban Quadrotor, designed for objective completion in an urban environment.

environment. The assumThe assumptions, proposed ptions, proposed design anddesign and project methodology is described in detail.

project methodology is described in detail.

Proposal

Proposal

(2)
(3)

 Table of Contents

 Table of Contents

LLISTIST OFOF FFIGURESIGURES...4...4

A

ABSTRACTBSTRACT...4...4

I.

I. INTINTRODRODUCTUCTIONION...55

G

GOALOAL...5...5

P

PERSONALERSONAL SS TATEMENT TATEMENT OFOF IINTERESTNTEREST...5...5

U

UNMANNEDNMANNEDAAERIALERIAL VVEHICLESEHICLES...5...5

P

PREVIOUSREVIOUS RRESEARCHESEARCH WITHWITH UAVUAVSS...6...6

 T

 THEHE IINTERNATIONALNTERNATIONAL AAERIALERIALRROBOTICSOBOTICSCCOMPETITIONOMPETITION...7...7

 Abstract...7  Abstract...7

Notional

Notional Mission...Mission...7...7

P

PAPERAPER OORGANIZATIONRGANIZATION...9...9

II.

II. STAGSTAGES ES OF OF DESDESIGN IGN AND AND CONSCONSTRUCTRUCTION.TION...1...100

S

S TAGES TAGES OFOF TTHESISHESIS WWORK ORK ...10...10

Research...10 Research...10 Inertial

Inertial Navigation System...Navigation System...1111 Manual Flight...

Manual Flight... ...1111 Flight

Flight Model Model for for Stabilization.Stabilization...11...11 Ground Station

Ground Station Communication...Communication...11...11 Preplanned

Preplanned Flight...Flight... 1111 Collision Sensing Array...11 Collision Sensing Array...11 Object

Object Identification...Identification...12...12 Object

Object Acquisition...Acquisition... ..12..12 Objective

Objective Planning...Planning... 1212

S

S TAGES TAGES OFOF QQUADROTORUADROTOR DDESIGNESIGN...12...12

M

MODULARODULAR DESIGNDESIGN...13...13

Naming

Naming convention...convention... 1414 III.

III. DESDESIGN IGN CONSCONSIDERAIDERATIONSTIONS...15..15

IARC R

IARC REQUIREMENTSEQUIREMENTS...15...15

 J

 JUSTIFICATIONUSTIFICATION OFOF QQUADROTORUADROTOR...16...16

IV.

IV. QUADQUADROTOR ROTOR BASICBASICS...S...17...17 V.

V. QUADQUADROTOR ROTOR COMPCOMPONENTONENTS....S...1...199

 T

 TRANSMITTERRANSMITTER/R/RECEIVERECEIVER...20...20

R

ROTOROTOR SS YSTEM YSTEM...21...21

DC

DC Motors...Motors...2121 Contra Rotating

Contra Rotating Blades...Blades... 2222 Electronic Speed

Electronic Speed Controllers...Controllers...22...22

B

BATTERYATTERY...23...23

M

MICROCONTROLLERICROCONTROLLER PPLATFORMLATFORM...24...24

S

(4)

U

USERSER IINTERFACENTERFACE...31...31

IDE...31

IDE...31

Live Display...31

Live Display...31

P PROGRAMMINGROGRAMMING IISSUESSSUES...32...32

 JAUS T  JAUS TERMINALERMINAL...32...32

VII. VII. OVERVOVERVIEW OF TESTIEW OF TESTING EQUIPING EQUIPMENT AND TOMENT AND TOOLS...OLS...33...33

 T  TESTINGESTING EEQUIPMENTQUIPMENT...33...33

Motor Motor Thrust Thrust Test-Bench...Test-Bench...33...33

Power Management... Power Management...3535 Tethered Flight Rig...36

Tethered Flight Rig...36

 T  TOOLSOOLS...36...36

VIII. VIII. QUADROTOR QUADROTOR DESIGN DESIGN AND AND CONSTRUCTION...CONSTRUCTION...37....37

S S TAGE TAGE1: M1: MANUALANUAL QQUADROTORUADROTOR FFLIGHTLIGHT...37...37

Procedure of Stage 1...37

Procedure of Stage 1...37

Construction of the Quadrotor...37

Construction of the Quadrotor...37

Testing...38

Testing...38

Verification of Stage Completion...38

Verification of Stage Completion...38

S S TAGE TAGE2: P2: PREPLANNEDREPLANNEDFFLIGHTLIGHT PPATHATH...39...39

Procedure of Stage 2...39

Procedure of Stage 2...39

 Adding the XBee communication system...39

 Adding the XBee communication system...39

Flight Controller... Flight Controller... ...3939 Testing...39

Testing...39

Verification of Stage Completion...40

Verification of Stage Completion...40

S S TAGE TAGE3: O3: OBSTACLEBSTACLE AAVOIDANCEVOIDANCE...41...41

Stage Stage Considerations...Considerations...41...41

Verification of Stage Completion...41

Verification of Stage Completion...41

S S TAGE TAGE4: O4: OBJECTIVEBJECTIVE PPLANNINGLANNING...42...42

Stage Stage Considerations...Considerations...42...42

Verification of Stage Completion...42

Verification of Stage Completion...42

IX. IX. FINANFINANCES.CES...43....43

S S TAGE TAGE 1...431...43

S S TAGE TAGE 2...442...44

S S TAGE TAGE 33 ANDAND4...4...44...44

X. X. TIMETIMELINE.LINE...46...46

XI. XI. CONCCONCLUSIOLUSION...N...47...47 XII.

XII. WORKS WORKS CITECITED....D...4...488

Error: Reference source not found Error: Reference source not found

(5)

List of Figures

List of Figures

FIGURE 1: ELECTRONIC COMMUNIQUÉ

FIGURE 1: ELECTRONIC COMMUNIQUÉ...8...8 FIGURE 2: STAGES OF RESEARCH

FIGURE 2: STAGES OF RESEARCH...10...10 FIGURE 3: QUADROTOR STAGES

FIGURE 3: QUADROTOR STAGES...12...12 FIGURE 4: MODULAR DESIGN

FIGURE 4: MODULAR DESIGN...13...13 FIGURE 5: NAMING CONVENTION

FIGURE 5: NAMING CONVENTION...14...14 FIGURE 6: QUADROTOR BODY DIAGRAM

FIGURE 6: QUADROTOR BODY DIAGRAM...17...17 FIGURE 7: QUADROTOR RENDERING

FIGURE 7: QUADROTOR RENDERING...19...19 FIGURE 8: SPEKTRUM DX6I (14)

FIGURE 8: SPEKTRUM DX6I (14)...20...20 FIGURE 9: TOWERPRO BRUSHLESS OUTRUNNER 2410-09 (17)

FIGURE 9: TOWERPRO BRUSHLESS OUTRUNNER 2410-09 (17)...21...21 FIGURE 10: CONTRA ROTATING EPP1045 (15)

FIGURE 10: CONTRA ROTATING EPP1045 (15)...22...22 FIGURE 11: TURNIGY PLUSH 18A SC (19)

FIGURE 11: TURNIGY PLUSH 18A SC (19)...22...22 FIGURE 12: TURNIGY 4000MAH (19)

FIGURE 12: TURNIGY 4000MAH (19)...23...23 FIGURE 13: ARDUINO MEGA (21)

FIGURE 13: ARDUINO MEGA (21)...24...24 FIGURE 14: IMU 6 DEGREES OF

FIGURE 14: IMU 6 DEGREES OF FREEDOM - V4 WITH BLUETOOTH CAPABILITY (22)FREEDOM - V4 WITH BLUETOOTH CAPABILITY (22)...25...25 FIGURE 15: IMU 5 DEGREES OF

FIGURE 15: IMU 5 DEGREES OF FREEDOM (23)FREEDOM (23)...26...26 FIGURE 16: GYRO BREAKOUT BOARD -

FIGURE 16: GYRO BREAKOUT BOARD - DUAL 500 DEGREE/SEC (24)DUAL 500 DEGREE/SEC (24)...26...26 FIGURE 17: COMPASS MODULE WITH TILT COMPENSATION - HMC6343 (25)

FIGURE 17: COMPASS MODULE WITH TILT COMPENSATION - HMC6343 (25)...26...26 FIGURE 18: MAXBOTIC LV-EZ1 (26)

FIGURE 18: MAXBOTIC LV-EZ1 (26)...27...27 FIGURE 19: SHARP GP2Y0A02YK0F (27)

FIGURE 19: SHARP GP2Y0A02YK0F (27)...28...28 FIGURE 20: HOKUYO URG-04LX (29)

FIGURE 20: HOKUYO URG-04LX (29)... ...28...28 FIGURE 21: CARBON FIBER SQUARE TUBES (30)

FIGURE 21: CARBON FIBER SQUARE TUBES (30)...29...29 FIGURE 22: DATA TRANSMISSION

FIGURE 22: DATA TRANSMISSION...30...30 FIGURE 23: XBEE PRO 50MW SERIES 2.5 RPSMA (33)

FIGURE 23: XBEE PRO 50MW SERIES 2.5 RPSMA (33)...31...31 FIGURE 24: XBEE EXPLORER USB (36)

FIGURE 24: XBEE EXPLORER USB (36)...32...32 FIGURE 25: MOTOR TEST BENCH

FIGURE 25: MOTOR TEST BENCH...33....33 FIGURE 26: FC22 COMPRESSION LOAD CELL (38)

FIGURE 26: FC22 COMPRESSION LOAD CELL (38)...34...34 FIGURE 27: MAXPRO BATTERY MONITOR 3S (39)

FIGURE 27: MAXPRO BATTERY MONITOR 3S (39)...35...35 FIGURE 28: TETHERED RIG RENDER

FIGURE 28: TETHERED RIG RENDER...36...36 FIGURE 29: STAGE 1 COSTS

FIGURE 29: STAGE 1 COSTS...43...43 FIGURE 30: STAGE 2 COSTS

FIGURE 30: STAGE 2 COSTS...44...44 FIGURE 31: PROPOSED TIMELINE

FIGURE 31: PROPOSED TIMELINE... ...46...46 FFIGUREIGURE32: P32: PROPOSEDROPOSEDTTIMELINEIMELINE

Abstract

Abstract

(6)

urban environm

urban environment. ent. The assumptionThe assumptions, proposed design as, proposed design and project methodnd project methodology isology is described in detail.

described in detail.

I.

I. In

Intr

trod

oduc

ucti

tion

on

 This section introduces the overall project and the concepts that are involved.  This section introduces the overall project and the concepts that are involved.

Goal

Goal

To design and build a small UAV that is capable of competing in the International Aerial To design and build a small UAV that is capable of competing in the International Aerial Robotics Competition 6

Robotics Competition 6thth Mission.Mission.

Personal Statement of Interest

Personal Statement of Interest

 The futu

 The future of technolore of technology always lies in makingy always lies in making things smag things smarter. rter. UnmaUnmanned aerinned aerialal vehicles (UAVs) are the next st

vehicles (UAVs) are the next stage in aviation. age in aviation. While human piWhile human pilots will always exist,lots will always exist, UAVs provide unparallel performance for dangerous situations, both for military and UAVs provide unparallel performance for dangerous situations, both for military and civilian appl

civilian applications. ications. While this is certainlWhile this is certainly a lucrative field, my a lucrative field, my interest in UAVs isy interest in UAVs is not based on financ

not based on finances. es. I decided on this topic bI decided on this topic because I wanted to builecause I wanted to build something.d something. I wanted a project that

I wanted a project that will challenge me to design and implement an entire system,will challenge me to design and implement an entire system, enco

encompasmpassing a rasing a range of tnge of topicopics. s. This pThis projecroject will ret will requirquire subste substantiaantial work in l work in aa va

varirietety y of of enengigineneererining g fifieleldsds. . WhWhilile e my my exexpepertrtisise e is is prprimimararilily y in in cocompmpututerer engineering, designing digital systems, this project will allow me to expand my engineering, designing digital systems, this project will allow me to expand my kno

knowlwledgedge e ininto to a a wiwide de ranrange ge of of appappliclicatiations ons in in aeraerodyodynamnamicsics, , guguidaidance nce andand navigation, and

navigation, and intelligent systems. intelligent systems. I cannot fathom a I cannot fathom a more comprehensive tomore comprehensive topic topic to get a chance to study a little bit of everything than in the interdisciplinary field of  get a chance to study a little bit of everything than in the interdisciplinary field of  robotics.

robotics.

Unmanned Aerial Vehicles

Unmanned Aerial Vehicles

 The unmanned

 The unmanned aerial vehicle is the fuaerial vehicle is the future of aviation technture of aviation technology. ology. After decades of After decades of  resear

research and testinch and testing, the UAV indusg, the UAV industry has finaltry has finally startely started to take off. d to take off. In the lastIn the last decade, the funding for UAV research, and especially VTOL UAVs, has increased decade, the funding for UAV research, and especially VTOL UAVs, has increased dra

dramamaticticallally. y. In the next decadIn the next decade, accoe, accordirding to ng to a a recrecent marent market studket study y by Tealby Teal Group

Group, UAV spending wil, UAV spending will more than double to $8.7 billl more than double to $8.7 billion annuaion annually(). lly(). This studyThis study po

poinints ts to to ththe e grgrowowining g dedemamand nd of of UAUAVs Vs in in ththe e inintetelllligigenencece, , susurvrveieillllanancece, , anandd reconnaissance sectors, especially in the US, which is predicted to account for 72% reconnaissance sectors, especially in the US, which is predicted to account for 72% of worldwide

of worldwide UAV funding(). UAV funding(). Of the UAV mOf the UAV market, VTOL vehicle darket, VTOL vehicle design is the fastestesign is the fastest growing secto

growing sector (). r (). Military use has Military use has historically been historically been the most comthe most common applicatmon applicationion for UAVs; however, VTOL vehicles have recently been introduced to the civilian for UAVs; however, VTOL vehicles have recently been introduced to the civilian secto

sector. r. CiviCivilian applian applicatlications for VTOL vehicions for VTOL vehicles includles include aerial photoe aerial photographgraphy, firey, fire suppression, search and rescue missions

suppression, search and rescue missions and surveillance.and surveillance.

One of the greatest advantages of VTOL vehicles is its high maneuverability, due to One of the greatest advantages of VTOL vehicles is its high maneuverability, due to its ability to hover and to f

its ability to hover and to fly at low altitudes. ly at low altitudes. This allows the vehicThis allows the vehicle to fly throughle to fly through narrow corridors and small openings, making it

narrow corridors and small openings, making it ideal for flight in urban and crowdedideal for flight in urban and crowded en

enviviroronmnmenentsts. . ThThe e momost st cocommmmon on VTVTOL OL UAUAV V dedesisign mimgn mimicics s ththe e shshapape e of aof a helicopter, provid

helicopter, providing excellent cing excellent control and maontrol and maneuverability. neuverability. While this dWhile this design hasesign has bee

(7)

payload capacity(). The four rotor helicopter commonly known as a Quadrotor or Quadcopter has been increasingly been used. This design boasts a number of  advantages over the traditional helicopter design(). The Quadrotor does not use any actuator control to change the direction of the thrust; rather it uses differential thrust to maneuver. This greatly simplifies the mechanical design of the vehicle. In contrast to the one large rotor with a traditional helicopter, the Quadrotor uses four smaller rotors for thrust. The smaller size reduces the possible damage to surroundings in the incident of a breakage and increases the response time associated with rotor drag. The simple design also lends itself well to reducing the aerodynamic complexity. One of the greatest advantages of the Quadrotor design however is the ability to support a heavier payload than other designs, making it ideal for carrying large cameras, powerful computers or communication devices. However, compared to other VTOL designs, the Quadrotor consumes a great deal of  power to stay aloft, which drastically limits flight time. In the RC hobbyist market, the most common designs are only capable of flight times under 20 minutes. However, for these short flights, no other design can match the stability and maneuverability of a Quadrotor, indoors or outdoors.

Previous Research with UAVs

A plethora of research groups are using the Quadrotor design as a test bed for a multitude of projects. Some of the more common uses are in developing control systems, improved autonomous control, sensor arrays and surveillance equipment. Prominent groups include the Autonomous System Laboratory at Eidgenössische  Technische Hochschule Zurich (ETH Zurich) in Switzerland (), the Stanford STARMAC group () in California, the Australian National University (ANU) X-4 Flyer Project ()(), and many others. Hoffman et al () provide an excellent review describing many of  these designs and their respective innovations. The numerous amount of published work using the Quadrotor design is indicative of the variety of topics involved. While the majority of the published work deals with control systems and sensor arrays, many papers have been published on developing accurate aerodynamic modeling(), improving the design ()() and developing point to point real-time communication between two Quadrotor vehicles(). In addition to published works, the hobbyist community has documented a great deal of work in regards to developing cheaper home built Quadrotors for personal use.

(8)

 The International Aerial Robotics Competition

  The International Aerial Robotics Competition (IARC) () is an international competition sponsored by the Association for Unmanned Vehicle Systems International (AUVSI)(). AUVSI sponsors a number of student competitions, including competitions for unmanned ground and underwater vehicles. The annual International Aerial Robotics Competition will celebrate its 20th year of competition

with the 6th Mission. In this new 6th mission, a small UAV will be launched to

infiltrate a hostile building. Once inside, the UAV is to find and acquire a specific small object without being detected by video surveillance. This is to simulate a covert operation for military usage.

  These competitions are designed to be challenging and generally beyond the abilities of the current technology. The 4th mission lasted for 8 years, and was never

fully completed. Even in the best of circumstances, it is unlikely that the project described in this proposal will be capable of competing in the IARC. The project will be undertaken with the intent of entering the IARC; however, the decision of  whether or not to enter will not be decided until a much later date. Even if this project is not capable of entering the IARC, the undertaking of this project will provide research opportunities in the field of unmanned aerial vehicles, inertial navigation systems, wireless communication, and collision avoidance algorithms and sensors.

Abstract

The 6th mission for the International Aerial Robotics Competition (IARC) will move the challenge to yet a higher level of autonomous aerial robotic behavior. The past two decades have seen a revolution in navigation technologies for operations in the open, but there is still much to be done in the area of indoor navigation. The goal is to create a small aerial robot capable of fully autonomous flight through a confined environment. In performing this task, the state-of-the-art in indoor  navigation, vehicle design and integration, and flight control will be  pushed to a higher level.

The 5th Mission of the IARC required collegiate teams to create fully  autonomous flying robots capable of negotiating a rather sterile environment. The new 6th Mission picks up where the 5th Mission left  off by demonstrating the fully autonomous aerial robotic behaviors necessary to more rapidly negotiate culturally-cluttered confined internal spaces of a structure once it has been penetrated by an air  vehicle, and intelligently interact with physical items encountered.

Notional Mission

Credible and actionable human intelligence (HUMINT) reports have been received from a mole within the Hesamic Republic of Nari’s Intelligence Organization. These reports indicate that highly sensitive information detailing plans to sabotage banking interests of a global organization may be stored in a security office located in the remote town of Rafq. A

(9)

breach in security has been identified which may allow a small autonomous air vehicle to penetrate perimeter defenses so that the sensitive information can be stolen by the global organization in order  to preempt any actions by the Nari government which would be deemed damaging to these unnamed global interests.

Before his untimely death, the mole was able to describe a number of  features within the Nari Intelligence Organization’s security compound and the desired target. Figure 1: Electronic Communiqué contains the electronic communiqué that is believed to contain reliable intelligence (NEC Pivot machine translation).

1. THE NARI SECURITY COMPOUND IS SURROUNDED BY RAISED TENSION [analyst: a high voltage] ELECTRIC FENCE WITH AN INNER RAZOR TELEGRAPH EXTENT [analyst: razor wire boundary or perimeter].

2. THE BUILDING IS CONSIDERED SECURE, BUT HAS VIDEO ÃOEBERWKöN [analyst: unreadable] AT VARIOUS POINTS. VIDEO IS MONITORED BY GUARDS IN AN

ADJACENT COMPOUND 300 METERS TO THE SOUTH.

3. CERTAIN HALLWAYS INTERNAL THE SECURITY COMPOUND GET LASER

“TRIPWIRES” WHICH ABLE TO BE DEACTIVATED MANUALLY. THE LASER BEAMS, IN CASE BROKEN IN TWO, INITIATE AN LOUD HEARING [analyst: audible] ALARM.

4. EVERYONE’S FLOORS HOLD PRESSURE-FEELING üBERKIDEN [analyst: unreadable; perhaps “switches”] THAT ARE MADE TO LIVE [analyst: armed or switched on] AT THE EVENING OF EACH WORK DAY.

5. A UNMARRIED GUARD [analyst: sole or single guard] PATROLS THE SECURITY

COMPOUND ON A 10-MINUTE ROTATION. HE CAN BE SEEN WALKING ÃÜJ ÂRSƑ [analyst: unreadable] THE COMPOUND AND THEN ONCE EVERY 10 MINUTES, ENTERING THE COMPOUND FOR A SECURITY INSPECTION LASTING ABOUT 15 MINUTES. HE THEN EXITS PRESENT EIFGäNGE THE ENCLOSURE TO üBIRPRüFEN BEING EXTENT. BEGINNING THE ENCLOSURE KöNND ONLY BEING ACHIEVED WHEN THE GUARD AM OUT. ENTRANCE WAY OUT. [analyst: entering the compound can only be achieved when the guard is outside] INGRESS/EGRESS MUST OCCUR IN BELOW 10 MINUTES TO AVOID DETECTION.

6. THERE ARE SEVERAL OFFICES WITHIN THE COMPOUND. THE SECURITY OFFICE CAN BE FIND BY OBSERVING DIRECTIONAL SIGNS AND IS UNIQUELY IDENTIFIED BY A SIGN OVER THE DOOR.

7. ONE TWINKLE DRIVE CONTAIN HURTFUL INFORMATION [analyst: a flash drive

containing sensitive or damaging information] OF INTEREST TO OUR GLOBAL MASTERS AM OTHERWISE KEPT ON THE TOP THE LAYERS OF PAPERS IN THE CHIEF OF

SECURITY’S IN-BOX EXISTING ON HIS DESK.

8. A SINGLE UPPER STORY WINDOW IS BROKEN AND REMAINS OPEN ON THE STAGE [analyst: level or floor] BELONGING THE CHIEF OF SECURITY OFFICE.

END OF TRANSCRIPT (RECEIVED 31 JULY 2009; NOK KUNDI SATNODE)

(10)

replace the original drive to delay detection of the missing data.

The covert mission must be conducted without compromise of the organizations funding the mission, therefore no identifying markings or  information that can be traced back to the organizations shall be incorporated onto or into systems which may be compromised or  captured should the mission fail. The priority of mission options is as follows:

Clean Mission: Covert ingress; flash drive swap; covert egress; delivery of flash drive to your handler (Judge). Requires mission completion in under 10 minutes to avoid notice by patrolling guard.

Dirty Mission: Detection upon entry: Covert ingress fails, alarms activated; flash drive swap; rapid egress; delivery of flash drive to your  handler (Judge). Vehicle has tmisson to complete the mission and avoid guards which have been alerted.

tremaining5 minutes from alarm activation if alarm activation is at tmission <5 minutesRemainder of 10 minutes if alarm activation is at tmission is at tmission ≥5 minutes

Mission Failure Type 1 - Failure to enter Security Compound building: Abort mission (attempt terminates).

Mission Failure Type 2 - Failure to exit building in under 10 minutes (Clean Mission criteria) or in under 5 minutes (Dirty Mission criteria): Explode the air vehicle before the 10 minute/5 minute detection threshold, destroying the flash drive in the process (self destruction simulated by shutting down propulsion system either in flight or after landing while initiating a continuous audible “beeping tone”).

Paper Organization

  The rest of this proposal is organized as follows. The Stages of Design and Construction section examines the proposed plan for this project, including both the thesis work and the actual construction of the Quadrotor. The following section, Design Considerations, discusses in detail the requirements for the IARC 5th Mission

and begins an explanation of what components will be needed in the Quadrotor. Background information on how the Quadrotor achieves motion is discussed in the Quadrotor Basics section. The proposed components for the Quadrotor are discussed in detail in the next section, Quadrotor Components. In the PC Interface section, the requirements for communication with the Quadrotor are examined, and additional components are proposed. Additional equipment necessary for methodical testing is discussed in the Overview of Testing Equipment section. Quadrotor Design and Construction describes the proposed construction of the Quadrotor, broken down into individual stages. The financial requirements are discussed in the section. The final section, Conclusion, provides a short summary.

(11)

II.Stages of Design and Construction

Stages of Thesis Work

 The thesis aspect of this project will be done in a series of stages. The end goal is to achieve a complete objective planning autonomous Quadrotor. However, a project of this magnitude must be broken down into a series of intermediate stages. A listing of the stages and the requirements of each stage is seen in Figure 2: Stages of Research.

Figure 2: Stages of Research

Research

 The research stage is the most important, providing a knowledge base for all other stages. The research will come from a variety of sources, taking full advantage of the wealth of information in published research works as well as from hobby enthusiast sources, such as message boards and internet forums.  The research stage is a never-ending stage, though it is most beneficial at

the beginning of the project.

Inertial Navigation System

 The Inertial Navigation System is a critical element in the Quadrotor design. It is required to provide accurate data for flight stabilization. In the absence of GPS, the use of highly accurate accelerometers and gyroscopes will be necessary as a means to counteract the drift bias commonly found in similar inertial based navigation systems. Magnetic navigation may also be incorporated. The use of the Extended Kaman Filter has also been documented to greatly aid in inertial navigation.

Manual Flight

Hobbyist sources will be most beneficial in this stage of work. Very few Quadrotors are available commercially, most are only available only as kits. In this stage, a Quadrotor will be designed and constructed according to the physical requirements specified by the IARC. Manual flight will be controlled with a high-end RC helicopter transmitter. This transmitter/receiver system will remain in place for the entirety of the project, as it meets the IARC requirement of a manual override of the propulsion system.

Flight Model for Stabilization

(12)

Ground Station Communication

A ground station will be a highly beneficial addition early in the project development. In the early stages, it will be designed to receive the stream of  data from the various stabilization sensors to aid in the development of a flight model. In the later stages, it will be improved upon to allow for bi-directional communication. During the development of the stabilization system, it may be done with a desktop computer in a small indoor controlled environment. However, it may be beneficial to upgrade to a mobile laptop towards the completion of the project.

Preplanned Flight

Once the Quadrotor has demonstrated manual control, the inertial navigation system can be integrated with a separate control system to complete a preplanned flight path. This stage will have a separate control system replacing the input signals from the manual controller, with the observational state data coming from the inertial navigation system.

Collision Sensing Array

 The next stage consists of the development of an array of sensors that will be capable of providing information to the Quadrotor of potential obstacles. This will likely consists of a series of ultrasonic sensors for distance measurements in several directions. An additional module will likely be added to the system to control trajectory planning for obstacle avoiding preplanned flight.

Object Identification

Whereas collision detection will likely require only distance measurements, object identification will require a more complex sensor system such as audio and video processing. The required computation will likely exceed anything that could fit on the Quadrotor itself, and thus should be handled by the ground station. High data communication throughput will be required for this stage

Object Acquisition

Once an object has been identified, it needs to be acquired by means of a gripping mechanism. This necessitates both a gripping hand and a way of  direction the hand to the object. This system will likely incorporate an additional visual system and a retractable hand.

(13)

Objective Planning

 The final stage is to integrate the object identification system with the rest of  the Quadrotor system. The ground station will provide directions for the Quadrotor in order to locate a particular identified object based on the audio and video streams. The Quadrotor will be responsible for steady and stable movement, avoiding obstacles and object acquisition.

Stages of Quadrotor Design

 The stages for the development of the Quadrotor will also be done in stages, and follow the stages of thesis work. In a project of this magnitude, attainable short-term goals are a necessity to quantify success. The advancement of the stages is marked by accomplishment. The Quadrotor may advance to the next stage upon the completion of the requirement of that stage. The requirements of each stage are seen in Figure 3: Quadrotor Stages. The details of each Stage are described fully in Section III Quadrotor Design and Construction.

(14)

Objectiv e Planning  Trajectory Control and Planning INS and Motion Control System Flight Stabilizati on System Motor s PC Data link Gripping Actuator

Modular design

 The short-term goal progression of the project facilitates the modularization of the Quadrotor. Once a system has been completed to satisfaction, it is designed to not need further adjustment. A diagram of the modular design of the Quadrotor system is seen in Figure 4: Modular Design.

  The flight stabilization system takes values from the accelerometers and gyroscopes to maintain a stable flight in regards to the desired direction. Not pictured is the manual control input to the flight stabilization system that would dictate desired direction in Stage 1. The INS and Motion Control System also require the accelerometer's and gyroscope's data as well as navigational information based off an electromagnetic compass. Upon reaching Stage 2, this system will be able to control the basic movement of the Quadrotor as well as provide live data to the ground station. The sensory array will provide information for the next stage for trajectory control to avoid obstacles. The final component for Acceleration Data

Acceleration DataCompass DataGyroscopic DataGyroscopic Data Compass DataSensory ArraySensory Array

(15)

objective planning will largely be computed by the ground station, though an overlying system for interpreting objective commands will likely be needed on the Quadrotor. This final system also includes the control system for controlling the acquisition of specified objects.

(16)

Naming convention

 The naming convention for documenting the progress of each iteration of the Quadrotor's development is specified as follows.

Figure 5: Naming Convention

Example:

Mark 0.05.12.025: Quadrotor has not yet achieved manual flight. The Quadrotor physical design has been changed five times and the software has been adjusted 12 times. The control system gain values are on their 25th iteration.

Mark 2.03.07.010: Quadrotor has achieved Stage 2 and is in the   process of achieving obstacle-avoiding flight. Since achieving

Stage 2, the physical design has been adjusted three times, the software seven times and the controller gain values 10 times.

(17)

III.Design Considerations

In considering the design of the Quadrotor, a review of the requirements given by the IARC is needed. The following list is a compilation of the important requirements and broken down into three categories: Vehicle Requirements, Ground Station and External Navigation Aids Requirements, and Mission Requirements. ()

IARC Requirements

Vehicle Requirements

1. Vehicle must be unmanned and autonomous.

2. The maximum vehicle takeoff weight may not exceed 1.50 kg 3. The maximum dimension of the air vehicle may not exceed 1m.

4. The vehicle must be equipped with a method of manual-activated remote override of the primary propulsion system.

Ground Station and External Navigation Aids Requirements

5. Computers operating from standard commercial power may be set up outside the competition area boundary and bi-directional data may be transmitted to/from the vehicle.

6. The system (consisting of the vehicle and any ground station equipment) must be capable of providing vehicle navigational and status information to a remote JAUS-compliant terminal.

7. A maximum of two non-line-of-sight (NLOS) navigational aids may be used external to the designated flight area. The navigational aids must be

 portable. GPS is not allowed as a navigation aid.

8. The ground station equipment other than the navigational aids, manual kill switch mechanism and judges JAUS-compliant terminal interface must be  portable.

Mission Requirements

9. The vehicle will be required to enter the building through a one square meter opening, 3m away from the launch area and will be required to navigate through hallways.

10.The vehicle will have to search for a target area while avoiding obstacles as well as visible security systems.

11.Security systems include a scanning video camera system and a laser  barrier. The scanning video camera at the window entrance is simulated by a slow blinking blue LED light. The laser barrier must be manually  disarmed.

12.The building will contain several signs indicating the route to the target  area.

13.A flash drive will be in the target area on top of a stack of papers on a desk. The vehicle must be able to remove the flash drive and replace it  with a fake one.

14.The mission must be accomplished within 10 minutes of entering the building covertly, or under 5 minutes if detected.

(18)

 Justification of Quadrotor

  The Quadrotor design is ideal for this task. The relative small size compared to other UAV designs meets IARC requirements and . The use of the manual remote control from the first stage of Quadrotor development satisfies requirement . The hovering capability offered by the Quadcopter allows for high maneuverability, (requirement 9) and stability (requirement 11, 13). Requirements 10, 11, and 12 will necessitate a variety of sensors, which the Quadrotor with its superior carrying capacity can support. In addition, requirement 13 requires that the vehicle must be able to acquire a small object and replace it with a duplicate. The Quadrotors superior stability and carry capacity will allow for an effective gripping mechanism to be mounted on board. Speed may be an issue for the Quadcopter design, however the time restriction from requirement eliminates the Quadrotor’s main deficiency, long flight times. The requirements relating to the ground station can be met with the proper hardware.

(19)

I. Quadrotor Basics

  The Quadrotor is a mechanically simple vehicle, requiring only four stationary motors to achieve all flight motions. Despite its simplicity, it is very capable of  advanced flight. The key to flight movements beyond hovering is in differential thrust. A simple diagram is seen in Figure 6: Quadrotor Body Diagram.

In the above diagram, where ωi represents the angular velocity of rotor i, the

rotation movements roll, θ, pitch, Φ, and yaw, Ψ can be described as functions of ωi.

ω1and ω3 rotate clockwise, while ω2 and ω4 rotate counter-clockwise. This balance

of rotation allows the Quadrotor to counteract torque from rotor blades. To rotate the Quadrotor in the θ direction, angular velocity ω4 is increased and ω2 is

decreased. To maintain a stable altitude, ω1 and ω3 will have to be increased

slightly as the Quadrotor will tilt and start to fly sideways along the X-axis. Differentiating ω1 and ω3 will result in movement along the Y-axis. Movement in

the Z-axis is simply controlled by the combined thrust of all the rotors. Yaw ω1 ω2 ω3 ω4 Φ θ Ψ X  Y

(20)

 The thrust force from each motor can be described to be related to the size of the rotor blade and the torque caused by the rotor in the air. There are several variations of this formula, and it is premature to decide which variation will be used in this project. The development of these formulas will be critical in developing the aerodynamic model used in simulations.

 The control system will have to be an underactuated system (). An aerial vehicle composes of at least 12 state variables, three degrees of freedom (DOF) in each position, position rate, rotation, and rotation rate. Inputs, on the other hand, only consist of the acceleration of position and rotation in the three axes. In addition, the flight dynamics of UAVs are typically nonlinear () due to the aerodynamics of the vehicle flight as well as from coordinate transforms.

In all implementations, the movement of the Quadrotor is separated into two control systems: attitude and altitude. Attitude refers to the rotation of the Quadrotor, while altitude is specifically for up/down movement. The attitude controller is responsible for the bulk of the changes in motor thrust control. However, the altitude controller is valuable to compensating for the force of gravity as the Quadrotor rotates roll or pitch.

 The traditional solution for the control system is a monolithic linearization approach (). While this method has its drawbacks, it has been shown to be more than adequate in several papers ((), (), (),()). The PID control system of a Quadrotor is the most commonly used system in hobbyist Quadrotors as well because of its simplicity. However, the STARMAC group point out that this linear approximation is only suitable for hovering and moving at low velocities(). At higher velocities, the flight dynamics begin to resemble an airplane, which necessitates a different approach for aggressive maneuvers. They demonstrated an updated linear controller for aggressive maneuvers in a subsequent paper ().

(21)

II.Quadrotor Components

  The Quadrotor design itself requires an examination of the components. The majority of the design is based on hobby enthusiast designs because of its low cost and readily available parts. The physical chassis will measure 70 cm across. The dimensions of the center casing will be determined during construction, but will likely be under a diameter of 15 cm.

 The rest of the physical design consideration is to prevent damage to the Quadrotor and the user. The motors are to be placed at least a blade’s distance from the edge to protect the blade from collision. Additional stabilization rods framing the tips may be used. A protective cover over the center casing would also provide extra protection to the electronics inside. Landing gear would provide a stable base during takeoff and landing. See Figure 7: Quadrotor Rendering for a potential physical structure. This rendering includes the stabilization rods bracing the arms, a center dome for chassis protection and foam blocks as landing gear. The gripping mechanism has yet to be determined, but will be designed to be placed underneath the chassis.

(22)

 Transmitter/Receiver

 The radio-control system is one of the most important components of any remote system. In the higher end hobby RC systems, each of the major types of vehicles have their own dedicated radio transmitter/receiver system. However, the Quadrotor does not yet have its own dedicated system; as such the use of RC helicopter transmitter/receiver will be used.

 The latest generation of RC radios, however, uses spread-spectrum technology at 2.4GHz. Pioneered by Spektrum (), this allows nearly interference proof flight as well as increased resolution and latency. These radios transmit on a single selected frequency, but at a wide band that eliminates interference (). To further combat frequency interference, these receivers only accept data sent with the correct identification code. A 2.4 GHz transmitter/receiver radio system will be used for this Quadrotor project because of its robust performance.

In most RC helicopter radios, a minimum of four channels is necessary: Throttle, Elevator, Aileron, and Rudder. For higher end helicopters, more channels may be used to control collective-pitch and gyro gain. Many advanced radios have as many as 10 channels available with additional features such as adjusting the pitch curves of  the control-stick input, adjustable trims, programmable switches and a manual kill switch. In the first stage of development, the transmitter will provide manual control of the Quadrotor. For a Quadrotor, the main four channels are very similar to the helicopter system. The minimum required channels necessary for flight control is 4: ()

•  Channel 1 – Roll - Left/Right Stick  •  Channel 2 – Pitch - Up/Down Stick  •  Channel 3 – Throttle - Up/Down Stick  •  Channel 4 – Yaw - Left/Right Stick 

Channel 1 and 2 controls the amount of tilt in their respective directions. Throttle controls the power of the motors to achieve lift. Channel 4 controls the rotation of  the Quadrotor around its center axis.

In addition to these four channels necessary for manual flight, a minimum of 2 additional channels will be used to aid in the development of this project. Taking cue from other aircraft, Channel 5 will implement a landing gear system. This system will be particularly useful as a way to prevent damage in the event of a crash due to a system failure. The ability to turn the preplanned flight control on and off will be implemented in Channel 6. While a kill switch is a mandatory requirement of the IARC, this can be accomplished with the advanced features of  the RC transmitter.

(23)

  The specifications of the receiver for these RC systems mainly depend on the matching transmitter. Typically, the motors of the aircraft are plugged directly into the receiver, leaving any signal mixing or adjustment to the transmitter. This keeps the weight of this component as low as possible.

 The DX6i transmitter (see Figure 8: Spektrum DX6i (14)) by Spektrum () fulfills all the requirements of the transmitter for use in this project. This six channel transmitter operates at the 2.4 GHz band, is compatible with the lightweight AR6200 receiver and includes such features as a throttle cut, a large LCD display and programmable sequences. The complete DX6i system has a retail price of  $259.99, including the AR6200 receiver.

Rotor System

  There are three main components to the physical movement on the Quadrotor: motors, blades and electronic speed controllers. The most effective solution for propulsion of the Quadrotor is to use electronically controlled DC motors. However, due the power requirements of these motors, direct control of the motors by the microcontroller platform is not feasible. This necessitates that electronic speed controllers (ESC) will be used.

DC Motors

In an electric propulsion system, there are two types of DC motors: Brushed and Brushless motors. While brushed motors are significantly cheaper, the unequivocal preferred motor is the DC brushless motor. They are typically more powerful, more efficient, and more durable. Whereas a brushed motor contains metallic brush contacts to provide energy to the rotating armature, in a brushless motor, stationary coils along the edge of the motor cause the magnet connected to the output shaft to rotate().

Brushless motors are further classified between inrunner and outrunner motors. This specifies the

location of the rotating magnet in the motor. An inrunner has the magnet surrounded by the stationary coils, allowing for high speed rotations. An outrunner has the magnet on the outside, rotating around the stationary coils. While this results in a slower rotation, outrunner motors typically have more torque, removing the need for a gearbox().

The most important specification of a motor is the

(24)

2410-Outrunner 2410-09 (17)). This motor meets all the requirements of the Quadrotor design. It is a high torque motor with a Kv rating of 840 and a weight of only 64.5g. The power consumption reaches a maximum of 104 W, at max using 13.5 A. The most desirable aspect of this motor however is its extremely low cost at only $6.39 each.

Contra Rotating Blades

Special care is necessary in determining which set of blades to use. There must be both clockwise and counter clockwise rotating blades in order to cancel out rotational torque. In this system, the front and rear blades will rotate clockwise while the left and right blades will rotate counter clockwise. In addition, blade size and pitch are important to consider. Blades are commonly identified by a four digit number that can be parsed to indicate blade diameter and pitch. Pitch is measured at the theoretical distance a propeller moves forward in one revolution. A commonly recommended blade is the counter rotating EPP 1045 blades (see Figure

10: Contra Rotating EPP1045 (15)). In this model, the blade diameter is 10 inches, and

the pitch is 4.5 inches per revolution. This blade set retails for $4.49 (). At least two sets will be needed for the Quadrotor; however, several sets should be kept as spare parts since blades are easily chipped or broken.

Electronic Speed Controllers

An electronic speed controller (ESC) simply acts as an open-loop controller for maintaining the motor speed. The important specifications for ESCs are the current rating and the voltage

rating. These ratings are the maximum current and voltage that they can handle. Advanced ESCs also provide programming features that can adjust throttle curves and other tweaks.

A light, low cost ESC is the Turnigy Plush 18 A SC (see Figure 11:

(25)

 Turnigy Plush 18A SC (19)). This small ESC weighs only 19 grams and retails for $15.95. It can support up to 18 A and up to 16.8 V (). Despite the critical importance of these controllers, very little published data is available. Two very critical elements are commonly missing in ESC technical sheets: the update rate and the resolution. This necessitates that the entirety of the motor system will have to be rigorously examined to create an accurate model. A motor testing rig is described in Section VII.

Battery

 The two most important aspects in choosing a battery are the supply voltage and the current capacity. Lithium-ion Polymer (LiPoly) batteries are the most common battery and provide the highest power output per dollar. In this Quadrotor system, because the competition flight is limited to under 10 minutes, an off the shelf  battery will be more than adequate. A commonly recommended battery is the  Turnigy 4000mAh 3 Cell battery (see Figure 12: Turnigy 4000mAh ()()). While it weighs a hefty 337g, it will provide more than enough power to the system and retails at a reasonable $27.73 (). While LiPoly batteries can be charged in as little as an hour, this downtime can prolong the project unnecessary. Multiple batteries are recommended as well as a suitable system for battery charging and management.

(26)

Microcontroller Platform

 The Quadrotor will require a significant amount of processing power to accomplish all of its goals. However, not all of this processing power has to be from a single microcontroller. The modular design allows for multiple microcontrollers, each handling a different stage of the overall design. This places the microcontroller requirements more on size and cost rather than power and memory.

 The Arduino() is an open-source platform for programming the high performance, low power Atmel AVR microcontrollers. The greatest advantage of the Arduino platform is its ease of use. Providing all the necessary components needed to use the microprocessor as well as breaking out all the pins, all the Arduino needs is power and something to do. Evan though they are powerful and versatile, these Arduino boards are small and light, perfect for use in the Quadrotor. The Arduino Mega (see Figure 13: Arduino Mega) is the most powerful board, complete with 16 10bit analog inputs, four UARTs and 128 Kbytes of program memory space. This board is 5.4cm x 11.11cm and retails for $64.95 at Sparkfun Electronics ().While it is larger and more power intensive than the smaller boards, this board should be more than adequate for Stages 1 and 2 of the Quadrotor design. This is a departure from the initial design to combine Stages 1 and 2 on a single board. However, since the majority of the system inputs are shared in these stages, and require costly ADC processing, it appears better served to use one board. Multiple boards may be used for the latter stages. The controller for the remaining stages will be determined when needed. The modular system design suggests at multi-controller design, though the exact requirements needed for the later stages are currently undetermined. An Arduino platform is the ideal setting, though a more advanced microcontroller may be necessary.

(27)

Sensors

 The sensor array is one of the most critical elements of the Quadrotor design. In the development of Stage 1, rotational information from all three axes is necessary for stable flight with manual control. Positional information from accelerometers will also be valuable. In the development for Stage 2, compass heading information, adjusted for tilt, will aid in the preplanned flight control and to minimize yaw drift error. Collision detection will require a completely different set of sensor information, requiring range information which may come from infrared sensors, ultrasonic sensors or scanning laser rangefinders. Object identification for the final stage will require optical capability, likely a video system with object identification algorithms. The exact sensors for the Quadrotor will likely be determined on a need basis, arising once the Quadrotor has achieved the proper Stage.

Inertial Navigation System

 The Inertial Navigation System (INS) is critical to the development of Stage 1 and 2. This system will be comprised of a 6 degree of freedom (DOF) sensor array, measuring acceleration in x, y and z directions and rotation in roll (θ), pitch (φ), and yaw (ψ). There are a variety of accelerometers and gyroscopes that can be assembled into a 6DOF system.

While accelerometers and gyroscopes are commonly used in this way, they frequently are affected by drift error caused by the integration, which increases drastically over time. Many systems use GPS as a way of keeping the error low, however, due to the competition rules, GPS will not be used.  This limits the effectiveness of solely using accelerometers and gyroscopes for positioning information. In this project, an additional set of magnetic compass sensors will be used. This is predicted to help reduce the yaw drift error.

A novel implementation of an ultrasonic sensor pointing directly down has been successfully demonstrated by STARMAC (). This was proved to be effective in reducing the drift error in the z axis and will be attempted in this project as well.

 The requirements of this INS are steep due to its importance in the Quadrotor system. A suitable array of sensors can be found Sparkfun’s IMU 6 DOF v4 with Bluetooth Capability (see Figure 14: IMU 6 Degrees of Freedom - v4 with Bluetooth Capability) (). This comprehensive device provides not only 3 axes of acceleration data, 3 axes of rotation data and 3 axes of magnetic data, it also contains a control board that includes a built-in filters and a Bluetooth module for live measurement readings. This small device is only 4.28 cm x 6.35 cm, helping keep the size of the main module down. The downside of  this module is its high price at $449.95.

(28)

Figure 14: IMU 6 Degrees of Freedom - v4 with Bluetooth Capability ()

 Though the IMU 6DOF v4 fulfills the necessary capabilities needed, it is an expensive component. A cheaper alternative would be to build an IMU with breakout boards. Sparkfun also sells IMU 5 DOF Combo Board (see Figure 15: IMU 5 Degrees of Freedom()()) for $99.95(). This smaller board breaks out a 2-axis gyroscope for roll (θ) and pitch (φ) and a 3-axis accelerometer. An additional gyroscope will be necessary for yaw (ψ) rotation. This can be done with the Gyro Breakout Board – Dual 500 degree/sec also by Sparkfun for $59.95(). This breakout board uses the same gyroscope model as on the IMU 5 DOF Combo Board. If a compass heading is desired, it can be done with a Compass Module with Tilt Compensation – HMC6353, also from Sparkfun for $149.95().

Figure 15: IMU 5 Degrees of Freedom() Figure 16: Gyro Breakout Board - Dual 500 degree/sec ()

(29)
(30)

Distance Measuring Sensors

As mentioned earlier, an altitude measurement will be done with a range finding ultrasonic sensor. This method has been demonstrated to help reduce altitude drift errors from a purely inertial measurement system (). As with most components on the Quadrotor, small size and low power consumption are the most important specifications. The Maxbotix LV-EZ1 (see Figure 18: Maxbotic LV-EZ1()()) meets the requirements. This small module is only 22.1mm x 19.9 mm x 16.4 mm and weighs only 4.3 grams.  The input voltage ranges from 2.5V to 5.5V and has a current draw of 2mA.  The module is limited to a resolution of 2.5 cm; however that should be

accurate enough to reduce high drift error. This particular model has an analog output and can operate at 20Hz. It is available at many locations, including Sparkfun() for $24.95.

Figure 18: Maxbotic LV-EZ1()

In the later stages of the Quadrotor development, distance measuring sensors will be utilized for collision avoidance. More research will be needed to determine the exact nature of this sensor array. There are ultrasonic sensors with a narrow range that can be arranged in an array to provide a rough depth mapping of the surroundings. This array would likely require its own controller to analyze the data into a meaningful fashion. This has been examined as a potential working system by Bouabdallah et al (). As an alternative to ultrasonic sensors, infrared range finders (see Figure 19: Sharp GP2Y0A02YK0F() ()) may also be used. Infrared sensors work in a similar fashion, and are generally cheaper. However, they have a much more limited range, usually less than 1m. This may be not be an issue in collision avoidance; however it may not be useful for the trajectory planning required for the final Stage.

(31)

An expensive, yet more comprehensive system is a scanning laser rangefinder. These systems are able to report very precise range information in a precise angular resolution. One particular model, the Hokuyo UBG-04LX Scanning Laser Rangefinder (see Figure 20: Hokuyo URG-04LX) is a small module that weighs only 160grams. It is a significant power draw, consuming 500 mA, but with its superior resolution and fast data transfer, it could be a valuable tool for trajectory mapping. A navigation system using the Hokuyo Scanning Laser Rangefinder was developed for indoor flight and mapping out of unknown terrain with a great deal of success ().

Figure 19: Sharp GP2Y0A02YK0F()

(32)

Building Material

 The actual building material for the Quadrotor should be not only durable but light weight as well. Balsa wood is light and easy to work with, however is not durable and would likely not be able to support the weight of the motors. Other wood varieties might be feasible, though the amount of wood needed to support the weight may be too large. ABS plastic is an extremely durable compound; however it is difficult to work with, and may be too heavy. Metals such as aluminum would be feasible, as long as it is light enough. However, the preferred building material is carbon fiber. Carbon fiber (see Figure 21: Carbon Fiber Square Tubes ()()) is a commonly used building material for a wide range of applications, and is especially desirable in the R/C hobbyist community. While not as cheap as other materials, the high strength-to-weight ratio makes it highly desirable. For improved strength, it is recommended to combine it with wood strips and epoxy to form a composite material. The dimensions of the Quadrotor design dictate that less than 2m of  carbon fiber will be necessary, however, having additional pieces of material would be beneficial. A 0.75m square tube can be purchased from Hobbyking.com for $4.26 each().

 The casing for the control system, sensors and batteries will be formed out of thin pieces of aluminum sheets. The actual physical design of the casing will likely be constantly adjusted for maximum efficiency. A preliminary design contains a three tier casing, with the bottom level for the battery, the second level for the microcontroller and the top level for data communication devices.

(33)

I. PC Interface

In the later stages of the project, communication between the Quadrotor and an external ground station will be critical. The key features required of this communication are the high data throughput and the 6dBm loss at a range of  100m(). Ideally, all sensor data should be sent to the ground station for testing purposes. This includes data from all three axes of accelerometer data, rotation data from all three gyroscope axes, magnetic compass data, position information that the Quadrotor uses for its location and data from the collision detection sensors. All of this information also must be transmitted at a high sampling rate to ensure accuracy. A rough estimate of the required transmission is seen in Figure 22: Data Transmission.

Figure 22: Data Transmission

Communication

A sampling rate of 40Hz should be adequate enough to analyze the data from the Quadrotor. Assuming that each Stage will require at least 48 bytes per sample, a total of 1.152kbits of data will have to be transmitted every 25 ms. This results in a minimum data transmission rate of 46kb/s. Taking into consideration a conservative overhead of 30%, a transmission rate of at least 60 kb/s is required. Stage 4 (not seen in Figure 22: Data Transmission) however will also require a data connection, but from the PC to the Quadrotor. In the previous stages, any user communication from the ground station to the Quadrotor can be neglected because they would be based on human commands, a notoriously slow bandwidth requirement. In Stage 4, the ground station will be actively guiding the Quadrotor in its object search. This can be assumed to require at least 20kb/s of transmission rate, increasing the required transmission data rate to at least 100 kb/s (including an additional conservative overhead). The audio and video transmission due to its immense bandwidth requirement will likely operate on a separate dedicated transmission system.

While the transmission distance is not particularly large, the power loss requirement is designed to simulate the transmission traveling through a concrete building. While commercially available Bluetooth modules easily support the required data rate, the 100m transmission range is just at the edge of the connection range. The Bluetooth communication built into the desired sensor system will prove to be valuable for testing purposes, but not adequate for the requirements of the IARC. As such, a more powerful and robust communication will be needed. The XBee family of RF communication devices is commonly used in the electronic hobbyist projects. It boasts many advantages including its ease of use and wide range of  configurations(). The XBee Pro 50mW Series 2.5 RPSMA module is ideal for this project (see Figure 23: XBee Pro 50mW Series 2.5 RPSMA). It has a max data rate of 250 kb/s with a range of over 1600m. The 50mW output should be enough to

(34)

Figure 23: XBee Pro 50mW Series 2.5 RPSMA ()

User Interface

 The user interface is not a required element to this project; however it may be a useful addition. The proposed user interface will provide a variety of features. The user interface will likely be programmed in C++ or C#. There are two main functions to this interface: an integrated development environment (IDE) and a live display.

IDE

While the proposed Arduino microcontroller platform has its own IDE, the Quadrotor’s IDE will provide additional resources related to the control system and Quadrotor specific configurations. The control system gain values are likely to be the most adjusted part of the system. With the proposed system, these values will be easily adjusted and the related response function can be displayed visually.

  The IDE will also provide a configuration setup for the Quadrotor. Even though the proposed system will be using identical motors, blades and ESC’s for each of the arms, there may be some slight disparities which may impede stable flight. The configuration tool will provide a way to adjust each of the motors for maximum effect.

 The features provided by the proposed IDE have been used to great effect by the MikroQuad Quadrotor project() and its successor, AeroQuad().

Live Display

 The live display is designed to provide the user with a visual representation of the data coming from the Quadrotor. During the first and second stage of  Quadrotor development, the data will consist of both sensor data and the calculated values. This will provide valuable data to evaluate the effectiveness of the algorithms. In addition to sensor and algorithm data,

(35)

status information from the Quadrotor is also extremely valuable. This status information includes power consumption, flight time, and motor speed.

Programming Issues

In examining the proposed user interface, there are a couple of issues that need to be addressed. In order for communication, a full duplex link is needed between the Quadrotor and the computer. Since the Quadrotor will be relying on XBee for reliable communication, another XBee module will be needed to receive data. Another component will be needed to connect the Xbee to the computer. This can be easily accomplished with an XBee Explorer USB (see Figure 24: XBee Explorer USB ()()) also available from Sparkfun for $24.95 (). This module provides an easy interface for serial transmission over USB.

Figure 24: XBee Explorer USB ()

 The programming of the GUI will also be required. This will involve several real-time graphical displays and access to the IDE of the Arduino. Since the Arduino source code is open-source, it will ease the implementation. Due to its ease of use, C# may be preferred programming language for implementing the Quadrotor IDE.

 JAUS Terminal

 The Joint Architecture for Unmanned System (JAUS) is a standard for interoperability and technology insertion in the field of unmanned systems. The IARC requires this standard for an interface terminal to ensure the robot’s navigational and status information. While the current standards are proprietary, older legacy versions still exist and will be the basis for the communication standard. A lengthy detailed review of these standards is underway. The latest JAUS documentation will be provided by the IARC upon registration.

(36)

II.Overview of Testing Equipment and Tools

 This section describes the tools and extra equipment that may be useful in the development of the Quadrotor.

 Testing Equipment

  The following equipment will be useful in rigorously testing the components. The motor thrust test-bench will provide valuable data for determining the characteristics of the motor, the ESC, vibration analysis on the chassis, and total thrust system identification. The battery monitor system will likewise provide valuable information on the power consumption of the Quadrotor and help prevent damage to the LiPoly batteries.

Motor Thrust Test-Bench

A test-bench for determining the thrust output from the motors will be necessary. This can be accomplished by measuring the force exerted with an accurate weight scale. A first-class leverage system can be built using the Arduino and a pressure sensor (see Figure 25: Motor Test Bench). The Arduino can control the motor speed through the ESC, and then measure the resulting force from the pressure sensor. The lever can be built using a spare carbon fiber rod and the pivot structure can be assembled from some aluminum.

  This type of test-bench has been commonly used by several research groups. The STARMAC group at Stanford used a similar test-bench to examine the aerodynamic effects of flight, including total thrust during forward flight, blade flapping effects, and airflow disruption(). The X-4 Flyer Project at ANU also examined the aerodynamic effects as well as Motor Assembly Direction of   Thrust Direction of  Leveraged Force Balancing Weight Pressure Sensor PC Ardui no

(37)

developed highly accurate control system transfer equations for their simulation models. In their simulations they modeled their thrust system as a two-pole, one-zero system(). In addition to research groups, a group of  hobbyists have also developed a similar test-bench and used it to examine a variety of motors and the effect of vibration on thrust performance ().   Their comprehensive examination provided the information that helped decide the motor for this project. In addition to testing motors, their work examined vibration from the motors and resonances and possible ways to dampen it.

 The test-bench in this project will focus on the examination of vibration and the modeling of the system, including a comprehensive study of the ESC.  To accomplish this, the test-bench will require several features. For the

development of the model, the system will need to be able to record the amount of thrust, the speed of the revolutions as well as the input signals to the ESC. The thrust can be measured using a pressure sensor, while the revolutions will be counted using an optical tachometer. For examining the vibrations in the system, the sampling rate of the system must be high. In the informal studies, the majority of the vibration was found to be in the 1 to 2 kHz range, suggesting it stems from the motor itself. As a result, the proposed test-bench must be able to sample at above 4 kHz to fulfill the Nyquist requirement.

 The comprehensive study of the ESC will also be performed using this motor thrust test-bench. Commercially available ESCs lack any useful documentation. While it is not a major concern in hobby applications that is was developed for, the lack of relevant data has made the development of Quadrotors difficult. Most research groups have used custom made speed controllers, with the exception of the STARMAC group (), however little work has been done on what is needed from these controllers. Gurdan et al. has suggested that a control system that operates at 1 kHz provides superior stability than other slower systems that operate at 50Hz().

 The motor test bench will provide valuable data regarding the performance of the motors. The test bench can be built using spare parts from the Quadrotor design, including the motor, blades, Arduino, carbon fiber square rod and spare aluminum. The only extra piece of equipment necessary is a pressure sensor. An FC22 Compression Load Cell (see Figure 26: FC22 Compression Load Cell) is available from Digikey for $55.61 (). Cheaper alternatives include finding a digital scale that supports an RS-232 connection.

(38)
(39)

Power Management

 The Quadrotor design contains many advantages over other aerial vehicles. Low power consumption is not one of them. The power required to run four motors simultaneously in addition to the electronics often prevents the Quadrotor from staying aloft for more than 15 minutes. Even though the IARC only requires the vehicle to be operational for 10 minutes, it would be useful to do an examination of the power consumption of the Quadrotor. This will require use of accurate multi-meters, which are freely available in the labs. The motor thrust test-bench will allow for accurate power consumption statistics from the rotor assembly, while the electronic control systems and sensors can be measured prior to flight. An additional system may be developed to record power consumption during flight.

While a power consumption recording system is developed, battery life can be monitored with a simple device. A cheap battery monitor that emits a warning buzzer when the battery is low on power is available from HobbyKing.com for $4.95 (). The Maxpro Battery Monitor 3S (see Figure 27: Maxpro Battery Monitor 3S) will be useful during the early stages of  development. Special care must be taken when using LiPoly batteries. LiPoly batteries can suffer permanent damage if the voltage on the battery drops below 3V per cell. Battery monitoring should not be taken lightly, even early in development. In the later stages, a more sophisticated battery monitoring system can be developed. A special battery charger is recommended, but not required. This can be purchased at a later date if it is deemed necessary.

(40)

Tethered Flight Rig

Prior to full-flight, the communication system and stability systems must be thoroughly tested to avoid damage to the Quadrotor and its surroundings. A simple tether rig will enable the Quadrotor to be tested without incurring any damage from crashes or malfunctions. This rig will both hold the Quadrotor above the ground to prevent falls and pull the Quadrotor down to prevent it from flying into the ceiling. A diagram of this rig is seen below in Figure 28:  Tethered Rig Render. This can be constructed out of cheap materials such as

wood and some rope.

Figure 28: Tethered Rig Render

 Tools

 The construction of the Quadrotor will require access to machining equipment: A saw capable of cutting metal, a hand drill, a drill press, and a rotary tool. These tools are common and do not need to be purchased. However, a suitable workspace is required with sufficient ventilation. Ideally, once the Quadrotor has been constructed, the machining equipment would no longer be required.

Miscellaneous screwdrivers, wrenches, clamps, and hex keys will also be required, but similarly commonly available. Other tools recommended for construction include a hobby knife, metric rulers, electrical tape and zip-ties. Glues such as silicone or epoxy will be especially valuable in repairing damage.

References

Related documents

22) K. Zeeberg “Stromal Composition And Hypoxia Modulate Pancreatic Ductal Adenocarcinoma Pdac Cancer Stem Cell Behavior And Plasticity By Controlling The Angiogenic Secretome”. 23)

In addition, if I marked &#34;Yes&#34; to any of the above questions, I hereby authorize release of information from my Department of Transportation regulated drug and alcohol

b In cell B11, write a formula to find Condobolin’s total rainfall for the week.. Use Fill Right to copy the formula into cells C11

A weaker effect of incongruent relative to congruent audiovisual pairs was also observed in the left medial temporal cortex. Again, this suggests sensitivity to

The 18 interdisciplinary doctoral students interviewed in this study discussed their socialization experiences as occurring across several major themes: (a) their

While not seeking to ignore the scale of the problems in Gaza, we were concerned that immediate exposure to the full scale of the challenges facing the country would prove

cell type, page size, and block size, on system-level metrics such as performance, cost, and power consumption in various applications with different characteristics, e.g..

The broadcast count is defined as the number of transmissions (time-slots) re- quired to make a broadcast in a frame (to reach all the nodes if possible, or at least the maximum