INTEGRATION AND EVALUATION OF SENSOR MODALITIES FOR POLAR RO

INTEGRATION AND EV ALUATION OF SENSOR MODALITIES FOR POLAR ROBOTS

By

Richard S.Stansbury

B.S.-Computer Engineering

University of Kansas,2002

Submitted to the Department of Electrical Engineering and Computer Science and the Faculty of the Graduate School of the University of Kansas

in partial ful?llment of the requirements for the degree of Master of Science

Prof.Arvin Agah(Committee Chair)

Prof.Victor Frost(Committee Member)

Prof.Costas Tsatsoulis(Committee Member)

Date of Acceptance

ABSTRACT

As part of the Polar Radar for Ice Sheet Measurement(PRISM)project,at the University of Kansas,a mobile robot was designed constructed to support radar measurements in Greenland and Antarctica.The PRISM mobile robot was constructed using an ATV as a mobile base vehicle and building a protective enclosure to support computer and radar equipment.

This research has focused on the integration and evaluation of sensor modalities to support the PRISM rover,as well as other polar robotic endeavors.

A general focus was to?nd commercial o?-the-shelf(COTS)sensors.Given the rover’s sensing needs and the prior work on planetary and polar robots,a suite of sensors was selected where each sensor met the needs of the PRISM project.The sensors were integrated with the PRISM rover.Integration of the sensors involved three main tasks of physical mounting,connectivity,and software integration. Mounting involved both external mounting of sensors as well as the construction of rack-mount cases for internal sensors.Connectivity involved connecting all of the sensors with the necessary data links so that the information is properly shared.Software integration included the development of a software API for the PRISM sensors.Sensor fusion also took place within the software to provide additional functionality.

The PRISM rover’s sensor suite was next validated to demonstrate the survivability of the sensors,measure properties of the sensors not available from the speci?cations and data sheets,and verify correct operation,integration,and mounting of the sensor suite.

ii

ACKNOWLEDGMENTS

Special thanks goes to my adviser and committee chair Professor Arvin Agah.His knowledge and support has been invaluable.I would also like to thank Professor Victor Frost and Professor Costas Tsatsoulis for serving on my committee.I look forward to working with all of you as I proceed forward toward my Ph.D.in the coming years.

Without the assistance of my fellow roboticists,Hans Harmon and Eric Akers,this thesis would not be possible.I’d also like to thank Torry Akins for dispensing advice whenever needed,or otherwise.I would also like to thank Dennis Sundermeyer for his assistance with fabricating whatever case,mount,or gizmo that I needed.I would also like to thank the students and faculty of the PRISM project whom I have worked with both o?and on the ice.Special thanks also goes out to MARVIN,the not so little robot that could.

I would like to thank my wife,Amy,for all of her support over the past ?ve years.Without her patience,care,and guidance,I would not be where I am today.I thank my family and friends for all of their support as I continue forward into academia.

This work was supported by the National Science Foundation(grant#OPP-0122520),the National Aeronautics and Space Administration(grants#NAG5-12659and NAG5-12980),the Kansas Technology Enterprise Corporation,and the University of Kansas.

iii

CONTENTS

LIST OF FIGURES (ix)

LIST OF TABLES (xi)

1.INTRODUCTION (1)

1.1Motivation (2)

1.1.1Polar Traversal (2)

1.1.2Polar Radar for Ice Sheet Measurement(PRISM)Project.3

1.1.3The PRISM Mobile Robot (5)

1.2Approach (6)

1.3Thesis Structure (7)

2.BACKGROUND (8)

2.1Navigation and Localization (8)

2.1.1Position Sensors (9)

2.1.1.1Dead Reckoning (10)

2.1.1.2Triangulation Using Land-Based Beacons (11)

2.1.1.3GPS (11)

2.1.2Heading and Orientation (13)

2.1.2.1Compasses (14)

2.1.2.2Gyroscopes (14)

2.1.2.3Inclinometers (14)

2.1.2.4GPS (15)

2.2Collision Detection and Avoidance (16)

2.2.1Obstacle Detection (16)

2.2.1.1Machine Vision (17)

2.2.1.2IR Detectors (18)

2.2.1.3Sonar (19)

2.2.1.4Laser Range Finders (19)

2.2.1.5Millimeter Wave Radar (20)

2.2.2Crevasse and Sastrugi Detection (21)

iv

2.2.2.1Traditional Obstacle Detection Sensors (21)

2.2.2.2Ground Penetrating Radar (22)

2.2.3Collision Detection (22)

2.2.3.1Bump Sensors (23)

2.2.3.2Accelerometers (23)

2.3Proprioception (24)

2.3.1Fuel Sensors (24)

2.3.2Power Levels (24)

2.3.3Internal Climate (25)

2.3.4External Climate (25)

2.4Sensor Fusion (25)

2.4.1Algorithms and Approaches (26)

2.4.2Neuroscience and Behavior-based Robotics (27)

2.4.3Sensor Fusion for Visualization and Mobile Robot Tele-

operation (27)

3.RELATED WORK (28)

3.1Planetary Rovers (28)

3.1.1Challenges (28)

3.1.2Field Integrated Design and Operation(FIDO)rover (29)

3.1.3Sample Return Rover(SRR) (30)

3.2Polar Rovers (30)

3.2.1Dante I (31)

3.2.2Nomad (31)

3.2.3Robot Antarctico di Super?cie-ENEA (32)

4.SENSOR SELECTION (33)

4.1Sensing Requirements (33)

4.1.1Task Requirements (33)

4.1.2Environmental Requirements (34)

4.1.3Outreach Requirements (34)

4.2Criteria for Selection (34)

4.2.1Cost (34)

v

4.2.2Power Consumption (35)

4.2.3Ruggedness (35)

4.2.4Accuracy (35)

4.2.5Size/Weight (35)

4.2.6Complexity and Reliability (35)

4.2.7Software Interface (36)

4.2.8Physical Mounting (36)

4.2.9Connectivity (36)

5.ANALYSIS OF SENSOR SUITE (37)

5.1Topcon Legacy-E GPS System (37)

5.2SICK LMS221Laser Range Finder (40)

5.3BEI Technologies MotionPak II Gyroscope (42)

5.4Precise Navigation Inc.TCM2-50 (44)

5.5Rainwise WS-2000Weather Station (45)

5.6Pelco Esprit Pan-Tilt-Zoom Camera (47)

5.7Cruz-Pro TL30Fuel Gauge (49)

6.SENSOR INTEGRATION (51)

6.1Sensor Mounting (51)

6.1.1External Sensor Mounting (51)

6.1.1.1WS-2000 (51)

6.1.1.2GPS Antennas (52)

6.1.1.3Pelco Pan/Tilt Camera (53)

6.1.1.4MotionPak II (53)

6.1.1.5Sick LMS221 (53)

6.1.2Internal Sensor Mounting (55)

6.1.2.1GPS Cases:Base and Rover (56)

6.1.2.2Sensors Case (57)

6.1.2.3Power Case (59)

6.2Sensor Connectivity (59)

6.2.1Serial RS-232Connectivity (60)

6.2.2Network Connectivity (62)

vi

6.2.3Computing (62)

6.3Software Integration (62)

6.3.1Overview of PRISM Sensor API (63)

6.3.1.1Basic Sensor Interfaces (64)

6.3.1.2Events and Event Listeners (65)

6.3.2Instantiation of API for PRISM sensors (66)

6.3.2.1TopconGPSReceiver.java (66)

6.3.2.2SickLaserRangeFinder.java (67)

6.3.2.3Motionpak2.java (68)

6.3.2.4TCMTiltTempSensor.java (68)

6.3.2.5WeatherStation.java (68)

6.3.2.6TL30FuelSensor.java (68)

6.3.3Sensor Fusion (69)

6.3.3.1Position2HeadingSensor.java (69)

6.3.3.2MarvinHeadingSensor.java (70)

6.3.3.3Waypoint Navigation (70)

7.FIELD EXPERIMENTS (73)

7.1Goals (73)

7.2Greenland2003Field Season (74)

7.3Local Experiments (75)

7.4Greenland2004Field Season (75)

8.EXPERIMENTAL SETUP (77)

8.1Climate Survivability (77)

8.2GPS Performance (78)

8.2.1Measurement of Relative Accuracy (78)

8.2.2Satellite Visibility (79)

8.2.3GPS Measurement Stability (80)

8.3Obstacle Detection (81)

8.4Orientation Measurements (82)

8.4.1Drift (82)

8.4.2Vibration (83)

8.5Waypoint Navigation (84)

vii

9.RESULTS AND ANALYSIS (86)

9.1Climate Survivability (86)

9.2GPS (88)

9.2.1Relative Accuracy (88)

9.2.2Visibility (89)

9.2.3Stability (90)

9.3Obstacle Detection (92)

9.3.1Con?guration#1:Positive and Negative Obstacle Detector93

9.3.2Con?guration#2:Positive-only Obstacle Detector (93)

9.3.3Discussion (99)

9.4Orientation (100)

9.5Waypoint Navigation (101)

10.CONCLUSION (104)

10.1Contributions (106)

10.2Limitations (106)

10.3Future Work (107)

REFERENCES (109)

viii

LIST OF FIGURES

1.1PRISM mobile robot (5)

5.1Topcon GPS base station (38)

5.2Sick LMS221laser range?nder (40)

5.3MotionPak II gyroscope/accelerometer (42)

5.4TCM2-50orientation sensor (45)

5.5Rainwise WS-2000weather station (46)

5.6Pelco Esprit (48)

5.7TL30digital fuel gauge (50)

6.1External mounting of sensors on PRISM Rover (52)

6.2LMS221con?gurations (54)

6.3Sick LMS221con?guration (55)

6.4Sensor cases mounted within PRISM rover (56)

6.5Rack-mountable case for rover GPS (57)

6.6Portable case for GPS base station (58)

6.7Rack-mountable case for sensors and sensor components (59)

6.8Rack-mountable case for power supplies (60)

6.9PRISM robotic sensor connectivity (61)

6.10Edgeport serial hub (61)

6.11Netgear ethernet switch (62)

6.12GoBook Max ruggedized laptop (63)

6.13PRISM robotic sensor API?owchart (64)

6.14Flowchart of waypoint navigation system (72)

ix

8.1Rover’s waypoint path for bistatic SAR (85)

9.1Topcon GPS:24-hour satellite visibility (90)

9.2Topcon GPS:measurement standard deviation vs.distance to base.92 9.3LMS221:igloo (94)

9.4LMS221:snow pit (95)

9.5LMS221:snow mobile (96)

9.6LMS221:sastrugi (97)

9.7LMS221:?ags and power line (98)

9.8Plot of rover’s path while performing waypoint navigation (102)

9.9Plot of rover’s path while performing waypoint navigation(zoomed

out) (103)

x

LIST OF TABLES

5.1Topcon Legacy-E and PDL Data Link speci?cations (39)

5.2Sick LMS221speci?cations (41)

5.3BEI MotionPak II speci?cations (43)

5.4PNI TCM2-50speci?cations (45)

5.5Rainwise WS-2000speci?cations (47)

5.6Pelco Esprit speci?cations (49)

5.7CruzPro TL30speci?cations (50)

9.1Topcon GPS:known UTM position(in meters) (88)

9.2Topcon GPS:measured UTM position(in meters) (88)

9.3Topcon GPS:relative positioning error(in meters) (89)

9.4Topcon GPS:average satellite visibility at NGRIP (91)

9.5TCM:stability and response to vibration (100)

9.6MotionPak II:stability and response to vibration (100)

xi

1.INTRODUCTION

With the possible melting of the polar ice caps and the gradual rise in sea level, the polar regions of the Arctic and Antarctica have become key targets for earth science researchers.The potential impacts from these ecological changes are still not well understood.By monitoring changes to the ice sheets,researchers may be able to develop and improve theories regarding the long-term impacts and how the ice sheets play a role globally.Due to the adverse conditions of polar environments,?eld research and the collection of data can be quite di?cult.Mobile robots provide useful means for automating the collection of research data in the?eld by reducing the risks due to human involvement and improving accuracy.

At the University of Kansas,the Polar Radar for Ice Sheet Measurement (PRISM)project[61]focuses on the task of developing a radar system capable of measuring the thickness and other characteristics of the polar ice sheets.To accomplish this goal,we are developing an autonomous mobile robot.It is our goal to utilize the PRISM rover to operate in both Greenland and Antarctica with limited supervision(refueling and remote monitoring).

In the past,robotics research has been focused on developing robust sensor suites and algorithms in order to provide reliable,robust,precise,and safe autonomy.However,less research has been conducted regarding sensing and navigation for robotic vehicles on an ice sheet.Sensor modalities must be fully tested and evaluated for operations on the ice sheets to ensure that the robot will have the same functionality,reliability,and precision that is expected in other types of environments.

1

1.1Motivation

Extensive research exists in the area of mobile robotics and sensors.This research addresses a speci?c application area that presents unique challenges.When designing a mobile robot for working in polar environments,special care must be taken with the selection of sensors with a focus on both survivability and reliability.

In this section,the challenges of operating in polar environments is discussed. We then provide a brief overview of the PRISM mobile robot that has been designed developed to address these challenges.An overview of the approach to address these issues is also presented.

1.1.1Polar Traversal

Polar research often relies on data collection and traversal of the ice sheets of Greenland and Antarctica.In June of2001,a workshop analyzed the feasibility and resourcefulness of using mobile robots similar to planetary rovers for collection of scienti?c data on the ice sheets[13].This group de?ned several tasks with which an autonomous rover could aide,including:traverses with detailed and tedious paths,extremely remote and/or inhospitable environments,data collection parallel to a manned traversal,and data collection at slow speeds.In2002, researchers at NASA’s Jet Propulsion Laboratory proposed additional research that proposed that utilizing mobile robotics for polar exploration can provide future bene?ts toward planetary exploration[7].

Both groups discussed the numerous challenges that exist with polar traversals.These challenges include deep and blowing snow,crevasse detection and avoidance,sastrugi detection and avoidance,and limited supervision.Several of these challenges directly relate to the sensing capabilities of an autonomous rover,which will be addressed in this thesis,along with a number of additional challenges.

2

Sensing in the polar regions can be more challenging than in more hospitable environments including sub-zero temperatures and high wind speeds.Equipment may be damaged as a result of such temperatures and wind speeds.

Vision can be limited due to the lack of contrast on the icy surface.With this lack of depth perception,natural obstacles are often overlooked.For instance, sastrugies,snow drifts that form as a result of wind erosion,have heights of up to one meter.In addition,crevasses in the ice sheet are encountered that remain invisible because they are covered in snow.

Sensors for positioning and localization also experience di?culties in the polar 7d21773343323968011c92dfpasses are less e?ective due to their close proximity to the magnetic poles.GPS satellite coverage near the polar regions is reduced because the satellite network focuses primarily on more populated or strategic areas. If there are not enough visible satellites,GPS position accuracy is reduced,or position information becomes unavailable.

Despite these limitations,research in polar environments has proven invalu-able.For the PRISM project,the radar images that are taken of the ice are used to support glaciologists and geophysicists who address research areas such as global warming.Atmospheric researchers can utilize data collected from ice cores and other measurements to uncover the history of the planet’s climate and its geological history over thousands of years.By developing a robust mobile robot for polar environments,we can uncover details that will support future research in polar environments.

1.1.2Polar Radar for Ice Sheet Measurement(PRISM)

Project

The Polar Radar for Ice Sheet Measurement project is currently underway at The University of Kansas[61].The project’s goal is to develop radar systems to measure polar ice sheet properties in order to accurately determine their

3

mass balance and other features.Such data will help researchers determine the contributions of polar ice sheet melting to global climate change and rising sea levels.

Two radar systems have been developed for this task[23].A monos-tatic/bistatic synthetic aperture radar(SAR)has been developed to generate a two-dimensional view of basal conditions of the Polar Ice Sheet.A wide-band, dual mode radar will also be utilized to determine aspects of ice sheet depth and accumulation[23].

In order accommodate these two radar systems,two vehicles have been employed.The?rst is a manned tracked vehicle that will tow an antenna array for the bistatic/monostatic SAR and also carry an on-board dual-mode radar.A second uninhabited rover will be utilized to carry a second SAR.When operating in bistatic mode,the rover will operate as a receiver.The two vehicles will move along side one another in a coordinated pattern.The autonomous rover will traverse a wide area along side the tracked vehicle collecting data to build the radar image.

Intelligent on-board systems are responsible for coordinating the two vehicles.Based on data received from the radars and the current state of the autonomous rover,this system will indicate areas of interest and assign autonomously waypoints for coordinated navigation.

In addition,the PRISM project has an extensive educational outreach program directed at both K-12students and the general public.The outreach opportunities also include collaboration with Haskell Indian Nations University in Lawrence,Kansas[25].Outreach activities include maintaining a website with polar news and live updates from PRISM team members from the?eld[11,61].

4

1.1.3The PRISM Mobile Robot

The research presented in this thesis directly facilitates the goal of PRISM to develop a semi-autonomous mobile robot[1,24,56].This robot is responsible for the transportation of the radar system and the computing equipment across polar ice sheets.It will primarily operate autonomously with the ability for a user to override the control system and operate it remotely.In the past,this robot has been also refereed to by the acronym“Mobile Antarctic Robotic Vehicle with Intelligent Navigation,”or MARVIN.For this thesis,it shall be referred to as either the PRISM mobile robot or the PRISM rover.

Figure1.1:PRISM mobile robot.

The PRISM rover,as shown in Figure1.1,is the result of our e?ort[1,24,56].

A signi?cantly modi?ed version of a six-wheeled Bu?alo All Terrain Vehicle(ATV)

5

[27]was used as the mobile base.Actuators were utilized for controlling the throttle and steering.An enclosure was constructed on top of the base platform to provide a protected environment suitable for computing equipment.To facilitate intelligent and reliable mobility,a suite of sensors were integrated to provide the vehicle with perception capabilities.These sensors will be discussed in more detail throughout this thesis.

1.2Approach

The approach used to address the challenges related to the sensor suite for the PRISM mobile robot was performed in a number of steps.The?rst step involved an extensive literature search.By reviewing existing literature regarding sensors, a list of potential sensors was compiled.In addition,a literature search was performed to study existing research projects that utilize mobile robots for polar or planetary applications.By learning from the work of past researchers,some sensor modalities can be eliminated based on previously known weaknesses encountered in these environments.

The next main task was to de?ne the criteria for the sensors that were to be selected for the PRISM rover.Project related criteria were de?ned such as the needs of other researchers and their equipment.In addition,the educational outreach requirements were de?ned.Since the vehicle will be operating in a hostile environment,environmental requirements such as operating temperature range and ruggedness were also de?ned.Based on the requirements de?ned,a suite of sensors were selected.

Upon selecting the suite of sensors,each sensor was integrated into the PRISM polar robot.A software API was developed in Java.Drivers were written for each sensor based on the API.Fusion of sensor data was also performed to create robust virtual sensors.Sensor connectivity was also addressed such that multiple sensors could be connected to a single computer.Finally,the task of

6

physically mounting each sensor at the point where it would be most a?ective was addressed.

Once the sensors were integrated into the PRISM rover,series of experiments were designed and performed to evaluate the sensor suite’s performance.These experiments were performed both locally at the University of Kansas and on Greenland’s ice cap.These experiments were designed to verify important sensor speci?cations,determine the a?ects of locality on sensor performance, and determine any potential weaknesses that must be addressed for future?eld experiments.

1.3Thesis Structure

This thesis is pided into10chapters.Chapter2presents the background knowledge relative to this work.In particular,three major areas of robotic sensing: navigation and localization,collision detection and avoidance,and proprioception are presented.Sensor fusion is also discussed.Chapter3discusses related work in the areas of planetary rovers and polar rovers.Chapter4presents the sensing requirements and selection criteria used for developing the PRISM rover.Chapter 5describes the entire sensor suite for the PRISM rover and discusses how each sensor relates to the requirements and selection criteria.Chapter6discusses how the sensors were integrated into the PRISM rover.Integration was pided into three areas of mounting,connectivity,and software.

Chapters7,8,and9discuss the validation of the sensor suite.Chapter7 discusses the evaluation goals and provides an overview of experiments performed locally and during the2003and2004Greenland?eld seasons.Chapter8discusses speci?c experiments that were performed to validate the sensors.Chapter9 presents the results and their analysis.

Chapter10concludes this thesis,presenting the contributions,limitations, and the future work.

7

2.BACKGROUND

This research focuses on the selection,integration,and validation of sensors for polar robots.Many sensor modalities exist to support autonomous and semi-autonomous mobile robots.The choice of sensors to integrate is problem-dependent.

The desired level of autonomy is a major factor in?uencing the sensor selection.A fully autonomous mobile robot must be capable of sensing its environment,interpreting the sensor data,and carrying out actions to achieve its goals.Vehicles that are teleoperated have no on-board autonomy and rely on commands from a remotely situated operator.Vehicles with supervisory control are between these extremes.They have some autonomy;however,their state is monitored remotely.A remote operator is capable of assigning new tasks, providing guidance,or temporarily taking control of the robot.The amount of control given to an outside controller is mission dependent[22].

The PRISM project’s mobile robot is semi-autonomous.It operates autonomously once a set of goals is assigned.The goals can be formulated by the rover’s own control system,or assigned by a human operator.A remote operator is also capable of overriding the autonomous control of the vehicle.To select sensors that facilitate the needs of the PRISM polar robot,sensors for mobile robots were investigated,along with the topic of sensor data fusion.

2.1Navigation and Localization

A semi-autonomous mobile robot must be capable of traveling from one point to another point.Within its current environment,the mobile robot must be capable of determining its location relative to the target destination(localization).It

8

must also be capable of determining the proper change in trajectory such that it can travel toward the destination.The sensing requirements for navigation and localization can be classi?ed into two categories:position and heading/orientation.

2.1.1Position Sensors

Determining the mobile robot’s current position plays a dual role for the PRISM project.Intelligent navigation requires knowledge of the rover’s location within the current world model.Location information could involve the precise location on a local or global map,or it could involve the distance traveled since its movement.

The robot’s position also plays a role in the scienti?c objectives of the PRISM project.Position information is shared both with the radar and the intelligent systems.Precise positioning is necessary to form an accurate radar image.

Position is determined either locally or globally.Locally,relative positioning determines the position of the bot within a local region relative to some?xed point within that region.Globally,absolute positioning determines the robot’s precise location with respect to a geodetic world model.A geodetic world model is a representation of position on a coordinate system that represents the world. For instance,Latitude,Longitude,and Altitude may be utilized.The Universal Transverse Mercator(UTM)coordinate system is another common geodetic world model that pides the Earth’s surface in to grids where position is represented as x,y,z coordinates within these grids.Relative positioning satis?es the requirements of the PRISM project.

In this section,various positioning sensors will be discussed with focus on how these sensors may be used for relative positioning.The bene?ts and limitations of each sensor are also described.

9

2.1.1.1Dead Reckoning

Dead reckoning approximates the position of the mobile robot by measuring its rate of motion and determining the relative change in its position.Dead reckoning utilizes various devices such as shaft encoders,Doppler,and accelerometers.

Shaft or quadrature encoders are some of the least expensive and simplest devices used for dead reckoning.These sensors measure the rotational rate of the vehicle’s wheels or axles.Given these rates,the change in position and possibly orientation of the robot can be determined.Incremental encoders often use a photosensor to calculate the number of alternations of a pattern on a disk which rotates about the axis.Wheel slippage often results in error accumulation over time[31].

Doppler technology has been used for dead reckoning.Doppler-based sensors measure the time-of-?ight of ultrasonic or microwave bursts to determine the velocity of the 7d21773343323968011c92dfing this velocity,it is possible to approximate the vehicle’s position over time.Doppler is however susceptible to interference.Its accuracy is also reduced on rough terrain where sudden changes in elevation exist [20].

Accelerometers can also be employed.The vehicle’s acceleration must be doubly integrated in order to approximate the vehicle’s position.Similar to the Doppler technology,this sensor does not work well on rough terrain.The sudden impacts from the bumpy road will add noise to the acceleration measurements [20].

Dead reckoning provides decent estimates,but lacks accuracy.Error accumulation may become too great,and redundancy can likely improve accuracy. For applications requiring precise navigation,dead reckoning is infeasible.

10

本文来源：https://www.bwwdw.com/article/jsaq.html