Download Dri7 FILE COpy and more Summaries Computer Architecture and Organization in PDF only on Docsity!
Dri7 FILE COpy
NAVAL POSTGRADUATE SCHOOL
Monterey, California
NN
N SDSTATeS,
DT|C
ELECTE
THESIS -.
GUIDANCE AND CONTROL SYSTEM
FOR AN AUTONOMOUS VEHICLE
by Michael John Cloutier June, 1990 Thesis Advisor: (^) Yuh-jeng Lee Second Reader: John M. Yurchak Approved for public release; distribution is unlimited.
£ "! ',^ 08§
UNCLASSIFIED SECURITY CLASSIFICATION OF THIS PAGE
REPORT DOCUMENTATION PAGE 0M No. 0704-0IN
Ia. REPORT (^) SECURITY CLASSIFICATION l b RESTRICTIVE MARKINGS 2a. SECURITY^ UNCLASSIFIED CLASSIFICATION AUTHORITY 3. DISTRIBUTION /AVAILABILITY (^) OF REPORT 2b. DECLASSIFICATION/DOWNGRADING SCHEDULE^ ApprovedDistribution^ for^ publicis unlimited^ release;
- PERFORMING ORGANIZATION REPORT (^) NUMBER(S) 5. MONITORING ORGANIZATION REPORT NUMBER(S) 6.. NAME OF PERFORMING ORGANIZATION 6b. OFFICE (if applicable) SYMBOL 7a. NAME OF MONITORING ORGANIZATION Naval Postgraduate School (^) Code 39 Naval Postgraduate School 6c. ADDRESS (City, State, and ZIP Code) 7b. ADDRESS (^) (City, State, and ZIP Code) Monterey, CA 93943-5000 Monterey, CA 93943- Ba. NAMEORGANIZATION OF FUNDING /SPONSORING (^) Sb (^) (If OFFICE applicable) SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER Sc. ADDRESS (City, State, (^) and ZIP Code) 10. PROGRAM SOURCE OF FUNDING PROJECT NUMBERS TASK iWORK UNIT ELEMENT NO. NO. NO ACCESSION NO 11. TITLE (Include Security Classification) Guidance and Control System for (^) an Autonomous Vehicle
- CLOUTIER, PERSONAL AUTHOR(S) Michael j, 113a. TYPE OF REPORT 13b. TIME (^) COVERED 14. DATE OF REPORT (Year, Month, Day) **115. PAGE COUNT
- SUPPLEMENTARY**^ Thesis NOTATION^ TheFROM views^ expre-sedTO^ in this Junethesis^^1990 are^ those of the author^162 and do not reflect the (^) official policy or position of the Department of Defense or the U.S. Government. 17. (^) IELD GROUP COSATI (^) CODESSUB-GROUP 18. SUBJECT TERMS (Continue on reverse if necessary and identify by block number) Autonomous Vehicle; Autonomous Underwater Vehicle; AUV; Local Path Planning; Real-Time Control; Cubic (^) Spirals; 19. A (^) The CT Naval (Continue Postgraduate on reverse if necessary School and identify (NPS) (^) by is block number) currently involved (^) in a long-term project to investigatearchitecture and and developcontrol real-timesystems theorycontrol as software,they pertain artificial to U.S. intelligence,Navy autonomous (^) computer (^) vehicle programs. testbed autonomous In support (^) underwater of this vehicle.goal, the NPS is currently (^) designing and fabricating a
this Thistestbed^ work vehicledescribes which^ the usesdesign, portions^ development, (^) of cubic^ andspirals^ testing :as^ ofthe^ a-;Guidance desired path^ Subsystem to follow^ for between waypoints.control surfaces and Inmain addition, motors data translation firmware and real-time software (^) for the is designed, implemented and tested. support^ *^ The processof these^ ofgoals^ selecting is also^ and discussed^ implementing and detailed,an^ appropriate along^ withcomputer (^) the^ architecturechoice of associated^ in
- (^) -UNCLASSIFIED/UNLIMITEDDISTRIBUTION /AVAILABILITY (^) OF 0 ABSTRACT SAME AS (^) RPT DOTIC USERS 21. ABSTRACT UNCLASSIFIED SECURITY CLASSIFICATION 22a Y. NAME LeelJ. OF RESPONSIBLE Yurchak (^) INDIVIDUAL (^) 22b TELEPHONE (include Area Code)22c OFFICE SYMBOL DO Form 1473. JUN 86 Previous editionsare obsolete. 408-646-2361/390I SECURITY (^) CLASSIFICATION^ C/ILe^ OF C.SIY,1 THIS PAGE i UNCLASSIFIED
Approved for public release; distribution is unlimited.
Guidance and Control System for an Autonomous Underwater Vehicle
by
Michael John Cloutier
Lieutenant, United States^ Navy
B.S., Chapman College, 1982
Submitted in partial fulfillment
of the requirements for^ the^ degree^ of
MASTER OF SCIENCE IN COMPUTER SCIENCE
from the
NAVAL POSTGRADUATE SCHOOL
June 1990
Author:^2 Mic^ ke4Jo/hn^ Cloutierc^ -¢/x
Approved by:
Yuh-jeng Lee, Thesk Advisor
06nM. Yurcha^ Second^ Reader
Robert Mc3hee, Chairman
Department of Computer Science
iii
ABSTRACT
The Naval Postgraduate School (NPS) is^ currently^ involved^ in^ a^ long-term
project to investigate and develop real-time control software, artificial intelligence,
computer architecture^ and control^ systems^ theory^ as^ they^ pertain^ to^ U.S.^ Navy
autonomous vehicle^ programs.^ In^ support^ of^ this^ goal,^ the^ NPS^ is^ currently^ designing
and fabricating a testbed autonomous underwater vehicle.
This work describes the^ design,^ development,^ and^ testing^ of^ a^ Guidance
Subsystem for this testbed vessel which utilizes portions of cubic spirals as the desired
path to follow between waypoints.^ In^ addition,^ data^ translation^ firmware^ and real-
time control software for^ the^ control^ surfaces^ and^ main^ motors^ is^ designed,
implemented and tested.
The process of selecting and implementing an appropriate computer
architecture in^ support^ of^ these^ goals^ is^ also^ discussed^ and^ detailed,^ along^ with^ the
choice of associated computer hardware and real-time operating system software.
iv
LIST OF FIGURES
- INTRODUCTION
- A. OBJECTIVES
- B. BACKGROUND - 1. Vehicle Characteristics
- AUV Software Classification
- Real Time Vehicle Control Requirements
- Guidance Subsystem Requirements
- C. THESIS ORGANIZATION
- II. RELATED W ORK
- A. EDUCATIONAL INSTITUTIONS - 1. Texas A&M
- University of Florida
- Georgia Institute of Technology
- University of New Hampshire (UNH)
- MIT - Sea Grant College
- University of California - Santa Barbara
- B. INDUSTRY AND GOVERNMENT - 1. FMC Corporation - 2. R ockwell
- M artin-M arietta
- National Institute of Standards and Technology
- Woods Hole Oceanographic Institute
- N O SC
- International Submarine Engineering
- C. SUM M ARY
- III. EVOLUTION OF ONBOARD COMPUTER SYSTEM
- A. GRiDCASE 1535 EXP - 1. Data Translation Boards
- Real-Time Operating System
- a. TSR Program
- b. Regulus Operating System
- Hardware Problems
- B. G ESPA C - 1. Target System
- C. AUV II DISTRIBUTED COMPUTER SYSTEM
- D. DATA TRANSLATION SOFTWARE - 1. Digital-to-Analog Routines - 2. Analog-to-Digital Routines
- E. CONTROL OF HYDRODYNAMIC SURFACES - 1. Servo Control Testing - 2. Servo Control Signal Calculation - 3. Manipulation of Control Surfaces
- F. MAIN MOTOR CONTROL - 1. Main Motor Control Testing - 2. Main Motor Control Signal Calculation - 3. Control of Main Motors
- IV. GUIDANCE SUBSYSTEM
- A. INPUTS AND OUTPUTS
- B. SUBSYSTEM COMPOSITION
- C. INTERFACE WITH OTHER SOFTWARE MODULES - 1. M ission Planner - a. Grid Size Determination - b. Output from Mission Planner - 2. A utopilot - 3. Navigation
- D. OFF-LINE LOCAL PATH PLANNER - 1. Reference Heading - 2. Cubic Spiral Calculation
- E. ON-LINE LOCAL PATH PLANNER - 1. Cardinal Heading Maneuvers - 2. Theory of Operation - 3. W aypoint Processing - 4. Reference Value Calculation
- F. TRACKING CONTROLLER - 1. Basic Operation - 2. Error Value Calculation - 3. Desired Value Calculation - 4. Look-Ahead Limitation
- G. TESTING RESULTS FOR TWO-DIMENSIONAL GUIDANCE
- H. THREE-DIMENSIONAL CALCULATIONS - 1. Depth Descriptors - 2. Modifications to On-Line Local Path Planner
- VI. CONCLUSIONS AND RECOMMENDATIONS
- A. CONTRIBUTIONS - 1. Vehicle Control Computer System - 2. Guidance Subsystem - 3. Vehicle Firmware
- B. RECOMMENDATIONS FOR FUTURE WORK - 1. Real-Time scheduling
- Guidance Subsystem
- Communication
- Vehicle Firmware
- APPENDIX A. GRiD TIMED INTERRUPT SOURCE CODE
- APPENDIX B. DATA TRANSLATION SOURCE CODE
- APPENDIX C. SERVO CONTROL CODE
- APPENDIX D. MAIN MOTOR CONTROL CODE
- APPENDIX E. GLOSSARY
- APPENDIX F. OFF-LINE GUIDANCE SOURCE CODE
- APPENDIX G. REAL-TIME GUIDANCE SOURCE CODE
- APPENDIX H. THREE DIMENSIONAL SOURCE CODE
- LIST OF REFERENCES
- BIBLIOGRAPHY
- INITIAL DISTRIBUTION LIST
- Figure 1 - Basic Equipment Layout [GOOD 90]
- Figure 2 - Control Systems Software Architecture
- Figure 3 - Proposed Hardware Configuration
- Figure 4 - Pulse-Width Modulator Schematic
- Figure 5 - Servo Test Setup
- Figure 6 - Raw Servo Control Voltage
- Figure 7 - Motor Controller Schematic
- Figure 8 - Main Motor Test Setup
- Figure 9 - Raw Main Motor Voltages
- Figure 10 - AUV 11 Software Modules
- Figure 11 - Two-Dimensional Grid for AUV II
- Figure 12 - Allowable Maneuvers
- Figure 13 - Calculation of Desired Heading
- Figure 14 - Example of a LLL Maneuver
- Figure 15 - Guidance Subsystem Performance
ACKNOWLEDGEMENTS
No undertaking as complex as a thesis can be completed without the combined
effort, expertise, cooperation and understanding of many people. I am thankful to
LCDR John M. Yurchak, a true 'C wizard', who gave me an appreciation of how
effective properly written code can be. I am also indebted to Jim Bellingham for his
cooperation and assistance with the Vehicle Control Computer software, without his
help and advice I would still be lost. I am especially grateful to Yutaka Kanayama.
His impressive knowledge of vehicle path-planning methods, combined with his
extraordinary ability^ to^ clarify^ the^ most^ complex^ of^ problems,^ were^ instrumental^ in
the development of the Guidance subsystem. His willingness to listen, and the inner
peace which he radiates will remain with me always.
Lastly, I would like to acknowledge my wife, Chris, for her understanding, and
the love and support which she provided while I labored on this difficult project.
Without her sacrifices, this thesis would not have been possible.
xiii
" They can perform tasks which are (^) considered too dangerous for humans to accomplish. " They can operate at (^) tremendous depths, and can maneuver without (^) regard for human physiological limitations. Along with (^) these advantages come disadvantages, primarily the lack of human control and the inability to intervene in the event of unforeseen problems. To compensate for these shortfalls, a successful AUV must be able to incorporate the advantages (^) of artificial intelligence, (^) real-time control, environmental sensing and maneuverability into a compact, integrated package. The development of autonomous underwater vehicles which have these characteristics, and are capable of operating under a wide variety of situations and operating conditions, is an area of intense and diverse investigation. A prime problem in fielding an operational (^) AUV is the requirement for "...robust software to maintain vehicle integrity while accomplishing mission goals ...." [BELLINGHAM 89] At the Naval Postgraduate School (NPS), two AUVs have been constructed in an attempt (^) to address some of this issues. The first NPS submersible, designated AUV 1, was 27 inches long and displaced less than 20 pounds, which allowed operation (^) in a 40 foot long test tank. This vehicle had an umbilical to provide power to onboard (^) control systems and had no onboard computer system. Vehicle control (^) was provided by an off-hull PC AT microcomputer which transmitted plane commands through radio (^) control equipment to the AUVs
dive planes. [BRUNNER 88]
In order to allow completely autonomous behavior a second AUV (designated AUV II) is under construction. AUV II will support analysis of vehicle dynamics and
control, real-time^ control^ systems^ theory, higher^ level^ artificial^ intelligent^ processing, robotics and computer architecture for future large AUVs. This research is being conducted as part of a long-term study funded and sponsored by the Naval Surface Weapons Center, White Oak, Maryland [HEALY 89b].
A. OBJECTIVES This thesis covers^ several^ tasks^ in^ support^ of^ the^ initial^ launch^ of AUV^ II including: (1) Selection of the Vehicle Control computer and data translation hardware and real-time computer operating systems which meet the requirements specified above (2) Design, development and testing of the vehicle Guidance subsystem (3) Development and testing of real-time control software for the control surfaces and main motors on AUV II.
B. BACKGROUND
1. Vehicle Characteristics The basic layout of AUV-2 is illustrated in Figure 1. The main vehicle body is constructed as a rectangular aluminum box^ with^ a^ beam^ width^ of^16 inches, a height of 10 inches, and a length overall (LOA) of 93 inches. It is anticipated that the finished vehicle will displace approximately 370 pounds.
propulsion and the hydrodynamic forces on the control surfaces to maintain desired
pitch, roll and yaw rates. [KWAK 90]
There is a sonar system consisting of four individual pencil-beam sonar
transducers mounted in a flooded nose, an inertial sensor suite ( consisting of a flux
gate compass, three rate gyros, three accelerometers, a vertical gyro and a directional
gyro), a differential pressure sensing depth cell, a paddle wheel speed sensor, and
individual motor RPM sensors. The overall combination of sensors and effectors
results in a minimum requirement of 12 separate analog control output signals and
23 different analog input signals for AUV-II (as summarized in Table I).
There were numerous goals established for the development of AUV II.
The vessel computer system had to be low cost, utilizing off-the shelf hardware where
possible, be readily reconfigurable for different missions, and be expandable to multi-
processor capability at a later date. The operating system and control software had
to be easy to implement (preferably in the C language), had to use a commercially
available operating system, must be able to simulate complex behavior, should
support multi-tasking and should be easy to reconfigure for different missions.
2. AUV Software Classification
Software for NPS AUV-Il can be divided into two broad classes: Mission
Planning software (off-vehicle non-real-time software) and Vehicle Control software
(real-time on-vehicle software).
Mission Planning is the process of determining a reference path to be
followed by the AUV based on knowledge of pre-existent obstacles, threats and other
Table I - Analog Input and Output Signals of AUV II
ANALOGSIGNALS INPUT SIGNALSANALOG OUTPUT Port and Starboard Bow planes (2) Control Surface Position (5) Port and Starboard Stem planes (2) Propulsor RPM (6) Fore and aft Rudders (2) Pitch/Pitch Rate (2) Port and Starboard main engines (2) Roll/Roll Rate (2) Fore and aft horizontal thrusters (2) Yaw/Yaw Rate (2) Fore and aft vertical thrusters (2) Depth (1) Sonar Channels (4) Speed (1) Health and Well-being (4)
possible hazards to the vehicle. The Mission Planner constructs a path as a series of waypoints for the vessel to reach. Development of the Mission Planner for AUV II
is described in [ONG 90]. This obstacle-free path is downloaded to the vehicle via a detachable RS-232 serial link. The Vehicle Control software utilizes the path information from the Mission Planner and controls the vehicle from point of origin to mission goal. As the mission progresses, the Vehicle Control system analyzes all input signals from onboard sensors and provides output signals to control surfaces, thrusters and main
motors while piloting the vessel to the desired goal.
