Occupancy-Based Map Search

advertisement

Autonomous Flight Systems

Laboratory

Aeronautics & Astronautics

All slides and material copyright of

University of Washington Autonomous

Flight Systems Laboratory

Autonomous Flight Systems

Laboratory

Aeronautics & Astronautics

Research and Development at the

Autonomous Flight Systems Laboratory

University of Washington

Seattle, WA

Guggenheim 109, AERB 214

(206) 543-7748 http://www.aa.washington.edu/research/afsl

General Information

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

People

Dr. Rolf Rysdyk

Dr. Juris Vagners

Dr. Uy-Loi Ly

Dr. Kristi Morgansen

Dr. Anawat Pongpunwattana

Christopher Lum

Craig Husby

John Osborne

Richard Wise

Elizabeth Bykoff

Ben Triplett

Dan Klein

Jim Colito

Research Focus

• Multi-Vehicle Cooperative Control Flight Testing

• Cooperative Strategies for Teams of Autonomous Air & Surface Vehicles

• Probability Based Searching/Target Identification

• Coordinated Underwater Robotics

• Communications for Heterogeneous Cooperating Autonomous Vehicles

Mission Statement

To conduct research that advances guidance, navigation, and control technology relevant to Autonomous Vehicles.

University of Washington 3

Hierarchy of Autonomy

Autonomous Flight Systems Laboratory

Strategic (low bandwidth)

Path Planning

Task Allocation

Search Patterns

Human Mission Command

Tactical (medium bandwidth)

Target Observation

Path Following

Communication & Cooperation

Human Monitor Interaction

Aeronautics & Astronautics

Dynamics and Control (high bandwidth)

State Stabilization

Signal Tracking

Inner Loop or “autopilot”

Configuration changes

University of Washington 4

Topography of Autonomous Flight

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

University of Washington 5

Hardware-in-the-Loop Simulator

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

HiL Simulator

University of Washington

Avionics Tray

6

Hardware-in-the-Loop Simulator

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Groundstation

University of Washington

Aircraft

7

Distributed Real Time Simulator

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Five computers running

REAL TIME simulation software.

Used as a high fidelity testing environment to accurately simulate data transfer and communication aspects.

University of Washington 8

Infrastructure of Flight Tests

Autonomous Flight Systems Laboratory

In addition to simulation, direct access to actual hardware and systems.

Partnered with the Insitu Group for ScanEagle

UAVs, Northwind Marine for SeaFox Boats.

Extensive test infrastructure in place by working with these local companies

Includes sea launch & retrieval of UAVs

Aeronautics & Astronautics

University of Washington 9

Aspects of Autonomy

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Base

University of Washington 10

Aspects of Autonomy

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Base

STRATEGIC

Team Assembly

Task Assignment

TACTICAL

 Pattern Hold

DYNAMICS & CONTROL

 Auto Launch/Retrieval

University of Washington 11

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly

Aeronautics & Astronautics

Base

University of Washington 12

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

Base

University of Washington 13

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

STRATEGIC

 Path Planning

TACTICAL

 Obstacle/Threat Avoidance

 Path Following

DYNAMICS & CONTROL

 State Stabilization

University of Washington 14

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

Obstacle/Threat

Avoidance

Base

University of Washington 15

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

Base

University of Washington

Obstacle/Threat

Avoidance

STRATEGIC

Dynamic Task Allocation

Team-Based Cooperation

 Path Re- planning

TACTICAL

 Obstacle Avoidance

 Engagement Maneuvers

DYNAMICS & CONTROL

 State stabilization

16

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

Base

Coordination w/ surface vehicles

University of Washington

Obstacle/Threat

Avoidance

17

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

Base

Obstacle avoidance

STRATEGIC

 Provide improved target tasking and routing info to unmanned surface vehicles

TACTICAL

Orbit Coordination

Communication

Path Following

DYNAMICS & CONTROL

 Signal Tracking

Coordination w/ surface vehicles

University of Washington 18

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

Obstacle/Threat

Avoidance

Base

Searching/Target ID

Coordination w/ surface vehicles

University of Washington 19

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

Obstacle avoidance

 Map-Based and Probabilistic Searches

TACTICAL

 Path following

DYNAMICS & CONTROL

 State stabilization

Coordination w/ ground vehicles

Searching/Target ID

University of Washington 20

Aspects of Autonomy

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

Obstacle/Threat

Avoidance

Base

Searching/Target ID

Coordination w/ surface vehicles

University of Washington 21

Current Research Projects

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

 Real Time Strategic Mission Planning

 dynamic task and path planning for a team of autonomous vehicles to cooperatively execute a set of assigned tasks.

 Coordination of Heterogeneous Vehicles

 developing robust navigation and guidance algorithms to coordinate multiple vehicles to perform a cooperative task.

 Autonomous Search and Target Identification

 using total magnetic intensity measurements to search and identify magnetic anomalies in a predetermined area.

University of Washington 22

Real Time Strategic Mission Planning

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Pattern hold/Team assembly Transition

Obstacle/Threat

Avoidance

Base

Searching/Target ID

Coordination w/ surface vehicles

University of Washington 23

System Overview

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

University of Washington

Previously funded by DARPA & AFOSR

24

System Block Diagram

Autonomous Flight Systems Laboratory

Solving optimal control problems in real-time

Aeronautics & Astronautics

D

 task plans

Q

 paths

University of Washington 25

Stochastic Problem Formulation

Autonomous Flight Systems Laboratory

Aeronautics & Astronautics

V

Predicted probability of survival of each vehicle at time t q+1

 v

V

( q

1 )

  v

V

( q )

N

O 

1 j

1

B v j

( q

1 )

 j

O

( q )

 O j

 x i

F

Predicted probability that a task is not completed at time t q+1 x i

F ( q

1 )

 x i

F ( q )

N v

V 

1

 1

B v i ( q

1 )

 v

V ( q )

V v

N j

 T

1 d ij v



Team utility function

J

Mission Score

Cost

J

N

T

N

1   i

1

 p

 i

F

( ) i

F

( ) v

N

V 

1

1

 i

B q v

1)

V v

( )

V v

N

T j

1 d ij v

N

V v

1 v

V v

( s p

)

  v

V

N

 

Q

F Q s v

( ( v p

))

University of Washington 26

Distributed Architecture for

Coordination of Autonomous Vehicles

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Each vehicle plans its own path and makes task trading decisions to maximize the team utility function

There is one active coordinator agent at a time

 efficiency failure detection local/global information exchanges

Computational requirement for running coordinator agent is small compared to planning

Coordinator role can be transferred to another vehicle via a voting procedure

University of Washington 27

Evolution-based Cooperative Planning

System (ECoPS)

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Uses Evolutionary Computationbased techniques in the optimization of trading decision making and path planning

Task planner uses price and shared information in addition to predicted states of the world for making trading decisions

Task planner interacts with path planner and state predictor to simultaneously search feasible near-optimal task and path plans.

We call this system the “Evolution-

Based Collaborative Planning

System” – ECoPS, combining market based techniques with evolutionary computation (EC).

University of Washington 28

Evolutionary Computation (EC)

Aeronautics & Astronautics Autonomous Flight Systems Laboratory

 Motivated by evolution process found in nature

 Population-based stochastic optimization technique

Metaphor Mapping

University of Washington 29

Features of Evolution-Based

Computation

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

 Provides a feasible solution at any time

 Optimality is a bonus

 Dynamic replanning

 Non-linear performance function

 Collision avoidance

 Constraints on vehicle capabilities

 Handling loss of vehicles

 Operating in uncertain dynamic environments

 Timing constraints

University of Washington 30

Market-based Planning for

Coordinating Team Tasks

Autonomous Flight Systems Laboratory

Distributed Task Planning Algorithm

Aeronautics & Astronautics

Task allocation problem: max ( A

A

J )

At trading round n

A ( n )

T

1

( n ), T

2

( n ),  , T

N v ( n )

Each vehicle proposes ( ), i ( ) which are approved by the auctioneer based on bid price.

At the end of the trading round:

T i

( n

1 )

T i

( n )

B i

( n )

S i

( n )

The goal of task trading:

J ( A ( n

1 ))

J ( A ( n ))

University of Washington 31

Dynamic Path Planning

Autonomous Flight Systems Laboratory

 Generate feasible paths and planned actions within a specified time limit ( Δ Ts ) while the vehicles are in motion.

T s

 t s p

 t s p

1

Highly dynamic environment requires a high bandwidth planning system (i.e. small

Δ T s

).

Formulate the problem as a

Model-based Predictive

Control (MPC) problem

University of Washington

Aeronautics & Astronautics

32

EC-Based Path Planning

Autonomous Flight Systems Laboratory

Path Encoding

Aeronautics & Astronautics

Dynamic Planning

Mutation

University of Washington 33

Collision Avoidance

Autonomous Flight Systems Laboratory

Model each site in the environment as a

 i

Probability of intersection:

 use numerical approximation

 computationally easier than true solution

B i v   k

 i v

 z

V v

 

, C k i

 i

 v k

 t

Z i v

: possible intersection region

 i v

: probability density field function z v

V

: position on the path

C i

: expected site location v : velocity of the vehicle

University of Washington

Aeronautics & Astronautics

34

Collision Avoidance Example

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

University of Washington 35

Simulation Results

Autonomous Flight Systems Laboratory

Simulation on the Boeing Open Experimental Platform

Aeronautics & Astronautics

University of Washington 36

Some Aspects of ECoPS

Autonomous Flight Systems Laboratory

Aeronautics & Astronautics

Each vehicle computes its own trajectory and makes decision to trade its tasks with other vehicles.

Vehicles may sacrifice themselves if that benefits the team.

Each vehicle needs to have periodically updated locations of nearby vehicles only for collision avoidance.

Each vehicle needs to know the information about the environment. The accuracy of the information affects the quality of its decision making.

The rate of environment information updates should be selected based on how fast objects move in the environment.

Assuming vehicles are equipped with on-board sensors, sharing sensed data improves the performance of the team.

University of Washington 37

Coordination of Heterogeneous Vehicles

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Pattern hold/Team assembly Transition

Obstacle/Threat

Avoidance

Base

Searching/Target ID

Coordination w/ surface vehicles

University of Washington 38

Coordination and Communication with

Autonomous Surface Vehicles

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

At strategic level,

UAVs can provide improved target tasking and routing information to surface vehicles

Autonomous path planning for surface vehicles through nonstructured environments enhanced by UAV information

At tactical level, UAVs can track evasive targets and update world estimates

Currently funded under WTC Phase I Fall/Winter ‘05

University of Washington 39

Goals and Advantages

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Goals

 Use multiple low-cost UAVs to cooperatively track targets

 Ability to mark targets, report to central database, report to deployed surface vehicles

 Improve quality and quantity of

ISR data and battlefield awareness

Advantages

 Tracking targets with tactical UAVs can require high operator workload

 Evasive targets could fool a single

UAV

University of Washington 40

Simulation Visualization

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Autonomous Orbit Coordination for Multiple UAVs

University of Washington 41

Simulation Results

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Effects of Radius and Airspeed Manipulation

University of Washington 42

Simulation Results

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Effects of Radius and Airspeed Manipulation

University of Washington 43

Orbit Coordination

Autonomous Flight Systems Laboratory

 Maintains relative phase angle between two UAVs in presence of disturbance

 Nonlinear issues dealing with asymmetry of varying orbits

 Joint effort between

UW, Cornell, U of

Calgary, and The

Insitu Group

Aeronautics & Astronautics

Insitu SeaScan tracking moving target

University of Washington 44

Autonomous Search and Target

Identification

Autonomous Flight Systems Laboratory

Pattern hold/Team assembly Transition

Aeronautics & Astronautics

Obstacle/Threat

Avoidance

Base

Searching/Target ID

Coordination w/ surface vehicles

University of Washington 45

Probabilistic Searching

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Evaluation of Autonomous Airborne Geomagnetic Surveying

 Utilize magnetometer to measure local magnetic anomalies for known signature

 Identify and classify anomalies

 Search for and track anomalies cooperatively

Currently funded under WTC Phase II

University of Washington 46

General Architecture

Autonomous Flight Systems Laboratory

Obtaining local magnetic map

Aeronautics & Astronautics

Data from Fugro Airborne Surveys

University of Washington 47

General Architecture

Autonomous Flight Systems Laboratory

Agent 1

Agent 2

Aeronautics & Astronautics

Groundstation

Local Magnetic Map

University of Washington

Occupancy Map

48

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Agents

Target

False

Anomalies

University of Washington 49

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 50

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 51

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 52

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 53

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 54

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 55

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 56

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 57

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 58

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 59

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 60

Occupancy-Based Map Search

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Basic Algorithm

Score Cell

Evaluate possible control population

Execute control

University of Washington 61

Anomaly Encounter

Autonomous Flight Systems Laboratory

How to score each cell?

Aeronautics & Astronautics

Anomaly

Aeromagnetic Data from Fugro Airborne

Corresponding Line Data

Goal: Classify anomaly as target or false signature

University of Washington 62

Particle Filter

Autonomous Flight Systems Laboratory

How consistent is trace with trajectory over desired target?

Aeronautics & Astronautics

Which trajectory

(if any) would produce trace?

Classify using Particle Filter

Nonparametric Bayes filter. Similar to Unscented Kalman or discrete Bayes filter.

University of Washington 63

Particle Filter

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Fox, D., Thrun, S., Burgard, W. 2005, “Probabilistic Robotics”

Klein, D.J., Klink, J.O., 2005, “Mobile Robot Localization” x

2 t x

1 t

 t x m t function

 t

 particle_f ilter (

 t

1

, u t

, z t

) for m=1:M

Sample x t from f motion

 x t w t

 f sensor

 z t

| x t

 t

 

 x t

| u t

, x t

1

 end

 t sampled from

 t w/probability α w t

University of Washington 64

True Anomaly Encounter

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

University of Washington 65

Different Magnetic Signatures

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

What about for false anomalies?

University of Washington 66

Confidence Comparison

Autonomous Flight Systems Laboratory

Actual Target Encounter

Aeronautics & Astronautics

False Encounter

Features

Use combination of particle filter and neural net to identify target and quantify confidence.

University of Washington 67

Contact Us

Autonomous Flight Systems Laboratory Aeronautics & Astronautics

Investigators

Dr. Rolf Rysdyk

Dr. Uy-Loi Ly

Dr. Juris Vagners

Dr. Kristi Morgansen rysdyk@aa.washington.edu

ly@aa.washington.edu

vagners@aa.washington.edu

morgansen@aa.washington.edu

Dr. Anawat Pongpunwattana anawatp@u.washington.edu

Autonomous Flight Systems Laboratory

Guggenheim 109

(206) 543-7748 http://www.aa.washington.edu/research/afsl

Nonlinear Dynamics and Control Laboratory

AERB 120

(206) 685-1530 http://vger.aa.washington.edu

University of Washington 68

Download