23
Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair Brian Mendel Andrew Moran Linda Wills

Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

  • View
    215

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Hybrid WorkgroupPam Binns

Magnus Carlsson

David Corman

Bonnie Heck

Tom Henzinger

Gokhan Inalhan

Gabor Karsai, co-chair

Wallace Kelly

Edward A. Lee, chair

Brian Mendel

Andrew Moran

Linda Wills

Page 2: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Modeling Questions

Language (syntax, semantics, composition) Modeling a distributed physical plant Modeling the controller Modeling faults and their effects How are models used? Semantics of anytime computation

Page 3: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Uses of Hybrid Systems

Analysis– Analysis of discrete and continuous controllers

interacting with a continuous plant Simulation

– Evaluation of continuous models with discrete mode transitions

Embedded modeling– Embedding of continuous models in deployed

systems, integrated into the control algorithms (e.g. for fault identification)

Page 4: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Questions

What's in the OCP, and what's in the application? Run-time support for hybrid models? at what cost? Design-time support? (e.g. generation of software) Anytime computation? (resource management) QoS guarantees? Is real-time O/S the only option? Admission control? Support for domain-specific languages?

Page 5: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Examples Considered

UAV helicopter control (GATech) Free flight conflict avoidance (Rockwell) Platooning vehicles (Berkeley) Dragonfly – Multi-vehicle coord (Stanford)

In all cases, we talked about fault detection and adaptation using modal control.

Page 6: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Mode Transition Control (GATech)Overall Architecture

MissionPlanning

SituationAwareness

Mode TransitionControllers

ModeTransitionManager

MTC #1

MTC #2

MTC #N

FlightController

ModeSelection

Goal Mode

TransitionStart Flag

TransitionComplete Flag

Helicopter States

Mode Transition Control Manager

Page 7: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Mode Transition Specification

Mode1 Mode2

ModeN

TransitionStart Flag

TransitionComplete Flag

DesiredGoal Mode

11

22

N 2

2N

21

12

1N

N1

NN

ij

ii

---- denotes a transition from Modei to Modej

---- denotes a regulation about Modei

Page 8: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

FDI/FTC Demo in OCP (GATech)

Subsystem

Local Controller

Subsystem

Local Controller

InterconnectionStructure

ReconfigurationController

RedistributionController

Set-PointController

Fault TolerantControl Manager

Fault Detection& Identification

HighLevel

HighLevel

MidLevel

MidLevel

LowLevel

LowLevel

e1

v1

r1y1

Indicates separate OCP components

Page 9: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Free Flight (Rockwell)

Page 10: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Extensions

UCAVs with accurate models of one another – coordinated maneuvers rather than assuming the other aircraft will not react.

Errors in transmission, faulty/damaged UCAVs – fault identification and adaptation.

Create hybrid automaton model and use hybrid Mocha to verify that conflicts do not occur.

Page 11: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Scenario – Vehicle Tracking

Publishes:• time • driving force• velocity• position

Event Channel

Implements:• fault detection • modal control

Subscribes

Page 12: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Modeling Car Tracking

Force car m odel pub lishe r

lead ing car

pub lishe r

force

speed

position

m onitor

subsc

subsc

Alarm

pub l

pub l

too close

too far

fau ltde tection

car m odelsubsc

fo llow ing car

pub lishe r

TrackingController

(P ID)

Bang-BangController

m odalcontro lle r

Java Space (today)OCP (tomorrow)

thanks to Jie Liu and Xiaojun Liu

Page 13: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Execution

Page 14: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

sensors

leader

Br Acc

Ba

bang-bang PID

follower

controller actuators

S

Hierarchical ViewOCP here only?

Page 15: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Styles of Publish and Subscribe Interactions

time stamped events? globally time stamped? reliable delivery? ordered delivery? signal coordination? synchronous delivery? blending of multiple publishers? dynamic redirection/resourcing? persistence? history?

Page 16: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

OCP Domain #1

time stamped events? yes globally time stamped? no reliable delivery? yes ordered delivery? no signal coordination? yes globally synchronous delivery? no blending of multiple publishers? no dynamic redirection/resourcing? yes persistence? no history? no

Page 17: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

What is a Domain

The definition of the interaction of components, and the software that supports this interaction.

Multi-domain modeling means: Hierarchical composition

– heterogeneous models allowed

Domains can be specialized– avoid creeping featurism– enable verification

Data replication in OCP/Boldstroke is another domain– separation of communication mechanisms.

Page 18: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

What technology can be shared when building domains?

Abstract syntax Type systems Components Interfaces …

sensors

leader

Br Acc

Ba

bang-bang PID

follower

controller actuators

S

Page 19: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Domains After Domain 1

High-sample-rate, periodic event handling Stream-based component interaction Control reconfiguration management Time-triggered, synchronous modeling Continuous-time domain

Domains need to be able to: Share a consistent notion of time Share signals/params across levels of the hierarchy. Export interfaces to other domains Import components designed in another domain

Page 20: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Milestones

Use OCP in demonstrations in each of– Active state models– On-line control customization– Coordinated multi-modal control

Steps– Define the challenge problems that justify domains.– Work out how mode transitions and other control

reconfiguration are handled in the OCP.– Work out the interaction semantics (domains)

supported by the OCP (mode transition control? continuous modeling? synchronous interactions?)

Page 21: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Architecture to Avoid

Poor common infrastructure. Weak specialization.Poor resource management and sharing. Poor planning.

Page 22: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Also to Avoid

Elegant, unified, and beautiful, but rigid, inflexible, and difficult to adapt. Plus, it takes 100 years to build.

Page 23: Hybrid Workgroup Pam Binns Magnus Carlsson David Corman Bonnie Heck Tom Henzinger Gokhan Inalhan Gabor Karsai, co-chair Wallace Kelly Edward A. Lee, chair

Elegant Federation

Elegant federation of heterogeneous models.

Two Rodeo Drive, Kaplan, McLaughlin, Diaz

Sourc

e:

Kapla

n M

cLaugh

lin D

iaz,

R. R

appaport

, R

ock

port

, 1

998