[PPT] - Towards Integrated Multi-Formalism Tool Support for the Design of PowerPoint Presentation

SLIDE 1

Software Technologies Concertation on Formal Methods for Components and Objects (FMCO’10)

Towards Integrated Multi-Formalism Tool Support for the Design of Embedded Control Systems

Martin Hüfner, Christian Sonntag, Sebastian Engell Process Dynamics and Operations Group (DYN)

Dept. of Biochemical and Chemical Engineering (BCI)

Technische Universität Dortmund, Germany

SLIDE 2

2

Outline

The MULTIFORM project
Illustrative example

– Integrated design of a pipeless plant

Goals of the MULTIFORM project
Design flow example
Research areas

– Algorithmic model exchange – Controller specification and synthesis – Trans-level tool integration and verification – Code verification – The design framework

Summary & state of the project

SLIDE 3

3

History

– Initiated by TU Dortmund as a successor of HYCON WP3 (Tool Integration) – Application October 2007 – Start September 2008

Funding

– Financed by the EU within the 7th Framework Programme in the ICT domain – Funding: 2.800.000 € – 8 Partners, thereof two industrial

Duration: 42 months

– September 2008 – February 2012

More information

– http://www.ict-multiform.eu

The European Project MULTIFORM

SLIDE 4

4

TUDO (Coordinator)

– TU Dortmund, Germany Sebastian Engell

TUE

– TU Eindhoven, Netherlands Koos Rooda, Bert van Beek, Jos Baeten

Verimag/ UJF

– Universite Joseph Fourier, Grenoble, France Goran Frehse, Oded Maler

RWTH

– RWTH Aachen, Germany Stefan Kowalewski

AAU

– Aalborg Universitet, Denmark Kim Larsen, Brian Nielsen

ESI

– Stichting Embedded Systems Institute Ed Brinksma, Boudewijn Haverkort

UJF

The MULTIFORM Consortium

VEMAC

– Aachen, Germany Michael Reke

KVCA

– “Danish Cooling Cluster” Jens Andersen – Closely working with DANFOSS

TUDO RWTH/VEMAC Verimag/UJF TUE/ESI AAU/KVCA

SLIDE 5

5

Control PC Camera AGVs Storage station Charging stations Product Color stations Mixing station

Example: Design of a Pipeless Plant

SLIDE 6

6

Challenges for Model-based Design (1)

Design and validation on different

levels of abstraction

– Specification

Specification of the tasks and of

the performance of the system

– High-level design

Choice of the equipment,

feasibility and bottleneck analysis, throughput maximization, plant layout

ptimization

– Low-level design

Optimization and control of

processing steps and motion dynamics, logic control

Choice of sensors and actuators,

communication system

– Implementation

PLCs, embedded controllers,

communication system

Control PC Camera AGVs Storage station Charging stations Product Color stations Mixing station

System specification High-level design Low-level design Implementation Implementation tests Low-level tests High-level tests Performance analysis

Design Validation

SLIDE 7

7

Challenges for Model-based Design (2)

The control system spans the

complete control hierarchy

– Coordination control

Scheduling and performance
ptimization

– Advanced control

Control of batch processes
AGV path planning

– Regulatory control

AGV motion control
Docking control
Sequence control in the

processing stations

Low-level continuous control

– Low-level safety-related control

Control PC Camera AGVs Storage station Charging stations Product Color stations Mixing station

Discrete-event, timed, and hybrid models Discrete-event, hybrid, and continuous models Timed or hybrid models Continuous models

SLIDE 8

8

Extension of the model-based approach beyond the scope of

classical feedback controller design to cover the complete control hierarchy.

The long-term goal is to support a fully model-driven design

process of a controlled system over its full life cycle

– Build systems that are correct by design and where the interaction of the components is fully transparent

Goal of the MULTIFORM Project

SLIDE 9

9

Integrated modeling and design of the system itself and of the multi-

layered and networked control system

– Including a structured approach to the management of specifications, design decisions, models, and results

Coverage of all layers of the automation and design hierarchy

– Integrated tool support on all layers of the automation and design hierarchies

Current state: Islands of support for specific design and analysis tasks
Trans-level integration of model-based design approaches
Support of iterations in the design process
Propagation of faults and unexpected behaviors
Modifications over the life cycle without top-down redesign

 Tool integration and Design Framework  Exchange of models between tools via the CIF (Compositional Interchange Format)  Improvement of the tool support for the design steps

Vision for Integrated Model-based Design

SLIDE 10

10

Design tasks Requirement specification Feasibility analysis Plant layout design AGV speed analysis Controller design Controller code generation

Design Flow: Design Tasks for the Pipeless Plant

SLIDE 11

11

Boderc key drivers Feasibility analysis Plant layout design AGV speed analysis Controller design Controller code generation

Design Flow: Refinement of Requirements

Design tasks Models Requirement specification

SLIDE 12

12

Storage station Complex & interesting Plant dosing station mixing station cleaning station AGV broad AGV even AGV round AGV with wheels Standard AGV AGV with Microcontroller AGV with WLAN Vessel separable from AGV Transparent vessel Open space/no rails for AGV Process times/AGV speed > 2? AGV operating time/charging time > 3? Colored product Complex recipes Different recipes with different steps Small production time Non-toxic materials Recipes can be modified Quality control for product Avoid cross contamination Parameterization of plant and recipes Parallel production Path planning & routing Vessel tracking Camera localization system Simple & powerful GUI Control PC Lab size plant boundaries Emergency stop Sensors for safety Sensors for control Same power supply Electrical power No hydraulics Avoid pneumatics Support documentation of all steps Appropriate design approach Integration to Multiform safety Standard parts Interesting & Impressive product Easy handling & robustness Multiform must-haves

Price / budget hast to be kept Demonstrator plant For Multiform Design flow and tool chain Reuse as a lab experiment / fair exhibit

Key Drivers for the Pipeless Plant

SLIDE 13

13

Boderc key drivers

Requirements

System configuration:

Production Tasks
Functionality of the Stations
Type of AGVs

+ Design choices

Number of stations
Station types (Color, Mixing, Storage,…)
Number of AGVs

Design Flow: Initial Design

Design step

SLIDE 14

14

Boderc key drivers Timed Chi Requirement specification Plant layout design AGV speed analysis Controller design Controller code generation

– Complete coarse plant model – Here: Timed model – Purpose: Simulation

Information propagation via Design Framework

Design tasks Models

Design Flow: Feasibility Analysis

Feasibility analysis

SLIDE 15

15

Feasibility Analysis Design step Requirements Design flow Design step

Boderc key drivers & Design Decisions Timed Chi model – Complete coarse plant model

– Timed model – Simulation

Model of the complete plant using approximation System configuration:

# and type of stations
Recipes
# and type of AGVs

Design Flow: Feasibility Analysis

Another View Control View

Plant Model

Chi model

Structure View

System block

SLIDE 16

16

Chi model

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
StationTypes=[5x1]
StationPositions=[5x2]
# AGVs=3
RecipeList=[6x1]
MovementTimes=[5x5]

Design Flow: Feasibility Analysis

SLIDE 17

17

Experiment: Set input: Model parameters #Stations:=5 StationTypes #AGVs:=3 MovementTimes:=[5x5]

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]

Set input:

Chi model

Design Flow: Feasibility Analysis Experiment

SLIDE 18

18

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]

Experiment: Set input: Model parameters #Stations:=5 StationTypes #AGVs:=3 MovementTimes:=[5x5] Tool parameters Chi model

Design Flow: Feasibility Analysis Experiment

SLIDE 19

19

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]

Chi Simulator Select tool: Experiment: Set input: Model parameters #Stations:=5 StationTypes #AGVs:=3 MovementTimes:=[5x5] Tool parameters Chi model

Design Flow: Feasibility Analysis Experiment

SLIDE 20

20

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]

Chi Simulator Select tool: Chi Simulation results Save results: Experiment: Set input: Model parameters #Stations:=5 StationTypes #AGVs:=3 MovementTime:=[5x5] Tool parameters Chi model

Design Flow: Feasibility Analysis Experiment

SLIDE 21

21

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=12

Chi lot-time diagram

Number in block represents vessel

Design Flow: Feasibility Analysis Experiment

SLIDE 22

22

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=12

Experiment: Chi Simulator Select tool: Save results: Set input: Model parameters #Stations:=5 StationTypes #AGVs:=3 MovementTime:=[5x5] Tool parameters Chi model Chi Simulation results Add new parameter and value from results

Design Flow: Feasibility Analysis Experiment

Iteration until result satisfactory

SLIDE 23

23

Requirements Feasibilty Analysis Design flow Design step

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=12

Design Flow: Feasibility Analysis Experiment

Another View Control View Structure View

Plant

Results

Experiment Configuration

Chi model

SLIDE 24

24

Boderc key drivers Timed-Chi Uppaal Requirement specification Feasibility analysis AGV speed analysis Controller design Controller code generation

– Complete plant model – Timed Automata model – Scheduling – Complete plant model – Timed model – Simulation

Information propagation via Design Framework

Design Flow: Plant Layout Design/ Scheduling

Design tasks Models

Algorithmic model transformation

Plant layout design

SLIDE 25

25

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=12

Algorithmic model transformation

Design Flow: Plant Layout Design / Scheduling

Requirements Feasibilty Analysis Design step Scheduling Design flow

Another View Control View Structure View

Plant Chi

SLIDE 26

26

Experiment: Set input: Model parameters Tool parameters Chi model Uppaal model file Tool chain: Chi- to-Uppaal transformation Select tool: Save results:

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=12

Design Flow: Transformation Experiment

SLIDE 27

27

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=12

Design Flow: Plant Layout Design / Scheduling

Requirements Feasibilty Analysis Design step Scheduling Design flow

Another View Control View Structure View

Plant

Uppaal

Chi

SLIDE 28

28

Experiment: Set input: Model parameters Tool parameters Uppaal model Uppaal scheduling Uppaal Select tool: Save results:

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=12

Uppaal trace file

Design Flow: Plant Layout Design / Scheduling

SLIDE 29

29

Experiment: Set input: Model parameters Tool parameters Uppaal model Uppaal scheduling Uppaal Select tool: Save results:

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=14

Uppaal trace file

Design Flow: Plant Layout Design / Scheduling

SLIDE 30

30

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=14

Design Flow: Plant Layout Design / Scheduling

Requirements Feasibilty Analysis Design step Scheduling Design flow

Another View Control View Structure View

Plant

Schedule Trace

Uppaal

Chi

SLIDE 31

31

Development tool chain

Boderc key drivers Timed Chi Uppaal PPSim Requirement specification Feasibility analysis AGV speed analysis Controller design Controller code generation

– Complete plant – Timed Automata – Scheduling – Complete plant model – Geometrical plant layout – Routing heuristics – Simulation – Complete plant – Timed model – Simulation

Design tasks Models

Information propagation via Design Framework

Plant layout design

SLIDE 32

32

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=14

PPSim – Complete plant

– Geometrical plant layout – Basic routing – Simulation

Complete plant model based on the scheduling trace of Uppaal

Design Flow: Detailed Plant Layout Design

Requirements Feasibilty Analysis Design step Scheduling Plant layout Design flow

Another View Control View Structure View

Plant

Uppaal

PPSim Chi

SLIDE 33

33

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=14

PPSim – Complete plant

– Geometrical plant layout – Basic routing – Simulation

Plant editor Recipe editor

Design Flow: The PPSim Tool

SLIDE 34

34

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=14
avgAGVSpeed=500

PPSim – Complete plant

– Geometrical plant layout – Basic routing – Simulation

PPSim model

Extract parameters:

Average AGV speed

Design Flow: Detailed Plant Layout Design

SLIDE 35

35

Experiment: Set input: Model parameters Tool parameters PPSim model PPSim Select tool: Save results: System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=14
avgAGVSpeed=500

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim – Complete plant

– Geometrical plant layout – Basic routing – Simulation

Simulation results

Design Flow: PPSim Experiment

SLIDE 36

36

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim – Complete plant

– Geometrical plant layout – Basic routing – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=14
avgAGVSpeed=500

Design Flow: Detailed Plant Layout Design

Requirements Feasibilty Analysis Design step Scheduling Plant layout Design flow

Another View Control View Structure View

Plant

Uppaal

PPSim Chi

SLIDE 37

37

1. Rerun experiments

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim – Complete plant

– Geometrical plant layout – Basic routing – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=14
avgAGVSpeed=500

Design Flow: Parameter Propagation

Plant

Uppaal

PPSim Chi

SLIDE 38

38

1. Rerun experiments
2. Update results

Uppaal – Complete plant

– Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim – Complete plant

– Geometrical plant layout – Basic routing – Simulation

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=13
avgAGVSpeed=500

Design Flow: Parameter Propagation

Plant

Uppaal

PPSim Chi

SLIDE 39

39

Boderc key drivers Timed Chi Uppaal PPSim Requirement specification Feasibility analysis AGV speed analysis Controller design Controller code generation

– Complete plant – Timed Automata – Scheduling – Complete plant – Geometrical plant layout – Basic routing – Simulation – Complete plant – Timed model – Simulation

Information propagation via Design Framework Model transformation via the CIF

Design Flow: Parameter Propagation

Design tasks Models Plant layout design

SLIDE 40

40

Boderc key drivers Timed Chi Uppaal PPSim gPROMS Requirement specification Feasibility analysis Plant layout design Controller design Controller code generation

– Complete plant – Timed Automata – Scheduling – Complete plant – Geometrical plant layout – Basic routing – Simulation – Hybrid dynamics, PDE models – Optimization – Complete plant – Timed model – Simulation

Information propagation via Design Framework Model transformation via the CIF

Design Flow: AGV Speed Analysis

Design tasks Models AGV speed analysis

SLIDE 41

41

Uppaal

– Complete plant – Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim

– Complete plant – Geometrical plant layout – Basic routing – Simulation

gPROMS – Dynamics

– Optimization

Model of AGV movement and fluid dynamics

Design Flow: AGV Speed Analysis

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=13
avgAGVSpeed=500

Requirements Feasibilty Analysis Design step Scheduling Plant layout Speed analysis Design flow

Another View Control View Structure View

Plant

Uppaal

PPSim

Stations +AGVs

AGVs

gPROMS

Chi

SLIDE 42

42

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=13
avgAGVSpeed=500

Uppaal

– Complete plant – Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim

– Complete plant – Geometrical plant layout – Basic routing – Simulation

gPROMS – Dynamics

– Optimization

Time Speed

500

Acceleration

max.

max.

Optimization of acceleration profile with constraints:

Avoidance of spilling during transport

Design Flow: Dynamic Optimization

SLIDE 43

43

Experiment: Set input: Model parameters Tool parameters: Maximize avgAGVSpeed Without spilling liquid gPROMS model

Uppaal

– Complete plant – Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim

– Complete plant – Geometrical plant layout – Basic routing – Simulation

gPROMS – Dynamics

– Optimization

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=13
avgAGVSpeed=432

gPROMS Select tool: Save results: gPROMS

ptimization

results

Design Flow: gPROMS Experiment

SLIDE 44

44

Uppaal

– Complete plant – Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim

– Complete plant – Geometrical plant layout – Basic routing – Simulation

gPROMS

1. Rerun experiments

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=13
avgAGVSpeed=432

Design Flow: Parameter Propagation

Plant

Stations +AGVs

AGVs

gPROMS

PPSim

Uppaal

Chi

– Dynamics – Optimization

SLIDE 45

45

Uppaal

– Complete plant – Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim

– Complete plant – Geometrical plant layout – Basic routing – Simulation

gPROMS

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=13
avgAGVSpeed=432
1. Rerun experiments
2. Update results

Design Flow: Parameter Propagation

Plant

Stations +AGVs

AGVs

gPROMS

PPSim

Uppaal

Chi

– Dynamics – Optimization

SLIDE 46

46

Uppaal

– Complete plant – Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim

– Complete plant – Geometrical plant layout – Basic routing – Simulation

gPROMS

1. Rerun experiments

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=13
avgAGVSpeed=432

Design Flow: Parameter Propagation

Plant

Stations +AGVs

AGVs

gPROMS

PPSim

Uppaal

Chi Uppaal

– Dynamics – Optimization

SLIDE 47

47

Uppaal

– Complete plant – Timed Automata – Scheduling

Boderc key drivers Timed-Chi

– Complete plant – Timed model – Simulation

PPSim

– Complete plant – Geometrical plant layout – Basic routing – Simulation

gPROMS

System configuration:

# Stations=5
Station Types=[5x1]
Station Positions=[5x2]
# AGVs=3
Recipe List=[6x1]
Movement Times=[5x5]
Recipes/Day=11
avgAGVSpeed=432
1. Rerun experiments
2. Update results

Design Flow: Parameter Propagation

Plant

Stations +AGVs

AGVs

gPROMS

PPSim

Uppaal

Chi Uppaal

– Dynamics – Optimization

SLIDE 48

48

Boderc key drivers Timed Chi Uppaal PPSim gPROMS Requirement specification Feasibility analysis Plant layout design Controller design Controller code generation

– Complete plant – Timed Automata – Scheduling – Complete plant – Geometrical plant layout – Basic routing – Simulation – Dynamics – Optimization – Complete plant – Timed model – Simulation

Information propagation via Integration Framework Model transformation via the CIF

Design Flow: Parameter Propagation

Design tasks Models AGV speed analysis

SLIDE 49

49

Boderc key drivers Timed Chi Uppaal PPSim DC/FT gPROMS Requirement specification Feasibility analysis Plant layout design AGV speed analysis Controller code generation Supervisory Control

– Complete plant – Timed Automata – Scheduling – Complete plant – Geometrical plant layout – Basic routing – Simulation – Dynamics – Optimization – Complete plant – Timed model – Simulation

DES – Docking process

– Timed model

Design Flow: Logic Controller Design and Synthesis

Information propagation via Integration Framework Model transformation via the CIF

Design tasks Models Controller design

SLIDE 50

50

Boderc key drivers Timed Chi Uppaal PPSim DC/FT PHAVer / SpaceEx gPROMS Requirement specification Feasibility analysis Plant layout design AGV speed analysis Controller code generation Supervisory Control

– Complete plant – Timed Automata – Scheduling – Docking process – Hybrid dynamics – Complete plant – Geometrical plant layout – Basic routing – Simulation – Dynamics – Optimization – Complete plant – Timed model – Simulation

DES – Docking process

– Timed model

Design Flow: Verification

Design tasks Models

Composition with HA docking model

Controller design

SLIDE 51

51

Boderc key drivers Timed Chi Uppaal PPSim DC/FT Code gPROMS Requirement specification Feasibility analysis Plant layout design AGV speed analysis Controller design Supervisory Control

– Complete plant – Timed Automata – Scheduling – PLC/Ethercat/AGV code – Verification with [mc]square – Complete plant – Geometrical plant layout – Basic routing – Simulation – Dynamics – Optimization – Complete plant – Timed model – Simulation

DES– Docking process

– Timed model

Design Flow: Code Generation

Design tasks Models PHAVer / SpaceEx

Composition with hybrid docking model

– Docking process – Hybrid dynamics

Controller code generation

SLIDE 52

52

Different views
n same

system and models

Not all models

present in every view

Require- ments Feasibilty Analysis Design step Scheduling Plant layout Speed analysis Design flow Docking Control Dead ends Code generation

Structure View

Plant

PPSim

Stations +AGVs

AGVs

gPROMS DC/FT PHAVer SC Uppaal

Chi

Control View

High level

Uppaal

PPSim

Low level

Imple- men- tation

gPROMS mc² DC/FT SC

Final Design Flow

SLIDE 53

53

Boderc key drivers Timed Chi Uppaal PPSim DC/FT Code gPROMS Requirement specification Feasibility analysis Plant layout design AGV speed analysis Controller design Supervisory Control

– Complete plant – Timed Automata – Scheduling – PLC/Ethercat/AGV code – Verification with [mc]square – Complete plant – Geometrical plant layout – Basic routing – Simulation – Dynamics – Optimization – Complete plant – Timed model – Simulation

DES – Docking process

– Timed model

Design Flow: Algorithmic Model Exchange

Design tasks Models PHAVer / SpaceEx

Composition with hybrid docking model

– Docking process – Hybrid dynamics

Controller code generation See second MULTIFORM presentation by

B. van Beek et al.

SLIDE 54

54

Boderc key drivers Timed Chi Uppaal PPSim DC/FT Code gPROMS Requirement specification Feasibility analysis Plant layout design AGV speed analysis Controller design Supervisory Control

– Complete plant – Timed Automata – Scheduling – PLC/Ethercat/AGV code – Verification with [mc]square – Complete plant – Geometrical plant layout – Basic routing – Simulation – Dynamics – Optimization – Complete plant – Timed model – Simulation

TA – Docking process

– Timed model

Design Flow: Logic Controller Design

Design tasks Models PHAVer / SpaceEx

Composition with hybrid docking model

– Docking process – Hybrid dynamics

Controller code generation

SLIDE 55

55

Logic Controller Design and Synthesis

Industrial practice: Logic controller design is a largely manual,

unsystematic, and error-prone procedure

Benefits of systematic specification refinement and analysis

– Improved and systematic communication between the plant designer and the control engineer → less “misunderstandings” and errors – Algorithmic requirements analysis → error avoidance early in the design phase

Benefits of the synthesis tool chain

– Largely automated design procedure – Comfortable choice between different synthesis approaches

Informal and vague specifications Systematic analysis Refinement Formal and precise specifications Control system Algorithmic synthesis Plant model

See third MULTIFORM presentation by

A. David et al.

SLIDE 56

56

Boderc key drivers Timed Chi Uppaal PPSim gPROMS Requirement specification Feasibility analysis Plant layout design AGV speed analysis Controller design Controller code generation

– Complete plant – Timed Automata – Scheduling – Complete plant – Geometrical plant layout – Basic routing – Simulation – Dynamics – Optimization – Complete plant – Timed model – Simulation

TA – Docking process

– Timed model

Verification, Trans-Layer Integration, Abstraction & Refinement

Design tasks Models DC/FT Code Supervisory Control

– PLC/Ethercat/AGV code – Verification with [mc]square

PHAVer / SpaceEx – Docking process

– Hybrid dynamics

Composition with hybrid docking model

SLIDE 57

57 57

WP3 Objectives

informal specifications

abstract models (discrete / timed) formal specifications linear hybrid automata models detailed models (nonlinear)

Implementation

control system

MC2 StateEx/ Phaver Uppaal e.g. Simulink

Develop techniques for checking refinement relations between models and artifacts at different design levels.

SLIDE 58

58

The MULTIFORM Design Framework Putting Chaos under Control [ESI]

Objective

– To enable more effective model-based design of complex systems – By means of making the decisions and dependencies in models and other design artifacts explicit

Approach

– Self-triggered error-detection that enables the connection of not directly related models – Not aiming at one big (system) model, but cherish multi-disciplinarity

software, hardware, mechanics, etc.
Application areas

– High-tech (cyber-physical) systems

58

SLIDE 59

59

The MULTIFORM Design Framework [ESI]

Consistent integration of design

models into a common software framework

Support of a generic design flow

model

– Design decisions – System design

Consistency management

– Communication of design parameters – Conflict detection – Models and results management

INPUT OUTPUT FORMAL INFORMAL DESIGN FLOW DESIGN STEP CONCRETE MODEL a n a l y s i s r e s u l t s analysis results

SLIDE 60

60

Case Study: Analysis of the Sensor and Actuator Wiring in Supermarket Refrigeration Systems by Online Testing

Large supermarket refrigeration systems: many sensors and

actuators, complex wiring

Faulty wiring is a

significant problem

Goal: Automated,

robust methods that reliably identify faulty wiring by tests that do not damage the equipment

SLIDE 61

61

Motivation

– Reduction of CO2 emissions and of the fuel consumption using PCC (piezo-controlled carburetor) in small combustion engines

Present development process (ineffective, expensive)

Case Study: Engine Control for Small-power Combustion Engines

Scooters Motor Bikes Industry Engines Go Karts Chain Saws

error requirements specification

mech. design

electronic design software design

mech. prototype

elect . prototype software prototype test bench test under real condition s mass production software testing error

time

error requirements specification

mech. design

electronic design software design

mech. prototype

elect . prototype software prototype test bench test under real condition s mass production software testing error

time

SLIDE 62

62

Case Study: Supporting the System Design

f Automatic Case Picking for Warehousing

Goals:

Design traceability, less problems (explicit assumptions,

explicit parameter dependencies), less design time

Link models of subsystems (Excel, POOSL, UPPAAL, …) at

the outside

Main system aspects:

– System performance – Controls architecture

SLIDE 63

63

Summary

There is a need for efficient model-based support of the

design of complex automated systems with trans-level propagation and iteration and re-use of models

An all-encompassing mega-tool for the design of complex

automated systems is not realistic!

Three different routes to tool and model integration and

design support are pursued in MULTIFORM: – Model exchange and tool chains via the CIF – Direct coupling of tools for testing – Propagation of parameters via the integration framework

SLIDE 64

64

State of the Project

CIF Vers. 2 is stable  significant simplification of the

semantics

– CIF2 models can be simulated via the CIF1 simulator

Transformations to and from gPROMS, Modelica, and from

SFC available

Connections to and from PHAVER (SpaceEx), UPPAAL,

Matlab/Simulink, and MUSCOD II are in progress

Ongoing developments in testing
Ongoing development of verification tools (SpaceEx,

[mc]square)

Definition of a generic design process and of the integration of

tools in the design process

Design framework layout finished, development of a prototype

is under way

SLIDE 65

65

TUDO (coordinator)

– TU Dortmund, Germany – Sebastian Engell

TUE

– TU Eindhoven, Netherlands Koos Rooda, Bert van Beek, Jos Baeten

Verimag/ UJF

– Universite Joseph Fourier, Grenoble, France Goran Frehse, Oded Maler

RWTH

– RWTH Aachen, Germany Stefan Kowalewski

AAU

– Aalborg Universitet, Denmark Kim Larsen, Brian Nielsen

ESI

– Stichting Embedded Systems Institute Ed Brinksma, Boudewijn Haverkort

The ideas presented here are the outcome of the joint effort

f all partners of the MULTIFORM Consortium!
VEMAC

– Aachen, Germany Michael Reke

KVCA