A common meta-model for the interoperation of tools with heterogeneous data models ECMFA 2010 Third...
-
Upload
naomi-ball -
Category
Documents
-
view
221 -
download
0
Transcript of A common meta-model for the interoperation of tools with heterogeneous data models ECMFA 2010 Third...
A common meta-model for the interoperation oftools with heterogeneous data models
ECMFA 2010
Third Workshop on Model-Driven Tool & Process Integration
2010-06-16
Andreas Baumgart
2
11.06.10
Motivation
• Embedded system development processes with various tools and different meta-models
• Cover requirements and models at different abstraction levels and with different viewpoints during process
• Meta-model transformations expensive when required between various meta-models
• Giant meta-model covering all specific artefacts difficult to handle
• Common meta-model provides common understanding of model artefacts and traceability
• Currently developped in european research project CESAR
3
11.06.10
Related Work
• SysML (OMG, UML based unification of standards for systems engineers)
• Structural modeling with blocks, properties, connectors, data types, physical units, dimensions
• Behavioral descriptions in terms of activity diagrams and state machines
• Requirements descriptions, use cases, trace links and test cases
• HRC (SPEEDS)
• Generic and formal description of hierarchical components with connectors, ports, port specifications and data types
• State Machines, Expression Language, Action Language (C like)
• Contracts with assumption / promise pairs on components and viewpoints
• EAST-ADL (ATESST, Automotive Domain Specific Language)
• Hierarchical modelling at different abstraction levels, features, functional and hardware models, connectors, flow / client-server / hardware ports, variability
• External nominal behavior, error behavior, timing descriptions
• Textual requirements, trace links, verification and validation
4
11.06.10
Concepts and artefacts of a common meta-model
• Heterogeneous Components• hierarchical, reuse, connectors, ports, port specifications, data flows, services,
data types, behavior
• Requirements, models and verification / validation along development process• requirements, levels of abstraction / formalization, traceability,
viewpoints (safety, realtime, …), fault models, product lines, test cases
5
11.06.10
Identication of concepts in different meta-models
Artifact Reference EAST-ADL 2 Reference
Component Part typed by Component
ADLFunctionType ADLFunctionPrototype.type [1]
Part Component has Parts
ADLFunctionPrototype ADLFunctionType.part [*]
Ports Component has Data Ports
ADL[In|Out]FlowPort ADLFunctionType.flowPort [*]
Component has Service Interface
ADL[Client|Server]Port ADLFunctionType.clientServerPort [*]Service
Connector Interfaces connected
ADLConnectorPrototype ADLConnectorPrototype.[flow, clientServer]Port [*]
Behaviour Component has Behaviour
ADL[Native, External]Behavior
ADLFunctionType.behavior [0..1]
Data Type Port is typed ADLDesignDataType ADLFlowPort.type [1]
6
11.06.10
Tool interoperation using the common meta-model
• Common understanding of model artefacts
• Mapping needed
• Exchange format
• Native format (i.e. for analysis)
• Common interface
• Identify foreign model artefacts
• Provide traceability between heterogeneous model elements
• Provide model with elements coming from different models
• Further trace model can provide linkage to original model elements
• Get language specific information from original model
7
11.06.10
• Modelling and analyzing a hybrid engine controller
• EAST-ADL (Papyrus)
• Automotive model
• HRC (A/P-Editor)
• Common model
• HRC (Dominance analysis)
• Model with contracts
• ModelBus
• Model driven tool integration framework
• Plugged Papyrus, A/P-Editor, Dominance Analysis Service
• Plugged model transformation service (QVT, EAST-ADL HRC)
• Model repository
• Common meta-model provides common interface to requirements and structure
Scenario
8
11.06.10
• A common-meta-model for the interoperation of tools with heterogeneous data-models
• No combination of all elements of different data-models
• Provides common concepts
• Hierarchical components with data flows and behaviour
• Requirements, traceability, abstraction levels, verification / validation, product lines
• … (to be identified)
• Mapping between data-models and common-meta-model required
• Provide common view / interface for a heterogeneous set of model artefacts
Conclusion