Institute for Software Integrated Systems Vanderbilt University MODEL-INTEGRATED DESIGN IN SOFTWARE, SYSTEMS AND CONTROL ENGINEERING Janos Sztipanovits ISIS, Vanderbilt University SERC Workshop October 5, 2011
Model-Based Design Tools Key Idea: Use models in domain-specific design flows and ensure that final design models are rich enough to enable production of Domain Specific artifacts with sufficiently predictable properties. Design Automation Impact: significant productivity increase in design technology Environments: • Automotive Production • Avionics Domain-Specific Design Environments • Sensors… Facilities Requirements Tools: • Modeling • Analysis • Verification • Synthesis Challenges: • Cost • Benefit only narrow domains • Island of doTransition (fsm as FSM, s as State, t as Transition) = Mathematical and require s.active Automation step exitState (s) step if t.outputEvent <> null then physical foundations emitEvent (fsm, t.outputEvent) step activateState (fsm, t.dst)
Metaprogrammable Design Tools Key Idea: Ensure reuse of high-value tools in domain-specific design flows by introducing a metaprogrammable tool infrastructure. Domain Specific VU-ISIS implementation: Model Integrated Computing (MIC) tool Design Automation suite ( http://repo.isis.vanderbilt.edu/downloads /) Environments: • Automotive Production • Avionics Domain-Specific Design Environments • Sensors… Facilities Requirements Semantic Backplane Metaprogrammable Tool Infrastructure • Model Building Metaprogrammable • Model Transf. Tools, Environments • Model Mgmt. • Tool Integration Explicit Semantic Foundation doTransition (fsm as FSM, s as State, t as Transition) = Semantic Foundation • Structural require s.active step exitState (s) step if t.outputEvent <> null then Component Libraries emitEvent (fsm, t.outputEvent) • Behavioral step activateState (fsm, t.dst)
Use Case 1: Cyber Physical Systems Battery Components ISG Engine span: Transmission VMS • Multiple Servos physics /Linkages • Multiple domains • Multiple tools Physical Cyber Cyber-Physical Functional: Computation and Physical with implements some communication deeply embedded function in the that implements computing and design some function communication Interconnect: acts Requires a physical as the facilitators platform to run/to for physical communicate interactions DARPA AVM Program
CPS Design Flow Requires Model Integration Architecture Design Integrated Multi-physics/Cyber Design Detailed Design Modeling Analysis Exploration Modeling Modeling Simulation V&V SW Physics-based Deep Structure/CAD/Mfg Rapid exploration analysis Exploration with integrated optimization and V&V • Architecture • Architecture Modeling • Architecture Modeling • Design Space + Constraint Modeling • Design Space + Modeling • Dynamics, RT Constraint • Dynamics Modeling (ODE) Software, CAD, Modeling • Computational Behavior Thermal, … • Low-Res Modeling • Detailed Domain Component • CAD/Thermal Modeling Modeling (FEM) Modeling • Manufacturing Modeling Domain Specific Modeling Languages
Model Integration Challenge: Physics Heterogeneity of Physics Electrical Mechanical Hydraulic Thermal Domain Domain Domain Domain Theories, Theories, Theories, Theories, Dynamics, Dynamics, Dynamics, Dynamics, Tools Tools Tools Tools Physical components are involved in multiple physical interactions (multi- physics) Source of resilience: explicit modeling of multi-physics interactions.
Model Integration Challenge: Implementation Layers B t ( ) ( B t ( ),..., B t ( )) Dynamics: = κ Plant Dynamics Controller p 1 j • Properties : stability, safety, performance Models Models • Abstractions : continuous time, functions, signals, flows,… Physical design Heterogeneity of Abstrac<ons Software : B i ( ) ( B i ( ),..., B i ( )) = κ Software Software c 1 k • Properties : deadlock, invariants, Architecture Component security,… Models Code • Abstractions : logical-time, concurrency, Software design atomicity, ideal communication,.. System Resource Systems : B t ( ) ( B t ( ),..., B t ( )) = κ j p 1 i k i Architecture Management • Properties : timing, power, security, fault Models Models tolerance • Abstractions : discrete-time, delays, System/Platform Design resources, scheduling, Source of resilience: systems science principles for decoupling across design layers (such as passive dynamics to decouple stability from implementation induced time-varying delays
Model Integration Language Model Integra:on Language (MIL) Seman<c Backplane Hierarchical Ported Models /Interconnects Structured Design Spaces Meta‐model Composi<on Operators MIL Pro‐E MIL SL/SF abstrac<on abstrac<on abstrac<on MIL TD SL/SF CAD SEER Sem. IF Sem. IF Sem. IF MIL CAD SEER‐MFG Thermal Desktop Pro‐E CATIA Tools and Frameworks Assets / IP / Designer Exper:se Impact : Open Language Engineering Environment Adaptability of Process/Design Flow Accommodate New Tools/Frameworks , Accommodate New Languages
Use Case 2: “ C2 Wind Tunnel ” Mixed Context Dep. Adaptive Human Initiative Command Resource Controllers Controller Interpretation Allocation Assigned Abstract HCI Platform Commands Platform Commands Commands Decision Coordination Support Platform Status Data Distribution Network Model-Based Experiment Integration Environment: C2WT Issues to be studied experimentally: • Information Sharing • Distributed Command and Control – Synchronization and coordination – Shared situation awareness – Distributed dynamic decision making – Common Operation Picture (COP) – Network effects – Network effects • Advanced Cooperative Control – Cooperative search algorithms AFOSR PRET Program
Heterogeneous Simulation Integration Processing (Tracking) Controller/Vehicle Dynamics 3-D Environment (Sensors) Organization/Coordination Devs Delta3D SL/SF CPN CPN Adaptive Mixed Context Dep. Adaptive Human Initiative Command Resource Organization Controller Interpretation Allocation How can we integrate the models? Assigned HCI Abstract Platform Commands Platform How can we integrate the simulated heterogeneous system components? Commands Commands Decision Coordination Support How can we integrate the simulation engines? COP Platform COP COP Elements Status Elements Elements Data Distribution Network Model-Integrated System and Software Laboratory Environment: C2 Windtunnel GME GME Simulation Interaction Simulation Architecture OMNET Network Architecture
Model Integration Architecture in C2WT Simulator Integration models Dataflow models Delta3D RTDS Simulink Interaction models Simulink Federate(s) Delta3D Federate G e n e r HLA‐RTI a t o r s OMNet Federate CPN Federate Deployment models CPN OMNet
Simulation Integration Architecture in C2WT Experiment Model Integration Layer Specification & Configuration Component Controller Network Env. Org. Fusion Models Models Models Models Models Models Models Run-time Simulink DEVS OGRE OmNet++ CPN Federate Federate. Federate Federate Federate. Instrumentation Layer Instrumentation Layer Simulation Integration Platform (HLA) Simulation Data Distribution/Communication Middleware Distributed Simulation Platform https://wiki.isis.vanderbilt.edu/OpenC2WT
Example: Simulink model integration (Vehicle dynamics) GME integration model Original model (X4 simulator) Add input-output bindings Input binding Code generation Output binding Modified model Generated .m Receiver and Sender S-function code + .java code for representing Simulink federate RTI runtime communication Signal flow Signal flow HLA Run-Time Infrastructure (RTI)
Experiments: Impact of Cyber Attacks Network attack: A sub-network with hundreds of zombie nodes attacks a critical router on the main network. Flood attack on udp, tcp or ping Zombie subnet Full network
Summary Questions: What are challenging systems application domains? Heterogeneous SoS domains (like CPS and C2). How does practice diverge from theory, and how do we connect? Precise compositionality is hard to achieve in heterogeneous systems, still, we need predictability. Need systems science principles for simplifying interactions and dependences (decoupling). Where are relevant technologies to be found? In cross-disciplinary interactions. E.g. scalability in embedded software verification may require tradeoffs in systems dynamics. What would be the most critical tools and products? Component-based and model-based design approaches and tools are and will be increasingly essential.
Example: Architecture Modeling Sublanguage Formalism, Language Constructs, Examples Usage / Capability Systems Hierarchical Architect Module - Explore Architecture Interconnect Design Modeling - Components Space - Interfaces - Derive - Interconnects Candidate - Parameters Designs - Properties Hierarchically Systems Layered Architect Design Parametric - Define Space Alternatives Design Modeling - Alternatives/ Space Options - Define - Parameters Constraint - Constraints
Recommend
More recommend
