Upload
phunghanh
View
213
Download
0
Embed Size (px)
Citation preview
2017
Adopting Model Based System Development
Joshua Spiegel, Brian Ladson, M. Michael BriggsPresented by: Brian Ladson
Modeling and Simulation [email protected]
Millennium Engineering and Integration CompanyNovember 1, 2017
Millennium Engineeringand Integration Company
1
2017
Simulation-Driven Development• “An approach to engineering that uses models as an integral part
of the technical baseline that includes the requirements, analysis, design, implementation, and verification of a capability, system, and/or product throughout the acquisition life cycle.” Final Report, Model-Based Engineering Subcommittee, NDIA, 2/2011
• Traditional Systems Engineering• Manual processes executed throughout top-down decomposition &
bottom-up realization, i.e. the ‘V’• Process lacks verification that requirements are satisfied
• Model Based Engineering/System Development (MBE or MBSD) fits naturally into the workflow with Models & Simulations integrated into each process phase
• To fully take advantage of MBSD for complex systems, tools that can simulate continuous and discrete time dynamics and generate real-time code are needed
• MATRIXx, LabVIEW, MATLAB / Simulink, SCADE, etc have Automatic Code Generation
2Simulation-Driven Development Achieves Substantial Reduction in
Software Life Cycle Cost & Improves Software Operational Reliability
2017
Seamless Model-Based Engineering• Model Based Systems Engineering (MBSE)
& Model Based Engineering (MBE) are an integral part of the Systems Engineering ‘V’
• Graphically-programmed models enable simulation-driven concepts, design trades, and early optimization
• Model Based Development substantially reduces development time / cost of complex systems• Engineering automation software tools
accelerate design via a re-usable library of pre-programmed functional objects with an integrated simulation engine
• System simulation availability speeds up trade studies and sensitivity analyses
3
Algorithm Blocks
Model Software
Model Based Development Process
Simulink Template
Improved Coding Accuracy and Reduced Development Time
2017Model Based Systems DevelopmentToolchain & Processes
4
Requirements
& Traceability
UML/SysML
Diags Model
Files &
Code
Scripts &
Code Integ.
w/Middleware
Requirements
Database Test
Results
DB
Block-diagram Sim Models
of HW & SW functions• NRT System Simulation
Code Generator
Embedded
Target Processors• RT Plant Simulator for PIL/HIL
• Flight Processor or Surrogate
External Simulation
Frameworks
Requirements
Management Tool
Real-Time Integrated
Development Environmnt• RTOS & Runtime environment
• Compiler, Linker, Loader etc
• BSP & IPC software
CM/VC• Config Mgmnt
• Version Control
• Build Management
Load
Module
Standalone Model
Builds
Generated RT
Source Code
& Make Files
Test
Vectors
External Code
Comms Middleware, Code Gen Control & Parsing Scripts
Legacy Models & Code• C++ Source Code, Simulink Blk Diags
UML/SysML Tool• Structure/Object Diagrams
• State Charts
• Sequence Diagrams
Back-propagate changes
2-way Translator
Requirements
& Traceability
Model
Files
Model File
Test
Results
DB
UML/SysMLDiags
Database
Block Diagram
Model Files
Middleware(optional)
Ext
Test
Results
DB
MBSE & MBE Tools
MBE Tools
Multiphysics Model Data
2017
Phased Strategic MBSD Adoption Approach• Investment Required:
• Significant Adaptation in Corporate Software Development Process
• Purchase of Multi-User Licenses for Multiple Mathworks Software Tools: MATLAB/Simulink/RTW/Stateflow, etc.
• Training / Retaining of Personnel• Living With Resistance to Change and Perception of
Increased Performance Risk Until Benefits Have Been Demonstrated
• Transition Legacy Processes and Code to align with MBD
• Expand Modeling Guidelines, Model Libraries, and V&V Process
5
Investment
Be
nef
it
A Phased Strategic Approach and Key Investments are Required to Adopt Model Based System Development and Realize its Benefits
2017
Transitioning Legacy Processes and Code• Legacy Models, Simulations, and Design Processes represented as
MANY lines of code• Legacy Code & Processes need NOT be abandoned – most can be used
and maintained in the Model Based Development Environment• Rapid Plug-and-Play, Drag-and-Drop architecture leverages legacy
models and simulations
• Many organizations interested in transitioning to Model Based Development Processes• Why don’t they OR haven’t they transitioned yet?
• Time, Cost, OR Lack of Path to Convert Large Scale Proven Legacy Models and Simulations to Model Based Environment
• How can they transition?• Legacy Models and Simulations are transitioned to the Model Based
Environment via Automated Tools which reduce Time, Cost, and Human Error which were once precluding factors
6Legacy Models and Simulations are Integrated Directly or Transitioned
to the Model Based Development Environment
2017
Expanding Modeling Guidelines & Model Libraries
• Common model repository enables software verification and validation (V&V) in a parallel model environment before implementing on critical operational assets
• Automated evaluations ensure Model Based Systems are developed, implemented, and integrated as intended and follow set modeling guidelines for ease of readability & requirements traceability
• Well designed, Integrated Models Facilitate Configuration Control and Rapid Change Implementation • Traceability between requirements and other model elements become a set of living linkages
that provide key insight into system dependencies
7Expanding Modeling Guidelines & Model Libraries Enables a Seamless
Model Based Systems Development Process
2017
Automatic Code Generation• Automatic Code Generation is Vital to Model Based Development Processes
• Leverages graphical programming language to rapidly prototype simulations in Monte Carlo space and integrate with external software / hardware
• Hours-to-days to accomplish initial capability assessments early in system life cycle
• Custom Code Generation Tools target desired Middleware and significantly reduce time required / error associated with transitioning prototype to an integrated system• Automatic Code Generation techniques utilized to produce real-time software products for
complex systems including flight vehicles
8
2017MBSE Payoff Examples:Telerobotic Remote Ops With Latency Compensation& Virtual Haptic Feedback (for DARPA Phoenix)
9
2017
MBSD Applied to NASA Ames LADEE Flight Software• Lunar Atmosphere and Dust Environment Explorer (LADEE)
• Lunar orbiting spacecraft• Structure based on modular bus developed by NASA• Objectives
• Measure dust & atmosphere properties around Moon• LaserCom experiment
• Launch: Sept 2013, Wallops, on a Minotaur V, 7-Month Mission• Millennium Roles since early 2009
• Flight Software Development • Real-time Simulation Development• Mission Ops Support
• Rapid Development Infrastructure• Simulink-centric Model based Development with Embedded Coder• Loosely-Coupled Architecture (cFE)• Early and Frequent Integration
• Early Integration in P/HIL Testbeds• Evolving fidelity as H/W becomes available• Early Integration of Device Protocol Software
• Posix I/O Drivers Approach• Facilitates Mix’n’Match Evolution of transports• Meets LADEE Testbed Implementation Goals 10
2017
LADEE Loosely-Coupled Flight Software Environment
11
Command
& Mode
Processor
Actuator
Manager
State
Estimator
Safe Mode
Controller
Attitude
Control
System
Thermal
Control
System
Power
Control
System
SchedulerLimit
Checker
Stored
Commands
Memory
Scrub
Memory
Manager
Memory
Dwell
Hardware
I/O
Health &
Safety
Data
Storage
File
Manager
Telemetry
Output
CCSDS File
DeliveryChecksum
Command
Ingest
House-
keeping
Software Bus
Telemetry
Gnd Cmds
Sensor Data
OFSW
System Support, O/S Services and Device Drivers
Battery
Charge
System
FSW Internal
FSW ExternalSimulink
Task
cFS
Task
Hand
Written
Task
KEY
2017
LADEE Processor/Hardware-In-Loop Simulation: Spacecraft Sim & Embedded Code Generated by Simulink Embedded Coder
• Flight code generated from Simulink, runs on Flight Avionics EDU• Provides testing of FSW with Avionics I/O• Definitive answers on resource utilization• Highest fidelity simulations for verification/validation 12
2017MBSD Payoff Example:Payload Modular Open Architecture
13MBSD Enabled Rapid Integration, Test, and Proof of a Customer Defined
Architecture using Models & Algorithms from Common Repository
2017International Space StationFlying 1,000,000 Lines of Autocoded ADA• Pointing & Support Portion of Guidance,
Navigation & Control Flight Software• Flight Code for All Onboard Experiments
• Solved Boeing’s problem of verifiying & validating mandated Ada code from dozens of experiment suppliers
• All suppliers used the MATRIXx/SystemBuild/Autocode toolset to develop Ada code and test it on a Space Station “Application Generator” with ISS Real-time target processor
• Original intent was rapid-prototyping: NASA decided to use 1,000,000 lines of Autocoded ADA inflight when it was discovered that the generated code was bulletproof
14
ISS
2017PhantomWorks Post-Transition-to-MathWorksExperimental Vehicles: X-40a, X-43• X-40A Space Maneuver Vehicle (9000 ft drop
test to autolanding)• Simulink simulation validated against NASA
Fortran sim• 6 weeks to initial code delivery• 1 week turn on changes
• X-43 (Hyper-X) Scramjet Research Flight Experiments• Flight software for propulsion & flight control
systems developed using Simulink & RTW• Reduced development time by months• Aided Boeing in achieving SEI CMMI Level 5
software rating
15
X-43 (Hyper-X)
X-40A
2017
Summary: Model Based System Development Payoffs• Graphically-programmed models enable simulation-driven concepts, design trades,
and early optimization• Integrated Engineering Toolkit Enables Rapid Design/Development• Rapid Plug-and-Play, Drag-and-Drop architecture leverages legacy models and sims• Analysis activities that used to take months are now measured in hours or days• Modifications and technology injections are accurately characterized much quicker
than legacy development processes allow• We provide support and training for customers who want to adopt advanced MB
engineering techniques and processes
16MBSD Substantially Reduced Development Time and Cost, Increased Product
Robustness and Reliability, & Accelerated Integration, Test, and Validation