Simulation to Implementation: Hardware in the Loop Simulation and Code Generation

1
Simulation to Implementation Hardware in the
Loop Simulation and         Code Generation

• Steve Neuendorffer
• Edward Lee
• UC Berkeley

2
Outline

• Embedded Systems
• Hybrid Embedded Systems
• Component-Based Design and Embedded systems
• The Caltech Ducted Fan vehicles

3
Embedded Systems
Small, Cheap Resource Constrained Ubiquitous
interaction Digital Computation is
commonplace Digital Logic Microprocessors
4
Hybrid Embedded Systems
Embedded computation where interaction with the
physical world is critical.
Electronic Thermostat
Furuta Pendulum
Neutrino Detection
Non-Real Time Control Systems (Path planning)
Data Acquisition Systems
Real Time Control Systems (Servo)
Constant and Multi-mode Dynamics Constant and
Multi-mode Computation
5
Design Pressures
Market Customization
Increasing Complexity
Safety Requirements
Design Reuse is Key!
6
Component-Based Design
Actor-oriented models have been applied to
embedded software. Sequential and Concurrent
systems, Timed and Untimed Systems
Example Model PSK baud rate detector
7
Continuous Time State models
A bridge for representing the Physical world
Spring Constantk Mass m Initial Position x0
f(x) -kx x f(x)/m
X0
X
8
Code Generation
Automatic code generation enables rapid
implementation from high-level component-based
design. Weve developed a code generation
technique based on component specialization that
transforms Ptolemy II models into a Java system
implementation.
9
Abstraction
Using low-level tools gives great control over
the end results. High-level tools insulate
designers from many implementation
details. Incorporate appropriate low-level
details into high-level models, leveraging
automatic code generation.
original child (Iron on Copper)
10
Modeling and controlling the Caltech Ducted Fan
Vehicle
11
Caltech Vehicles

• Difficulties
• Complex control problem
• Complex implementation platform

Localization computer estimates vehicle locations
Command computer Waypoints, trajectories, Control
changes
30 feet
20 feet
Vehicles with onboard controllers and 802.11b
12
A Detailed Heterogenous Model
Measured Physical Parameters
Discrete Event model convenient for events that
do not occur at the same time.
Model of computation and communication delay.
Array of 50 Bytes TimeStamp, ID, X, Y,
Angle 60 times a second
Array of 3 Bytes 85, Left, Right Sent
immediately after controller computes value
13
A Detailed Heterogenous Model
Continuous time model of vehicle dynamics
Fan Thrust Map
Data formatting
14
A Detailed Heterogenous Model
Discrete-state model of vehicle software
Encapsulated Control Law
15
Whats the point?
Abstract models encapsulate complexity, and
emphasize the complexity that is
important. Automatic system synthesis leverages
the structure of good system models and avoids
creating a ball of code.
16
Partitioning and Refinement
Communication refinement to platform specific
communication
Partitioning results in a distributed system
Works well for event-based models since
components are highly decoupled from each other
17
Towards Implementation
802.11b
RS-232
18
Hardware-in-the-loop
Replace hardware-true simulation model with
actual vehicle. Allows validation of continuous
dynamics model, and hardware/software interface.
802.11b
RS-232
19
Simulation-in-the-loop
Code generation of the controller onto an
embedded platform. Allows validation of
generated code, and execution delay.
802.11b
RS-232
Embedded Java Platform
20
System Implementation
The generated code forms the final system
implementation.
802.11b
RS-232
Embedded Java Platform
21
However
The control engineers model ignores many
embedded system issues.
User input? In situ controller
updates? Misbehaving controllers?
Sensor error recovery? Sensor
degradation? Dropped network packets?
Variable computation speed?
Concurrent trajectory generation?
These require more detailed explicit models!
22
Controller Updates
Mobile model allows substitution of different
controllers
Simplified model of base station
Controller component transmitted over
publish/subscribe network
(Yang Zhao)
23
Conclusion
Heterogenous, component-based design addresses
many of the difficulties of implementing hybrid
embedded system. Discrete and Continuous
Models Code Generation Heterogenous
models Repeatable Concurrent Design Modeling
across abstraction levels