Embedded and Cyber-Physical Systems -selected topics-€¦ · 2 Courtesy of Kuka Robotics...

1 ECPS-S2014-Lecture1 (Stefan Resmerita)

Embedded and Cyber-Physical Systems

- Introduction -

Stefan Resmerita Summer 2014

2 ECPS-S2014-Lecture1 (Stefan Resmerita) Courtesy of Kuka Robotics Corp.

Cyber-Physical Systems (CPS): Orchestrating networked computational

resources with physical systems

Courtesy of Doug Schmidt

Power generation and distribution

Courtesy of General Electric

Military systems:

E-Corner, Siemens

Transportation (Air traffic control at SFO)

Avionics

Telecommunications

Factory automation

Instrumentation (Soleil Synchrotron)

Daimler-Chrysler

Automotive

Building Systems

Slide from Lee & Seshia

3 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Ken Butts

Toyota’s Direction

Infrastructure integration

Vehicle technologies

Our sustainable mobility strategy includes products, partnerships, the urban environment and energy solutions.*

*Toyota 2010 North American Environmental Report Highlights

Connections

4 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Cyber-Physical Systems

5 ECPS-S2014-Lecture1 (Stefan Resmerita)

When software error is combined with physical modeling error

6 ECPS-S2014-Lecture1 (Stefan Resmerita)

When sensing fails

• https://www.youtube.com/watch?v=O1L-rLZayWQ

7 ECPS-S2014-Lecture1 (Stefan Resmerita)

How can we deal with such complex systems?

8 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Modeling, Design, Analysis

•Modeling is the process of gaining a deeper understanding of a system through imitation. Models specify what a system does.

•Design is the structured creation of artifacts. It specifies how a system does what it does. This includes optimization.

•Analysis is the process of gaining a deeper understanding of a system through dissection. It specifies why a system does what it does (or fails to do what a model says it should do).

9 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Modeling

•Developing insight about a system, process, or artifact through imitation.

•A model is an artifact that imitates the system, process, or artifact of interest.

•A mathematical model is model in the form of a set of definitions and mathematical formulas.

10 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Model-Based Design

1. Create a model of all the parts of the embedded system Physical world Control system Software environment Hardware platform Network Sensors and actuators

2. Construct an implementation from the model Construction may be automated, like a compiler More commonly, portions are automatically

constructed

11 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Contents

• Modeling techniques

• Controller design (discrete event systems)

• System analysis (discrete event systems)

• Code generation and integration

• Practical projects: system modeling in Ptolemy II

12 ECPS-S2014-Lecture1 (Stefan Resmerita)

Modeling of continuous dynamics

13 ECPS-S2014-Lecture1 (Stefan Resmerita)

The two-tank example

T1 T2

v1 v2

F

x1

x2

14 ECPS-S2014-Lecture1 (Stefan Resmerita)

The problem

• Setting: – F(t)= c (constant)

– v1(0) = v2(0) = 0

– x1(0) > 0, x2(0) > 0

– v1(t) = c1 , v2(t) = c2 for t > 0

– c1 < c and c2 < c

– F can be switched either to T1 or to T2

• Requirement: find a suitable switching to keep both levels above zero.

15 ECPS-S2014-Lecture1 (Stefan Resmerita)

The Cyber Part

• Possible solution: a discrete controller

– if (when) xi == 0, switch to Ti (i=1,2)

• Advantage: Minimally interventive

• Is this a viable controller?

16 ECPS-S2014-Lecture1 (Stefan Resmerita)

The Physical Part

• Principle of mass conservation (assume isolated system, no work) Input mass = stored mass + output mass

F·t = M1(t) + M2(t) + (v1+v2)·t

M1(t) + M2(t) = (F-v1-v2)·t

What if (F-v1-v2) < 0 ?

17 ECPS-S2014-Lecture1 (Stefan Resmerita)

A mathematical model

22

.

11

.

22

.

11

.

2

1

vFx

vx

vx

vFx

18 ECPS-S2014-Lecture1 (Stefan Resmerita)

Solution

2212022

2111011

)(

)(

tvFtvxtx

tvtvFxtx

Does a viable controller exist? Check this:

0

0

2212

2111

tvFtv

tvtvF

19 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Another Example: Modeling Helicopter Dynamics

20 ECPS-S2014-Lecture1 (Stefan Resmerita)

Feedback Control Problem

A helicopter without a tail rotor, like the one below, will spin uncontrollably. Why?

Control system problem: Apply torque using the tail rotor.

21 ECPS-S2014-Lecture1 (Stefan Resmerita)

Physics teaser: helicopter spin

• Watch the two videos, with helicopters that are similar in construction.

http://www.youtube.com/watch?v=MhEXXgiIVuY http://www.youtube.com/watch?v=ijErOpjx_Ws • Why do they behave differently after loosing the

tail rotor control?

22 ECPS-S2014-Lecture1 (Stefan Resmerita)

Feedback Control

0 ,1

0 ,0)(

t

ttu

23 ECPS-S2014-Lecture1 (Stefan Resmerita)

Helicopter model in Ptolemy II

a Java class !

http://ptolemy.eecs.berkeley.edu/ptolemyII/

24 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Modeling Physical Motion

• Six degrees of freedom:

Position: x, y, z

Orientation: pitch, yaw, roll

25 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Torque

• Just as force is a push or a pull, a torque is a twist.

• Units: newton-meters/radian, Joules/radian

Ty(t ) = r f (t )

For a point mass rotating around a fixed axis:

26 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Rotational Version of Newton’s Second Law

27 ECPS-S2014-Lecture1 (Stefan Resmerita) Slide from Lee & Seshia

Simple Example

28 ECPS-S2014-Lecture1 (Stefan Resmerita)

Feedback Control