Electrical system 1927 1997 Electrical system 1927-1997 Virtual - - PowerPoint PPT Presentation

electrical system 1927 1997 electrical system 1927 1997
SMART_READER_LITE
LIVE PREVIEW

Electrical system 1927 1997 Electrical system 1927-1997 Virtual - - PowerPoint PPT Presentation

Roger Johansson/ 2010 Roger Johansson/ 2010 A premium passenger car is controlled and A premium passenger car is controlled and Time triggered real time communication Ti t i d l ti i ti managed by 80+ Embedded Systems Presentation


slide-1
SLIDE 1

Roger Johansson/ 2010

Ti t i d l ti i ti Time triggered real time communication

Presentation overview

 Background

automotive electronics, an application area for time triggered communication.

 Time triggered protocols

TTPC, first commercial implementation. Originally from TU Vienna. Operational p g y p in civil aircrafts. TTCAN, based on Controller Area Network (CAN) which is widely used in today's vehicular electronic systems today s vehicular electronic systems. FlexRay, based on BMW’s “ByteFlight”. Anticipated in next generation automotive electronic systems.

 Hybrid scheduling

combining static scheduling with fixed priority scheduling analysis.

Time triggered real time communication 1 Roger Johansson/ 2010

A premium passenger car is controlled and A premium passenger car is controlled and managed by 80+ Embedded Systems

Comfort Electronics: Thermal Management Chassis Control Infotainment: Telematics Solutions Car PC Wireless Connectivity Chassis Control Parking Assistant Wireless Connectivity Car-to-car communication Floating Car Data Po ertrain Powertrain: Engine Management Transmission Control Power Management Safety: Predictive Safety Systems Driver Assistance Systems Adaptive Cruise Control

Time triggered real time communication 2

g Adaptive Cruise Control Electric Power Steering

Courtesy of DaimlerChrysler, Bosch Roger Johansson/ 2010

Vi t l diff ti ti b t i t

Entertainment

Virtual differentiation between variants

  • All variants of a specific

Variant 1

configuration A

model are physically identical and differ only in their individual software configuration

Motor configuration A

  • The various included

physical components can be activated or deactivated by the software

Motor configuration Entertainment configuration F

the software

g B

Variant 2

Time triggered real time communication 3 Roger Johansson/ 2010

Electrical system 1927 1997 Electrical system 1927-1997

54 54

1200 1200

27 27

  • No. of fuses
  • No. of
  • No. of

meters of meters of electric electric wires wires 16 16 27 27

575 575

wires

Wiring diagram, ÖV4 (“Jacob”) 1927

4 4 5 5 7 7 9 9 16 16

283 283 183 183 83 83

Wiring diagram, ÖV4 ( Jacob ) 1927

1927 1927 1975 1975 1982 1982 1944 1944 1997 1997 1966 1966 1956 1956

4 4

83 83 50 50 30 30 1927 1927 1975 1975 1982 1982 1944 1944 1997 1997 1966 1966 1956 1956

Time triggered real time communication 4

slide-2
SLIDE 2

Roger Johansson/ 2010

The evolution of the electrical system

Power production

Features

Architecture Optimisation on many levels Standardised interfaces and distribution Simple components

300 350 400 450

More complex functions stand-alone systems ABS, Airbag

100 150 200 250

# of functions # of

Integration of systems Optimisation of information Common data busses

50 100 1930 1940 1950 1960 1970 1980 1990 1995 2000 2005

integrated functions

1970 1980 1990 2000 2010

Time triggered real time communication 5 Roger Johansson/ 2010

A t ti l t i d Automotive electronics roadmap

Time triggered real time communication 6 Roger Johansson/ 2010

The current electrical system The current electrical system

Mirror CAN Light Window Lift Lock Lock Universal Light CAN Light Seat Htng Instruments Htng Wiper Power Train ITS Interior Light g Central Body Ctrl Climate Roof x6 Htng p Trunk WHtg CAN Seat Htng Light St-Wheel Panel Seat Sub-Bus Lock Mirror Lock Universal Motor Universal Panel Mirror Time triggered real time communication 7 Roger Johansson/ 2010

Multiplex Networks

Conventional system Network

Identifier Identifier Data Data Command Command Control Control

system

Engine Control Control units Module Automatic Transmission Driver Information Central Module

Time triggered real time communication 8

slide-3
SLIDE 3

Roger Johansson/ 2010

B i t l

Electronic

By-wire control

Electronic information carrier Hydraulic information carrier

The F-8 Digital Fly-By-Wire (DFBW) flight research project validated the principal research project validated the principal concepts of all-electric flight control systems now used on nearly all modern high performance aircraft and on military high-performance aircraft and on military and civilian transports. The first flight of the 13-year project was on May 25, 1972.

Time triggered real time communication 9 Courtesy of Dryden Flight Research Center Roger Johansson/ 2010

Electronics in distributed control Electronics in distributed control

((( ))) ((( )))

traffic control train (consist) control

((( )))

traffic control train (consist) control

((( )))

local control

((( )))

local control

((( )))

Time triggered real time communication 10

((( )))

wayside control

((( )))

wayside control Roger Johansson/ 2010

Drive-by-wire

Time triggered real time communication 11 Roger Johansson/ 2010

Control system implementation strategies Control system implementation strategies

Local control

  • Local information processing
  • Independent control objects
  • Independent control objects

Centralized global control Centralized global control

  • Local and central information processing
  • Interconnected control objects

Distributed global control

  • Local and distributed information processing
  • Interconnected control objects

Time triggered real time communication 12

slide-4
SLIDE 4

Roger Johansson/ 2010

N f ti l i t Non-functional requirements

System life Maintainability System life time Safety Changeability Interoperability Testability Maintainability Extendability Portability Restructuring

System

Usability Performance/ Efficiency

System Architecture

Availability Security Reliability Cost-effectiveness Robustness Fault tolerance Produceability Understandability Timeliness y Fault tolerance Variability (variants configurations) Conceptual integrity Variability (variants, configurations)

Time triggered real time communication 13 Roger Johansson/ 2010

T d ff f S f t /R li bilit i t Tradeoffs from Safety/Reliability requirements

The extremes from reliability requirements leads to safety requirements. y q y q Safety requirements implies redundancy, (Fail-Operational, Fail-Safe, etc). Safety requirements also demands predictability, we has to show, a priori, that

 In a distributed environment, only time triggered protocols and d d t b id thi f t C t TTP’

the system will fulfill it’s mission in every surrounding at every time.

redundant buses can provide this safety. Contemporary TTP’s are:

TTP/C, first commercial implementation. Originally from TU Vienna. Operational in civil aircrafts. in civil aircrafts. TTCAN, based on Controller Area Network (CAN) which is widely used in today's vehicular electronic systems. FlexRay, based on BMW’s “ByteFlight”. Anticipated in next generation automotive electronic systems.

Time triggered real time communication 14 Roger Johansson/ 2010

TTCAN

– Based on the CAN protocol – Bus topology Media: twisted pair

TTCAN

– Media: twisted pair – 1Mbit/s

A

Nod

Node 1 Node 3 Node 6 Node 7

CPU/mem /CC

S S S Node 4 Node 5 Node 2

A second controller is required to implement the redundant bus

Time triggered real time communication 15

A second controller is required to implement the redundant bus

Roger Johansson/ 2010

TTCAN TTCAN

”Exclusive” – guaranteed service ”Arbitration” – guaranteed service (high ID), best effort (low ID) ”Reserved” – for future expansion... p

Transmission Columns

Basic cycle 0 Basic cycle 1 Basic cycle 2 Basic cycle 3 Basic cycle 3 t

Time triggered real time communication 16

Time is global and measured in network time units (NTU’s)

slide-5
SLIDE 5

Roger Johansson/ 2010

TTP/C TTP/C

– Double channels (one redundant). Bus topology or ”star” (optical) – Media: twisted pair, fibre – 10 Mbit/s for each channel

CNI works as a “firewall” Status, global time, membership

Nod Nod 1 Nod Nod 3 Nod Nod 6

Control, clock interrupt Watchdog, checking consensus Data the actual message

Nod Nod 4 Nod Nod 6 Nod Nod 5

A

CPU/mem

Nod

C

4 2 5 Nod Nod 1 Nod Nod 4

S S S

/CC

CNI

A

1 Nod Nod 2 4 Nod Nod 5

S

B

Nod Nod 3 Nod Nod 6 Time triggered real time communication 17

A network is built on either twin buses or twin stars.

Roger Johansson/ 2010

TTP/C TTP/C

All communication is statically scheduled

Guaranteed service

”TDMA-round”

”message slots” message slots N i di l h t b fitt d i t t ti l t b th li ti t

Time triggered real time communication 18

Non periodical messages has to been fitted into static slots by the application

Roger Johansson/ 2010

Flexray Flexray

– Double channels, bus or star (even mixed). – Media: twisted pair, fibre

– 10 Mbit/s for each channel

N d 3 Nod 1 Nod 3 Nod 6

B

Nod 7

A CPU/mem/ CC Nod

A B

S S S CC

Nod 4 Nod 2 Nod 5

R d d t h l b d f lt ti h d l

Time triggered real time communication 19

Redundant channel can be used for an alternative schedule

Roger Johansson/ 2010

Flexray

”Static segment” (compare TTCAN ”Exclusive”) – guaranteed service

Flexray

”Dynamic segment” (compare TTCAN ”Arbitration”) – guaranteed service (high ID), ”best effort” (low ID)

63 62 N Guaranteed periodical Guaranteed periodical/ aperiodical ”Best-effort” aperiodical Network Idle Symbol win 3 2 1 e Time ndow Static segment (m slots) Dynamic segment (n mini-slots)

M 64 d Fl t k

Time triggered real time communication 20

Max 64 nodes on a Flexray network.

slide-6
SLIDE 6

Roger Johansson/ 2010

Comparisons Comparisons

All protocols targets real time applications.

TTCAN and Flexay combines time AND event triggered paradigms well.

All protocols are suitable for scheduling tools.

TTP/C has commercial production tools. Tools for TTCAN and Flexray are ti i t d y gg g anticipated.

CAN, many years experiences, a lot of existing applications. CAN, many years experiences, a lot of existing applications.

Implies migration of existing CAN applications into TTCAN.

TTP/C considered as complex TTP/C considered as complex.

Poor support for asynchronous events. High complexity, lacks second (or multiple) sources.

Flexray is the latest initiative.

Supported by most automotive suppliers.

Time triggered real time communication 21 Roger Johansson/ 2010

C bi i ti t i i ith t Combining time triggering with events: Exam ple of Hybrid scheduling for TTCAN

Messages are sorted into three different categories: Messages are sorted into three different categories:

 Hard real-time, for minimal jitter with guaranteed response time.  Firm real-time, for guaranteed response time, but can tolerate jitter. , g p , j  Soft real-time, for “best effort” messages.

Time triggered real time communication 22 Roger Johansson/ 2010

TTCAN d t il d TTCAN detailed study

Q T Response time analysis B i i i i

Q T B R   

Time triggered real time communication 23 Roger Johansson/ 2010

Ti t i d

Transmission Columns time windows

Time triggered messages Mh

Basic cycle 0 Basic cycle 1 Basic cycle 2 time windows

After structuring: M : {Mh, Mf, Ms}, assume that at least Mh is defined. We now construct a matrix cycle Due to protocol constraints the schedule has to fulfil:

Basic cycle 3

  • cycle. Due to protocol constraints, the schedule has to fulfil:

LCM( Mh

p ) = x 2n

where: LCM i l t lti l i d f th Mh t  LCM is least common multiple period for the Mh message set;  x is the preferred length of a basic cycle within LCM;  n is the number of basic cycles. Hardware constraints: Hwc1: 1 ≤ x ≤ 2y, has to be consistent with a hardware register, y bits Hwc2: 0 ≤ n ≤ k, always a power of 2, constraint in hardware. H 3 # f t i ≤ T l i th t i l Li it d b th b f Hwc3: # of triggers ≤ Tr, columns in the matrix cycle. Limited by the number of available trigger registers.

Time triggered real time communication 24

slide-7
SLIDE 7

Roger Johansson/ 2010

Multiple solutions satisfies the equation...

Choose a strategy: Strategy 1 :

Minimize number of basic cycles, requires a longer basic cycle, and more triggers.

Strategy 2 :

Minimize length of basic cycles, increase probability of finding a feasible schedule for large message

Time triggered real time communication 25 Roger Johansson/ 2010

Persuing the strategies Persuing the strategies...

Construct a schedule for the following set: M h = ( M1, M2 , M3) with the following attributes (NTU): M1 = 1000 M1 = 168 M1p = 1000, M1e = 168 M2p = 2000, M2e = 184 M3p = 3000, M3e = 216 It’s obvious that: LCM( M1, M2 , M3 ) = 6000. and: 6000 = x 2n

Time triggered real time communication 26 Roger Johansson/ 2010

Strategy 1

Minimizing number of basic cycles yields: 2n = 1, so n = 0 and x = 6000. H 1 d H 2 f lfill d

Strategy 1

Hwc1 and Hwc2 are fulfilled. Total numbers of triggers for N messages in one basic cycle is:

N

) LCM( M

in this case:

 i i

M ) (

1

6000 6000 6000

# of triggers = So, strategy 1, leads to a solution with:

11 3000 6000 2000 6000 1000 6000   

So, strategy 1, leads to a solution with:  1 basic cycle and 11 triggers.  MAtrix cycle length is 6000 NTU.

Basic Cycle Triggers

168 352 1000 2000 2168 3000 3352 4000 4168 5000 M1 M2 M3 M1 M1 M2 M1 M3 M1 M2 M1

Time triggered real time communication 27 Roger Johansson/ 2010

Strategy 2

Basic cycle 1 (at 0) 2 (at 375) 3 (at 750) 4 (at 1125) 5 ( 1500)

  • 168
  • 352
  • 1000
  • Trigger

Information

Strategy 2

n = 0: 6000 = x 20  x = 6000

5 (at 1500) 6 (at 1875) 7 (at 2250) 8 (at 2625) 9 (at 3000) 10 (at 3375)

  • 3000
  • 2000
  • 2168
  • 3352
  • 6000 = x 20

 x = 6000 (same as strategy 1) n = 1: 6000 = x 21  x = 3000

11 (at 3750) 12 (at 4125) 13 (at 4500) 14 (at 4875) 15 (at 5250) 16 (at 5625)

  • 4125
  • 4168
  • 4000
  • 5000
  • Minimum

Triggers

6000 = x 21  x = 3000 n = 2: 6000 = x 22  x = 1500

1 M1 M2 M3 3 2 3 M1 1 4 5

n = 3: 6000 = x 23  x = 750 n = 4 :

6 M1 M2 2 7 8 9 M1 M3 2 10

n = 4 : 6 0 0 0 = x 2 4  x = 3 7 5 n = 5: 6000 = x 25  x = 187 5

11 M1 ? 1 12 ? M1 M2 2 13 14 M1 1 15 Time triggered real time communication 28

6000 = x 25  x = 187.5

16

slide-8
SLIDE 8

Roger Johansson/ 2010

Strategy 2

Avoid this conflict with the requirement that: a basic cycle shall be at least as long as the shortest period in the message set

Strategy 2

a basic cycle shall be at least as long as the shortest period in the message set. Applying this restriction we get: n = 2 (x = 1500) n = 2, (x = 1500) which yields a feasible schedule:

Basic cycle 1 2 3 4

  • 3000
  • 168
  • 352
  • 3352
  • 2000
  • 5000
  • 2168
  • 4000
  • 1000
  • 4168
  • Trigger

Information Minimum Triggers 4 5000 Triggers

1 M1 M2 M3 M1 4 2 M1 M2 2 3 M1 M3 M1 M2 4 4 M1 1

Time triggered real time communication 29 Roger Johansson/ 2010

Verifying the events (Mf) Verifying the events... (Mf)

Grey slots are supposed to be allocated for M h Basic Cycle NTU-slots (Columns) 1 q0 2 q1 q2 3 q3 q4 q5 3 q3 q4 q5 ….. … … … … 2n qN-3 qN-2 qN-1

for each message m in M f : for message m = 1 up to last_m for virtual message VMi = 1 up to last_VM if( Qm + Tm ) falls within ( VMi,start , VMi,completion ) Qm = VMi,completion Qm

i,completion

else

j P P j j m m

T t Q Q

 

        

1 :

endif end end end

P P j j

j m

   

:

Time triggered real time communication 30

end

Roger Johansson/ 2010

Conclusions Conclusions

 Applicable real time communication protocols for future

safety-critical applications has to provide strictly periodical y pp p y p (minimal jitter), periodical (jitter is negliable) and a-periodic communication to fully support control applications. Scheduling periodical and a-periodical events requires a new approach, hybrid scheduling.  Hybrid scheduling is sparsely found in today’s literature...

Time triggered real time communication 31 Roger Johansson/ 2010

Ti t i d l ti i ti Time triggered real time communication

Thank you for your attention

Time triggered real time communication 32

Thank you for your attention.