ICT CT Fra rame mework works for Analysi ysing ng, , Op - - PowerPoint PPT Presentation

▶
ict ct fra rame mework
SMART_READER_LITE
LIVE PREVIEW

ICT CT Fra rame mework works for Analysi ysing ng, , Op - - PowerPoint PPT Presentation

Mar 06, 2023 486 likes •800 views

ICT CT Fra rame mework works for Analysi ysing ng, , Op Optimiz imizing ing and Building ding Sma mart rt Grids ids and Ene nergy gy Awa ware re Bui uildings ldings Ki Kim m G. Lar . Larsen Aalborg University DENMARK

slide-1
SLIDE 1

ICT CT Fra rame mework works for

Analysi ysing ng, , Op Optimiz imizing ing and Building ding

Sma mart rt Grids ids and Ene nergy gy Awa ware re Bui uildings ldings

Ki Kim m G. Lar . Larsen – Aalborg University DENMARK

slide-2
SLIDE 2

CIS ISS S –

Center er For Embedded dded Softwar tware e Systems tems

Regional ional ICT Center er (2002 2002- )

  • 3 research groups
  • Computer Science
  • Control Theory
  • Hardware
  • Wireless Communication
  • 20

20 Employed

  • 25

25 Associated

  • 20

20 PhD Students

  • 50

50 Industrial projects

  • 10

10 Elite-students

  • 100

100+ + MDKK

  • ARTIST Design
  • ARTEMIS
  • ... ...

Kim G. Larse sen [2]

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Characte cteri ristica stica:

  • Dedicated function
  • Complex environment
  • SW/HW/Mechanics
  • Networked
  • Autonomous
  • Ressource constrained

: Energy : Bandwidth : Memory : …

  • Timi

iming g con

  • nstrain

straints

slide-3
SLIDE 3

ES are of

  • fte

ten Safet fety Critical itical

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [3]

300 horse power 100 processors

How to achieve ES that are:

  • correct
  • predicable
  • dependable
  • fault tolerant
  • ressource minial
  • cheap

..

Model-Based Based Develop lopment ment

slide-4
SLIDE 4

UPPAAL AAL Model el Checke ecker

Editor Simulator

Kim Larse sen [4]

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Verifier Performance Analyses

Discrete Control Concurrency Continuous Aspects Stochasticity Timing Constraints Resources

slide-5
SLIDE 5

UPPAAL AAL Model el Checke ecker

Editor Simulator

Kim Larse sen [5]

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Verifier Performance Analyses

Discrete Control Concurrency Continuous Aspects Stochasticity Timing Constraints Resources

slide-6
SLIDE 6

Ov Overview view

  • Da

DaNES NES

  • Climate Control – Synthesis
  • Bright

ight Green reen

  • Optimal Scheduling
  • Homeport
  • MT

MT-LAB & IDEA4CPS CPS

  • Chinese Smart Grid
  • Danish Defense Building Administration
  • ENC

NCOURAGE OURAGE

  • Embedded iNtelligent COntrols for

bUildings with Renewable generAtion and storaGE

  • SENSATION

Futu ture re Green een Tech chnolog

  • gy

y - Japanese ese-Da Danish sh Joint t Work rksh shop, , 2012

Kim Larse sen [6] VKR-Center of Excellence Sino-Danish Basic Research Center

slide-7
SLIDE 7

Climat limate e Con

  • ntr

trol

  • l

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [7]

slide-8
SLIDE 8

Zo Zone ne-based based Climat limate e Con

  • ntr

trol

  • l

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

Kim Larse sen [8]

Temp[i+1] Temp[i] Temp[i-1]

slide-9
SLIDE 9

Ob Obje jectiv ctive e & St Strate ategy gy

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

Kim Larse sen [9]

Stragegy

1296 cases BDD 289 nodes

Objective control: A[] (Temp>12 && Temp<14)

Simulink

slide-10
SLIDE 10

Ob Obje jectiv ctive e & St Strate ategy gy

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

Kim Larse sen [10 10]

Stragegy

1296 cases BDD 289 nodes

Objective control: A[] (Temp>12 && Temp<14)

Simulink

Quasimodo

Kim G Larsen ESI Symposium, December 4, 2008 Page 14

Plastic c Injecti ction

  • n Molding

ding Machine ne

  • Robust and optimal

control

  • Tool Chain

– Synthesis: UPPAAL TIGA – Verification: PHAVer – Performance: SIMULIN LINK

  • 40% improvement of

existing solutions..

slide-11
SLIDE 11

The he Ene nerg rgy y Awa ware re Ho Hous use

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

Kim Larse sen [11 11]

slide-12
SLIDE 12

The he Ene nerg rgy y Plan lanner ner

Kim G. Larse sen [12 12]

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

slide-13
SLIDE 13

The he Ene nerg rgy y Plan lanner ner

Kim G. Larse sen [13 13]

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

slide-14
SLIDE 14

Sma mart rt Bui uildings ldings

Kim G. Larse sen [14 14]

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012
slide-15
SLIDE 15

Ho Home me Por

  • rt

ARC RCHITEC HITECTURE URE

HOMEPO EPORT

  • Important diversity of Home

Automation networks and protocols

  • No common interface to access and

control devices

  • HomePort provide a common

interface to heterogeneous networked devices

Future re Green n Techno nolo logy y - Japan anese se-Da Danish nish Joint Worksh shop, p, 2012 2012 Kim G. Larse sen [15 15]

TECHNOLO LOGIES IES

  • Based on well known standards:
  • RESTful architecture for scalable and

easy interfacing

  • TLS 1.0 for secured connections
  • ZeroConf for LaN service discovery
  • Implemented in C to fit resource

constrained systems

  • Targets Linux based systems
  • Log system for local and remote

monitoring

  • Part of several projects

(e.g. ENCOURAGE)

slide-16
SLIDE 16

Ho Home me Por

  • rt

t Proj

  • jec

ect t Facts ts

(www www.e .ener ergy gybo box.dk) x.dk)

  • Period: 2009 – 2011
  • Budget: 10.6 mill.kr.
  • Funding: 8 mill. kr. (2.6 mill. kr. for companies)
  • Knowledge institution partners: AAU (CISS), AU/Alexandra Institute,

CSI, IHA

  • Initial private partners: Develco, Seluxit, Servodan
  • Additional partners: Zense Techology, Danfoss Heating, LIAB,

NEOGRID

  • Also a number of networked partners

Kim G. Larse sen [16 16]

Futu ture re Green een Tech chnolog

  • gy

y - Japanese ese-Da Danish sh Joint t Work rksh shop, , 2012

GOALS

  • Implementation for co-existence of

different home devices and energy optimization

  • Validation of a number of scenarios
  • Collaboration with industrial partners
  • Investigate the business potential
slide-17
SLIDE 17

MT MT-LAB & IDEA4CPS CPS

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [17 17]

MT MT-LA LAB

Future Green Technolo

  • logy - Japanes-Danis

ish Joint Wor

  • rkshop
  • p, 2012

Kim Larsen n [21 21]

MT-LAB - a VKR Centre of Excellence is a research centre, which explores and develops methods for formal verification of modern advanced software systems. MT-LAB aims at solving the problems imposed by IT-systems becoming embedded in the world around us: how can we make sure that they behave safely and give us an acceptable performance?

IDEA4CPS PS Fou

  • undati

tion

  • ns for CPS

Futur ure Green n Techn hnology - Japane nes-Dani nish h Joint nt Worksho hop, 2012

Kim Larsen n [21 21]

I D E A

  • Inst. of Software Chinese Academy of Sciences,

Beijing, China Technical University of Denmark, Lyngby, Denmark East China Normal University, Shanghai, China Aalborg University, Denmark

Basic Research Centers

VKR Center of Excellence 40MDKK / 5y Sino-Danish Basic Research Center 25MDKK / 3y

slide-18
SLIDE 18

Chi hinese nese Sma mart rt Grid id

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [18 18]

Motivation – The Grid

  • E. Yüksel et al. TASE 2012

Modelling and Analysis of Smart Grid 3

Motivation – Storm…

  • E. Yüksel et al. TASE 2012

Modelling and Analysis of Smart Grid 4

Electricity workers repair the damaged power transmission tower in Yihuang County of south China's Jiangxi Province,

  • Feb. 12, 2008. (Xinhua Photo)

Ender Yksel, DTU Hanna R Nielson, DTU Flemming Nielson, DTU Huibiao Zhu, ECNU Heqing Huang, Wuxi SensingNet Inst.

slide-19
SLIDE 19

Sma mart rt Elect ctricity ricity Tow

  • wer

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [19 19]

Bone node Bone node Bone node Bone node Bone node

Electricity Transmission Line

 linear arrangement of towers

 100 towers  200m intervals

 Range of bone node:

 closest two neighbours – regular link  less closer two other neighbours – backup link

 Backup links are obviously more costly  A bone node transmits to the end that is cheaper to reach

  • E. Yüksel et al. TASE 2012

Modelling and Analysis of Smart Grid 7 WSN Bone node Wind direction & velocity sensor Line swing sensor Leaning degree measurement sensor Vibration transducer sensor

Sensor Bone node

Electricity Tower

 1 bone node, 50 sensors

 distance measurement sensor  voltage measurement sensor  short circuit fault sensor  vibration transducer  leaning degree measurement sensor  line swing sensor  wind direction sensor  wind velocity sensor

 Star topology  Single-hop

  • E. Yüksel et al. TASE 2012

Modelling and Analysis of Smart Grid 8

Goals

  • Low probability of errors
  • Good capabilities to recover

from errors

  • Good performance wrt energy

consumption

slide-20
SLIDE 20

Mo Modeling eling & Ana Analysi lysis

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [20 20]

Probability of node failure in the long run

  • E. Yüksel et al. TASE 2012
Modelling and Analysis of Smart Grid 25

the influence of MTBF is diminished for larger MTBF values. PRIOR to deployment

Impact of Repair Service on the Failure Risk

  • E. Yüksel et al. TASE 2012
Modelling and Analysis of Smart Grid 26

recovery times have a significant effect on the robustness of the system After the deployment, we are always able to improve (or sometimes impair) recovery capabilities

Expected Energy Consumption

  • E. Yüksel et al. TASE 2012
Modelling and Analysis of Smart Grid 27

we can assist the system management in determining a convenient time for replacement What is the best time for replacing the batteries?

Interacting

  • Cont. Time Markov Chains
slide-21
SLIDE 21

Ene nergy gy Awa ware re Bui uildings ldings

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [21 21] Danish National Defense Building Administration Pilot – Sjælland’s Odde

slide-22
SLIDE 22

Ene nergy gy Awa ware re Bui uildings ldings

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [22 22] Danish National Defense Building Administration Pilot – Sjælland’s Odde

slide-23
SLIDE 23

Mo Models els

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [23 23]

User Weather

slide-24
SLIDE 24

Ana nalysi lysis

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [24 24]

Temp Heaters

slide-25
SLIDE 25

ENCOURAG COURAGE

  • Embedded iNtelligent

COntrols for bUildings with Renewable generAtion and storaGE

  • 6,37MEUR
  • directly optimize energy use in

buildings

  • ptimal control of internal sub-

systems

  • enable active participation in the

future smart grid environment.

  • effective interaction with external

world, including other buildings, local producers, or electricity distributors.

Future re Green n Techno nolo logy y - Japan anese se-Da Danish nish Joint Worksh shop, p, 2012 2012 25 25

slide-26
SLIDE 26

Approac proach

  • Developing supervis

rvisory

  • ry

control rol strategi gies es that will be able to coordinate larger subsystems

  • Development of an

intel elligent ligent gateway ay with embedded logic supporting inter-building energy exchange

  • Developing nov
  • vel

el virtual tual sub-met meterin ering technologies and event- based middleware applications that will support advanced monitoring and diagnostics concepts

ARCHITECTURE

Future re Green n Techno nolo logy y - Japan anese se-Da Danish nish Joint Worksh shop, p, 2012 26 26

slide-27
SLIDE 27

In Inno nova vation tion & & Demo monst nstrators rators

  • Technological innovations at

device level

  • Scalable processing and

inference of complex events

  • Use of cloud computing

capabilities

  • Optimal energy management

and control

  • Reduction of costs of system

development

  • Open architectures

Future re Green n Techno nolo logy y - Japan anese se-Da Danish nish Joint Worksh shop, p, 2012 2012 27 27

  • Private homes and office

buildings in the city of Aalborg, Denmark

  • A housing co-operative with 8 homes,

which recently installed 8 solar panel

  • units. The buildings have electric

heating.

  • A building with heat pump and solar

panel

  • Energy-Efficient Campus in

Terrassa, Barcelona, Spain

  • This real life campus district will allow

validation of the ENCOURAGE architecture as well as the social network.

  • Laboratory building of

Scuola Normale Superiore di Pisa, Italy

  • 4000 m2 in a recently refurbished

ancient building: energy efficiency and cost optimization, reliability and quality of energy supply

slide-28
SLIDE 28

From Ne Network twork Pro rotoco tocols ls to Power

  • wer Gri

rids ds

  • Renewable Energies are
  • n the rise
  • Photovoltaic generators
  • n roof tops
  • Good Strategies are

needed for photovoltaic microgenerators

Future re Green n Techno nolo logy y - Japan anese se-Da Danish nish Joint Worksh shop, p, 2012 2012 Kim Larse sen [28 28]

Hundreds of thousands Geographically distributed Distributed Selfstabilising Simple Fair Central

slide-29
SLIDE 29

Con

  • ntr

trol

  • l Str

trate ategie gies

Current rent Strateg egies ies (G)

  • 75% non-measured,

cannot be remotely controlled

  • EN 50438:2007, in

force since 2007

  • Switch off when

frequency > 50.2 Hz

  • VDE-AR-N4105,

required today:

  • Output linear function
  • f frequency in

[50.2,51.5]Hz

  • Emergency switch-off

above 51.5 Hz

  • Switch on again when

<50.05 for 1 minute

Inven vent new controll trollers ers

  • Be inspired by computer networks
  • Power Grid / Network
  • Up limit on generation /

limit on bandwidth

  • All generators contribute /

bandwidth is shared

  • Can measure frequency /

detect loss + collision

  • Fair distr. of capacity /

fair distr. of bandwidth

Let us try solutio tions ns from Ethern ernet, et, TCP

  • AIMD: Add Increase, Mult decr
  • Exponential Backoff
  • Freq dep probabilistic switching
  • Combinations

Future re Green n Techno nolo logy y - Japan anese se-Da Danish nish Joint Worksh shop, p, 2012 Kim Larse sen [29 29]

slide-30
SLIDE 30

Results ults

Future e Green Technolog

  • logy - Japan

anese-Dan anis ish Joint t Workshop,

  • p, 2012

2012

Kim Larse sen [30 30]

  • n-off

ff linear ar AIMD Exp Backo koff ff Freq prob switchin ching

Results ults – Mo More to to Com

  • me
slide-31
SLIDE 31

Con

  • nclusion

clusion

  • What can we offer?
  • Involvement in ARTEMIS projects (a’la ENCOURAGE) on

Smart Houses & Smart Grid

  • Infrastructure solutions (Homeport)
  • Wireless network competences
  • Control synthesis & optimization
  • Energy-aware embedded systems
  • Int. Transport Systems

Databases (spatial & temporal)

  • eHealth Tech
  • Security
  • THANKS 

Futu ture re Green een Tech chnolog

  • gy

y - Japanese ese-Da Danish sh Joint t Work rksh shop, , 2012

Kim Larse sen [31 31]