Hybrid Systems in Automotive Engine Control Andrea Balluchi - - PDF document

SLIDE 1

16

HYSCOM

IEEE CSS Technical Committee on Hybrid Systems

scimanyd suounitnoc enibmoc smetsys dirbyH lacipyt (snoitauqe ecnereffid ro laitnereffid) scimanyd etercsid dna stnalp lacisyhp fo fo lacipyt (snoitidnoc lacigol dna atamotua) fo senilpicsid gninibmoc yB .cigol lortnoc ,yroeht lortnoc dna smetsys dna ecneics retupmoc dilos a edivorp smetsys dirbyh no hcraeser ,sisylana eht rof sloot lanoitatupmoc dna yroeht fo ngised lortnoc dna ,noitacifirev ,noitalumis egral a ni desu era dna ,''smetsys deddebme`` ria ,smetsys evitomotua) snoitacilppa fo yteirav ssecorp ,smetsys lacigoloib ,tnemeganam ciffart .(srehto ynam dna ,seirtsudni

www.ist-hycon.org www.unisi.it

1 HYCON PhD School on Hybrid Systems

st

Siena, July 1 9-22, 2005 - Rectorate of the University of Siena

Hybrid Systems in Automotive Engine Control Andrea Balluchi

PARADES – Rome, Italy

balluchi@parades.rm.cnr.it

SLIDE 2

Hybrid Systems in Hybrid Systems in Automotive Engine Control Automotive Engine Control

Andrea Balluchi Andrea Balluchi

in collaboration with in collaboration with Luca Luca Benvenuti Benvenuti, Antonio , Antonio Bicchi Bicchi, , Claudio Lemma, Claudio Lemma, Emanuele Mazzi Emanuele Mazzi, , Alberto L. Sangiovanni Alberto L. Sangiovanni-

• Vincentelli

Vincentelli, , Gabriele Gabriele Serra Serra 1 1st

st HYCON PhD School on Hybrid Systems

HYCON PhD School on Hybrid Systems Siena, 19 Siena, 19-

• 22 July 2005

22 July 2005

PARADES PARADES

฀Università degli Studi di Pisa

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.2

Outline Outline

• Automotive: a promising domain for hybrid systems

Automotive: a promising domain for hybrid systems

• Model

Model-

• based design

based design

• Derivative design

Derivative design

• Design flow

Design flow

• Two automotive engine control applications of hybrid systems

Two automotive engine control applications of hybrid systems

• Actual Engaged Gear Identification: a Hybrid Observer Approach

Actual Engaged Gear Identification: a Hybrid Observer Approach

• Hybrid Modeling and Control of the Common Rail

Hybrid Modeling and Control of the Common Rail

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.3

Automotive networked control system Automotive networked control system

• Car manufacturers have to redesign their products more and more

Car manufacturers have to redesign their products more and more frequently to meet customers’ demands on innovation frequently to meet customers’ demands on innovation

• The pressure of competitiveness is even higher for control syste

The pressure of competitiveness is even higher for control system m development, since more than 80% of innovation is in electronics development, since more than 80% of innovation is in electronics

• In today cars, the electronic control system is a networked syst

In today cars, the electronic control system is a networked system em

• with more than 80 interconnected

with more than 80 interconnected ECUs ECUs, some of them safety critical , some of them safety critical

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.4

Automotive control systems architecture Automotive control systems architecture

• Today, a “one

Today, a “one-

• subsystem one

subsystem one-

• ECU” networked control system

ECU” networked control system

• This rigid partition between subsystems and electronics

This rigid partition between subsystems and electronics

• results in a higher cost of electronics

results in a higher cost of electronics

• prevents the design of efficiently integrated functionalities

prevents the design of efficiently integrated functionalities

• it is often not efficient in terms of communication and synchron

it is often not efficient in terms of communication and synchronization ization

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.5

Future scenario for automotive control systems Future scenario for automotive control systems

• The new trend is

The new trend is

• Break the “one

Break the “one-

• subsystem one

subsystem one-

• ECU” paradigm

ECU” paradigm

• Distribute functionalities over several nodes to optimize number

Distribute functionalities over several nodes to optimize number and cost of and cost of ECUs ECUs

• Advantages

Advantages

• flexibility, cost reduction, redundancy (fault

flexibility, cost reduction, redundancy (fault-

• tolerance)

tolerance)

• more sophisticated control enabled by more powerful hardware

more sophisticated control enabled by more powerful hardware

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.6

Model Model-

• based design

based design

• Model

Model-

• based design is becoming widely

based design is becoming widely used in automotive industry used in automotive industry

• algorithms are designed and analyzed using

algorithms are designed and analyzed using block diagram block diagram-

• based modeling tools

based modeling tools

• correctness of the algorithms is validated

correctness of the algorithms is validated against models of the plant against models of the plant

• models form the basis for all subsequent

models form the basis for all subsequent development stages development stages

• executable specification (instead of docs)

executable specification (instead of docs)

• automatic code generation

automatic code generation

• Advantages

Advantages

• Time

Time-

• saving and cost

saving and cost-

• effective

effective

• Design choices can be explored and

Design choices can be explored and evaluated quickly and reliably evaluated quickly and reliably

• Ideally, an optimized and fully tested system

Ideally, an optimized and fully tested system is obtained is obtained

SLIDE 3

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.7

Model Model-

• based design

based design

• However, today in the automotive industry

However, today in the automotive industry

• model

model-

• based design is often limited to control algorithm description

based design is often limited to control algorithm description

• not complete plant modeling prevents accurate validation of algo

not complete plant modeling prevents accurate validation of algorithms rithms

• Experimental validation is still extensively used, but

Experimental validation is still extensively used, but

• very expensive, time

very expensive, time-

• consuming, bounded coverage

consuming, bounded coverage

• due to the high cost, OEM will provide less support to experimen

due to the high cost, OEM will provide less support to experimentation in Tier tation in Tier-

• 1 companies

1 companies

• The partial implementation of model

The partial implementation of model-

• based design is due to

based design is due to

• insufficient investments in design process innovation

insufficient investments in design process innovation

• lack of methodologies and tools suitable to address critical ste

lack of methodologies and tools suitable to address critical steps in the design ps in the design flow, which are currently handled relying on the experience of t flow, which are currently handled relying on the experience of the designers he designers

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.8

Derivative design Derivative design

• The derivative design approach:

The derivative design approach:

• Every two

Every two-

• three years a new generation of products is designed

three years a new generation of products is designed

• Product generations are conceived to accommodate the specificati

Product generations are conceived to accommodate the specification of all

• n of all

customers for the next years customers for the next years

• For each commitment, the electronic control unit is obtained by

For each commitment, the electronic control unit is obtained by derivation derivation from the current generation from the current generation

• In the derivative design approach, reuse is extensively employed

In the derivative design approach, reuse is extensively employed to to minimize cost and development time minimize cost and development time

• for each class of applications, products are variants of a same

for each class of applications, products are variants of a same originating

• riginating

design design

PPC

OEM Technology DB ABS ACC Steer C. Brake C.

Control Algorithm HC11

Supplier Technology DB

SW 1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.9

Automotive industry design process Automotive industry design process

system system specification specification functional functional deployment deployment control control design design hw/sw hw/sw design design components components implementation implementation hw/sw hw/sw testing testing control control validation validation functional functional integration integration system system testing testing

design design integration & testing integration & testing

hybrid hybrid systems systems

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.10

Plant modeling Plant modeling -

• model development

model development

• 4

4-

• stroke internal combustion engine

stroke internal combustion engine

• 4

4-

• stroke engine cycle (FSM + DES + CT)

stroke engine cycle (FSM + DES + CT)

• in:

in: spark ignition spark ignition; ; injected fuel injected fuel; ; air charge air charge; ; EGR conc. EGR conc.; ; engine speed engine speed; ;

• ut:
• ut: engine torque, temperature

engine torque, temperature; ; dc events dc events; ; A/F A/F; ; exhaust gas exhaust gas; ;

• fuel injection (FSM + CT)

fuel injection (FSM + CT)

• inputs:

inputs: fuel injection signal fuel injection signal ( (rail pressure regulator command rail pressure regulator command

• DI);

DI);

• utputs:
• utputs: injected fuel

injected fuel ( (rail pressure rail pressure; ;fuel temperature fuel temperature -

• DI);

DI);

• spark ignition (FSM)

spark ignition (FSM)

• inputs:

inputs: ignition coil command ignition coil command; ; spark command spark command; ;

• utputs:
• utputs: spark ignition

spark ignition; ;

• air dynamics (CT)

air dynamics (CT)

• inputs:

inputs: throttle valve command throttle valve command; ; EGR command EGR command; ; VGT VGT command command; ;

• utputs:
• utputs: throttle valve angle

throttle valve angle; ; temperature; pressure temperature; pressure; ; air flow air flow rate rate; ; air charge air charge; ; EGR concentration EGR concentration; ;

CV CV DV DV CT CT DT DT Maserati Spider - V8 1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.11

Hybrid model of a 4 Hybrid model of a 4-

• stroke engine

stroke engine

cont / cont load torque load torque T TL

L

disc / disc clutch clutch clutch clutch Time / Value Time / Value Disturbances Disturbances cont / cont throttle α disc / disc ignition spark Time / Value Time / Value Controls Controls T(t) TL(t) n(t) m(t) α (t) spark clutch

Throttle Valve Angle POWER-TRAIN INTAKE MANIFOLD CYLINDERS Mass of Air Spark Ignition Engine Torque Engine Speed Load Torque Clutch Engagement

+

• intake manifold

powertrain (idle gear) 1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.12

Single cylinder FSM: engine cycle Single cylinder FSM: engine cycle

disc / cont spark advance angle ϕ disc / cont generated torque T disc / cont mass of air m Time / Value Time / Value States States

dc dc dc dc spk spk spk&dc dc

I H BS AS PA NA C E

negative spark negative spark advance advance positive spark positive spark advance advance I I →

→ BS

BS BS BS →

→ PA

PA PA PA →

→ AS

AS NA NA →

→ AS

AS AS AS →

→ H

H

t

T (t)

torque piece-wise profile spark

k - 4 k - 3 k - 2 k - 1 k k+1

a

H I BS PA AS H

fuel injection air intake

SLIDE 4

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.13

4 4-

• Cylinder Engine Hybrid Automaton

Cylinder Engine Hybrid Automaton

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.14

Control synthesis Control synthesis -

• algorithm development

algorithm development

• Characteristics of the overall electronic control system

Characteristics of the overall electronic control system

• Multi

Multi-

• rate control system composed of nested control loops that intera

rate control system composed of nested control loops that interact with ct with

• ther embedded controllers
• ther embedded controllers
• frequency and phase drifts between sampling frequencies

frequency and phase drifts between sampling frequencies

• event driven actions

event driven actions

• asynchronous communication on the network

asynchronous communication on the network

• Implements both continuous and discrete functionalities

Implements both continuous and discrete functionalities

• more discrete than continuous

more discrete than continuous

• control algorithms may have many operation modes

control algorithms may have many operation modes

nominal operation modes safety, protection and recovery modes

• computations performed at transition time are very important

computations performed at transition time are very important

switching conditions controller initializations

• A large part of algorithms devoted to diagnosis, fault tolerance

A large part of algorithms devoted to diagnosis, fault tolerance and safety and safety Complexity: more than 150 I/O and 200 algorithms in engine contr Complexity: more than 150 I/O and 200 algorithms in engine control units

• l units

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.15

Architecture of cruise control algorithms Architecture of cruise control algorithms

• Driver console and driveline systems interface

Driver console and driveline systems interface

• Cruise control supervisor

Cruise control supervisor

• Vehicle velocity regulation feedback

Vehicle velocity regulation feedback

VEHICLE Cruise Control Commands ECU Inputs: Brake, Clutch, Gear Accelerator Pedal VEHICLE VELOCITY REGULATION FEEDBACK CRUISE CONTROL SUPERVISOR DRIVER CONSOLE AND DRIVELINE SYSTEMS INTERFACE TORQUE CONTROL Commands CC Torque Feedback Driver/Cruise Velocity Set Point Cruise Control Mode Display Vehicle Velocity DRIVER Traction Force Engine Torque

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.16

Cruise control supervisor Cruise control supervisor

key_on_end KEYON LOCK OFF ON AWAIT REG_UP lamp = on; vref = 0; lamp = on; lamp = on; vref = 0; lamp = on; vref = 0; lock_ immediately lock_smoothly cruise_on cruise_off cruise_on & NOT cruise_release & NOT cruise_stop cruise_on & NOT cruise_release & NOT cruise_stop cruise_off ( set_plus l set_minus ) / vref = vehicle_speed; vehicle_speed >> vref VCONTROL TRACKING set_plus / vref = vref + v set_minus / vref = vref - v vehicle_speed << vref vehicle_speed >= vref vehicle_speed <= vref REG_DOWN ENABLED lamp = on; stand_by OVERRUN STANDBY aux_sys_req resume & NOT stand_by & NOT aux_sys_req ( set_plus l set_minus ) & NOT stand_by & NOT aux_sys_req / vref = vehicle_speed; ( set_plus l set_minus ) & NOT aux_sys_req / vref = vehicle_speed; driver_req l aux_sys_req NOT driver_req & NOT aux_sys_req REG_UP disenable_smoothly disenable_immediately

Actual Engaged Gear Identification: Actual Engaged Gear Identification: a Hybrid Observer Approach a Hybrid Observer Approach

Andrea Balluchi Andrea Balluchi(

(1,2 1,2) ),

, Luca Luca Benvenuti Benvenuti(

(1,3 1,3) ),

, Claudio Lemma Claudio Lemma(

(1 1) )

Alberto L. Sangiovanni Alberto L. Sangiovanni-

• Vincentelli

Vincentelli(

(1,4 1,4) ) ,

, Gabriele Gabriele Serra Serra(

(5 5) )

(1) (1) PARADES, Rome, I

PARADES, Rome, I

(2) (2) Centro

Centro Interdip

• Interdip. “E.

. “E. Piaggio Piaggio”, University of Pisa, Pisa, I ”, University of Pisa, Pisa, I

(3) (3) Dip.

• Dip. di Informatica

di Informatica e e Sistemistica Sistemistica, University of Roma “La , University of Roma “La Sapienza Sapienza”, Rome, I ”, Rome, I

(4) (4) Dept. of EECS, University of California at Berkeley, CA

• Dept. of EECS, University of California at Berkeley, CA

(5) (5) Magneti

Magneti Marelli Marelli Powertrain Powertrain, Bologna, I , Bologna, I 16 16th

th IFAC World Congress

IFAC World Congress Prague, July 4 Prague, July 4-

• 8, 2005

8, 2005

CC “Control and Computation” CC “Control and Computation” Applications Paper Prize 1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.18

Motivation Motivation

• Actual engaged gear identification is relevant to engine control

Actual engaged gear identification is relevant to engine control for for cars equipped with manual gear cars equipped with manual gear

• The gear and clutch states are used in

The gear and clutch states are used in

• Engine torque control

Engine torque control

• to improve drivability by compensating the equivalent inertia of

to improve drivability by compensating the equivalent inertia of the vehicle on the the vehicle on the crankshaft crankshaft

• to prevent engine stall by acting promptly when the transmission

to prevent engine stall by acting promptly when the transmission is opened is opened

• Tailpipe emissions control

Tailpipe emissions control

• particulate emissions for Diesel engines are particularly critic

particulate emissions for Diesel engines are particularly critical to control with first al to control with first gear engaged gear engaged

SLIDE 5

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.19

Outline Outline

• Automotive driveline modeling

Automotive driveline modeling

• Detailed hybrid model used for analysis and validation

Detailed hybrid model used for analysis and validation

• discontinuities and nonlinear dynamics

discontinuities and nonlinear dynamics

• Simplified hybrid model used for synthesis

Simplified hybrid model used for synthesis

• btained by abstraction and reduction
• btained by abstraction and reduction
• Hybrid design of the actual engaged gear identification algorith

Hybrid design of the actual engaged gear identification algorithm m

• Validation

Validation

• Against the detailed hybrid nonlinear model

Against the detailed hybrid nonlinear model

• robustness analysis

robustness analysis

• Using experimental data provided by

Using experimental data provided by Magneti Marelli Powertrain Magneti Marelli Powertrain

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.20

Automotive driveline Automotive driveline

1) engine suspension 2) cylinder block 3) crankshaft 4) connecting rod 5) piston 6) clutch 7) gear 8) differential 9) semi-axle 10) flexible joint 11) gear box 12) hub 13) body suspension 14) wheel 15) tire

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.21

Driveline detailed hybrid model Driveline detailed hybrid model

6048 discrete states 12 continuous states

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.22

Validation of the model with experimental data Validation of the model with experimental data

proposed model experimental data

engine clutch

shunt shuffle

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.23

Driveline simplified hybrid model Driveline simplified hybrid model

7 discrete states

• gear and clutch

4 continuous states ωe - crankshaft speed ωc - clutch speed ωw - wheel speed α

• torsion angle

1 discrete input

• gear lever ∈{1, 2,..5, RG, N }

3 continuous inputs Te - engine torque (m.v. known) Pc - clutch plate pressure Tw - wheel torque 2 continuous outputs ωe - crankshaft speed ωw - wheel speed with gear engaged and clutch closed:

crankshaft clutch plates gear road

ωe ωc ωw Te Pc Tw α

vehicle inertia

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.24

Abstraction of the discrete behavior Abstraction of the discrete behavior

clutch closed and 1st gear

OPEN SLIPPING CLOSED

idle gear or backlash or clutch open or clutch slipping

SLIDE 6

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.25

Wheel speed and engine speed comparison Wheel speed and engine speed comparison

• Limitations

Limitations

• Large time delays in the engaged gear identification

Large time delays in the engaged gear identification

• due to oscillations of the transmission shafts during transients

due to oscillations of the transmission shafts during transients

• particularly critical for idle speed and first gear identificati

particularly critical for idle speed and first gear identification

• n
• Frequent identification errors

Frequent identification errors

• Specification

Specification

• Identification within a delay of 250

Identification within a delay of 250 msec msec., sampling period of 12 ., sampling period of 12 msec msec. .

crankshaft clutch plates gear road

ωe ωc ωw Te Pc Tw α

vehicle inertia

ωw ωe

1st gear 2nd gear 3rd gear

τ3 τ2 τ1 1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.26

Hybrid observer approach for actual engaged gear Hybrid observer approach for actual engaged gear identification identification

Driveline Hybrid Driveline Hybrid Model Model

q q( (k k), ), x x( (t t) )

Hybrid Observer Hybrid Observer Continuous Continuous Observer Observer

q(k)

~

x(t)

~ Location Location Observer Observer

Discrete input σ

σ ( (k k) )

Discrete output ψ

ψ ( (k k) )

Continuous input u

u( (t t) )

Continuous output y

y( (t t) )

inputs

• gear lever
• clutch pedal
• engine torque
• wheel torque
• utputs
• crankshaft speed
• wheel speed

X

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.27

Location observer Location observer

Location Observer Location Observer FSM FSM Observer Observer

ψ ψ( (k k) ) q(k)

~ Location Location Identification Identification Logic Logic

• DES

DES Observer Observer Residual Residual Generator Generator . . . Decision Decision Function Function . . .

r1 rM r1(t) rM(t)

~ ~

Discrete input σ

σ ( (k k) )

Discrete output ψ

ψ ( (k k) )

Continuous input u

u( (t t) )

Continuous output y

y( (t t) ) u u( (t t) ) y y( (t t) )

X

Driveline Hybrid Driveline Hybrid Model Model

q q( (k k), ), x x( (t t) )

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.28

Continuous observer Continuous observer

• A switched

A switched Luenberger Luenberger observer with resets and switching controlled by the

• bserver with resets and switching controlled by the

identified plant location identified plant location

• Stability can be achieved by using dwell time approach

Stability can be achieved by using dwell time approach

• a possible delay in location identification has to be taken into

a possible delay in location identification has to be taken into account account

q q3

3

q q2

2

q q1

1

q=q2 /

~

x=(A3-G3C3)x+B3u+G3y

~ ~ .

x:=R1

13x+R0 13

~ ~

x=(A1-G1C1)x+B1u+G1y

~ ~ .

x:=R1

32x+R0 32

~ ~

x=(A2-G2C2)x+B2u+G2y

~ ~ .

q=q3 /

~

x:=R1

21x+R0 21

~ ~

q=q1 /

~

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.29

Engaged gear identification algorithm Engaged gear identification algorithm

• The location

The location q qN

N cannot be detected using a residual generator due to lack of fee

cannot be detected using a residual generator due to lack of feedbacks dbacks

• The corresponding signature is obtained by negation of the other

The corresponding signature is obtained by negation of the others s

Location Identification Logic

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.30

Residual generator and decision function Residual generator and decision function

• Residual generators

Residual generators

• Luenberger obs

Luenberger obs. .

• Unknown input

Unknown input obs

• bs.

.

• Kalman

Kalman filter filter

• Walkcott

Walkcott-

• Zak obs

Zak obs. .

• Sliding mode

Sliding mode obs

• bs.

.

• Decision function

Decision function

• Passive

Passive hysteresis hysteresis relay relay

• Debouncing

Debouncing algorithm algorithm

EXIT threshold EXIT threshold ENTER threshold ENTER threshold

• Thresholds function of engine torque

Thresholds function of engine torque

• Decision function output disenabled

Decision function output disenabled during residual transients during residual transients

SLIDE 7

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.31

Validation against the detailed hybrid model of the Validation against the detailed hybrid model of the driveline driveline

• Unknown inputs

Unknown inputs

• wheel torque (road slope)

wheel torque (road slope)

• clutch plate pressure

clutch plate pressure

• engine torque pulsation

engine torque pulsation

• gear lever

gear lever

• Sensors

Sensors

• engine speed quantization (1 RPM)

engine speed quantization (1 RPM)

• vehicle speed quantization (1 Km/h)

vehicle speed quantization (1 Km/h) and dead zone (5 Km/h) and dead zone (5 Km/h)

• Unmodel

Unmodel dynamics dynamics

• both discrete and continuous

both discrete and continuous

driveline driveline hybrid model hybrid model q q( (k k), ), x x( (t t) )

engine torque pulsation unknown unknown gear lever unknown quantization quantization dead zone

ωe ω w Tw Pc Te

unknown

engaged gear engaged gear hybrid observer hybrid observer

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.32

Validation against the detailed hybrid model of the Validation against the detailed hybrid model of the driveline driveline

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.33

Experimental results Experimental results

• Obtained in

Obtained in Magneti Marelli Powertrain Magneti Marelli Powertrain using an using an Opel Astra Opel Astra equipped with equipped with

• a Diesel engine and a robotized gearbox

a Diesel engine and a robotized gearbox SeleSpeed SeleSpeed

• The estimated engaged gear compared to the signal from the gearb

The estimated engaged gear compared to the signal from the gearbox control unit

• x control unit
• The proposed algorithm was tested on several maneuvers of differ

The proposed algorithm was tested on several maneuvers of different types, for a ent types, for a total of 250 gear engagements total of 250 gear engagements

• The actual engaged gear was successfully identified within a del

The actual engaged gear was successfully identified within a delay of 250 ay of 250 msec msec. . in 90% of cases in 90% of cases

• The unsuccessful cases have been obtained in very critical maneu

The unsuccessful cases have been obtained in very critical maneuvers such as vers such as

• gear engagements during sharp braking

gear engagements during sharp braking

• clutch abrupt releases

clutch abrupt releases

• In these cases, the residuals exhibit large oscillations

In these cases, the residuals exhibit large oscillations

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.34

Experimental results Experimental results

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.35

Conclusions Conclusions

• A detailed hybrid model of the driveline has been developed

A detailed hybrid model of the driveline has been developed

• The model has been analyzed to obtained a reduced model used for

The model has been analyzed to obtained a reduced model used for synthesis synthesis

• An algorithm for actual engaged gear identification based on hyb

An algorithm for actual engaged gear identification based on hybrid observer rid observer theory has been devised theory has been devised

• The proposed algorithm exhibits remarkable robustness with respe

The proposed algorithm exhibits remarkable robustness with respect to ct to unmodel unmodel dynamics, disturbances and uncertain parameters dynamics, disturbances and uncertain parameters

• The proposed algorithm has been validated by both

The proposed algorithm has been validated by both

• extensive simulation with the detailed hybrid model of the drive

extensive simulation with the detailed hybrid model of the driveline line

• experimental data obtained with an

experimental data obtained with an Opel Opel Astra Astra equipped with equipped with SeleSpeed SeleSpeed

• Efficient drivability control allows car manufactures to design

Efficient drivability control allows car manufactures to design lighter transmission lighter transmission systems characterized by higher elasticity, which will require t systems characterized by higher elasticity, which will require the use of he use of dynamical algorithms for actual engaged gear identification dynamical algorithms for actual engaged gear identification

Hybrid Modeling and Control Hybrid Modeling and Control

• f the Common Rail
• f the Common Rail

Andrea Balluchi Andrea Balluchi(

(1 1) ),

, Antonio Antonio Bicchi Bicchi(

(2 2) ),

, Emanuele Mazzi Emanuele Mazzi(

(2,4 2,4) )

Alberto L. Alberto L. Sangiovanni Sangiovanni-

• Vincentelli

Vincentelli(

(1,3 1,3) ) ,

, Gabriele Gabriele Serra Serra(

(4 4) )

(1) (1) PARADES GEIE, Rome, I

PARADES GEIE, Rome, I

(2) (2) Centro

Centro Interdip

• Interdip. “E.

. “E. Piaggio Piaggio”, University of Pisa, Pisa, I ”, University of Pisa, Pisa, I

(3) (3) Dept. of EECS., University of California at Berkeley, CA

• Dept. of EECS., University of California at Berkeley, CA

(4) (4) Magneti

Magneti Marelli Marelli Powertrain Powertrain, Bologna, I , Bologna, I

First HYCON Workshop on First HYCON Workshop on Automotive Applications of Hybrid Systems Automotive Applications of Hybrid Systems Rome, 26 Rome, 26-

• 27 May 2004

27 May 2004

SLIDE 8

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.37

Outline Outline

• Common rail injection system

Common rail injection system

• Detailed hybrid model of the fuel injection system

Detailed hybrid model of the fuel injection system

• Rail pressure controller design

Rail pressure controller design

• Simulation results

Simulation results

• Conclusions and future work

Conclusions and future work

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.38

New common rail injection system developed by New common rail injection system developed by Magneti Magneti Marelli Marelli Powertrain Powertrain

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.39

Hybrid model of the common rail injection system Hybrid model of the common rail injection system

Pump Pump DRV DRV

mduty Vbatt

mProp mProp Injectors Injectors

Qpump Qinj Qleak Qinj-SERV QDRV Tcoil

Prail

Rail Rail

enginespeed Tfuel QmProp

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.40

High pressure pump regulation valve ( High pressure pump regulation valve (mProp mProp) )

mduty Vbatt

PWM

0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.2 0.4 0.6 0.8 1 time (sec) 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 5 10 15 time (sec) tensione (V) sequenza ripetuta mduty V

( )

L V I L T R I

mprop coil mprop

+ − = &

V

Tcoil

Flow rate - Current

Tfuel

QmProp 50 100 150 200 250 20 40 60 80 100 time (sec) Tcoil (°C) Improp

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.41

High pressure pump High pressure pump

Intake Compression

QmProp

∫

Pump angle Amout of flow rate entering each raw

x

Geometric volume Fuel charge 1 = compression Flow rate Compressed volume Fuel charge Qpump

+

Qpiston_2 Qpiston_3 Qpiston_1 Pump angle deg mm3 mm3 / sec mm3

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.42

Injectors Injectors

) , , ( ) , ( ) , , ( giri ET P Q T P Q giri ET P Q Q

rail SERV inj fuel rail leak rail inj

• ut

−

+ + =

Prail Tfuel ET giri Qout 200 400 600 800 1000 1200 1400 −40 −20 20 40 60 80 100 200 400 600 800 1000 1200 ←−30 ←−10 ←10 ←20 ←40 ←60 ←80 ←100 Prail (bar) Tfuel (°C) QLeak (mm3/sec) engine angle (deg) Injector command

SLIDE 9

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.43

Common rail Common rail

( ) ( )

) ( ), ( ) ( ) ( ) ( ) ( ) ( 1 ) ( t T t P K V V t C t Q t Q t Q t C t P

fuel rail bulk pipe rail rail DRV

• ut

pompa rail rail

+ = − − = & Fuel comprimibility modelled using Bulk parameter

Prail(0) Tfuel Qpump Qout QDRV Prail(t)

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.44

Controller design Controller design

• Common rail pressure control

Common rail pressure control

• Design a tracking controller for the rail pressure that achieves

Design a tracking controller for the rail pressure that achieves tracking of a tracking of a reference pressure signal generated on reference pressure signal generated on-

• line by an outer loop controller

line by an outer loop controller

• Main challenges

Main challenges

• Time

Time-

• varying delay between valve control command and rail pressure

varying delay between valve control command and rail pressure measurement measurement

• Time

Time-

• varying uncertainty of the actuator electrical resistance

varying uncertainty of the actuator electrical resistance

• Design approach

Design approach

• Smith predictor controller

Smith predictor controller

• loop delay compensation

loop delay compensation

• Adaptive algorithm

Adaptive algorithm

• n
• n-
• line identification of the electrical resistance

line identification of the electrical resistance

• First attempt: controller synthesis based on a plant mean

First attempt: controller synthesis based on a plant mean-

• value model

value model

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.45

Mean Mean-

• value model for controller synthesis

value model for controller synthesis

• nonlinear CT model

nonlinear CT model

• piecewise polynomial CT model

piecewise polynomial CT model

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.46

Rail pressure controller Rail pressure controller

• Sampling time 5msec

Sampling time 5msec

+

• Feed Forward

+ +

• Injection system

d

sT

e−

Smith predictor

Delay-free model

d

T s

e

ˆ − PID anti-windup

mduty Prail +

Delay-free Plant

• +

Prailref

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.47

Hybrid closed Hybrid closed-

• loop system simulation results

loop system simulation results

Smith controlled pressure Reference pressure

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.48

Smith Predictor and adaptive control Smith Predictor and adaptive control

Adaptive Algorithm Improp-est 1° order Sliding mode observer Rest current estimator +

• Feed Forward

+ +

• Smith predictor

Delay-free model PID anti-windup Prailref mduty Prail +

Common Rail Hybrid Model

• +

d

sT

e−

Low Pass Filter

SLIDE 10

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.49

Closed Closed-

• loop hybrid model simulation results

loop hybrid model simulation results

Smith controlled pressure Reference pressure Magneti Marelli controlled pressure Smith controlled pressure Reference pressure

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.50

Limits of the controller designed using the mean Limits of the controller designed using the mean-

• value model approach

value model approach

3 4 5 6 7 8 9 10 200 250 300 350 400 450 time (sec) Prail (bar)

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.51

Hybrid multi Hybrid multi-

• rate controller

rate controller

QHP(t) p(t)

COMMON RAIL FLOW RATE VALVE HP PUMP valve fuel flow rate HP fuel flow rate CM pressure injectors fuel flow rate desired HP fuel flow rate flow rate valve command FLOW RATE VALVE CONTROLLER

m(l)

flow rate feedback

+

• QINJ(t)

+ pref(l)

pressure reference

• pressure

sensor

~

QHP(k)

CM PRESSURE CONTROLLER

QFRV(t) perr(k)

pressure error valve actuator

5ms fuel delivery 5ms

interpolator decimator

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.52

Tracking behavior of the hybrid multi Tracking behavior of the hybrid multi-

• rate

rate controller controller

2 3 4 5 6 7 8 200 220 240 260 280 300 320 340 360 380 time (sec) Prail (bar) Prail PrailREF

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.53

Conclusions Conclusions

• A detailed hybrid model of the common rail fuel injection system

A detailed hybrid model of the common rail fuel injection system has been has been presented presented

• The hybrid model describes the pulsating evolution of the rail p

The hybrid model describes the pulsating evolution of the rail pressure due to HP ressure due to HP pump supply and multiple fuel injections pump supply and multiple fuel injections

• The proposed switching controller has been designed using a mean

The proposed switching controller has been designed using a mean-

• value model

value model

• f the plant and employs
• f the plant and employs
• a Smith Predictor to compensate the time

a Smith Predictor to compensate the time-

• varying loop delay

varying loop delay

• an adaptive algorithm to adjust the static gain

an adaptive algorithm to adjust the static gain

• Simulation results obtained with the hybrid closed

Simulation results obtained with the hybrid closed-

• loop model show that the

loop model show that the controller perform satisfactorily if the reference pressure is n controller perform satisfactorily if the reference pressure is not too fast

• t too fast
• Controller design based on hybrid methodologies achieves better

Controller design based on hybrid methodologies achieves better performances performances and ensures tracking of fast pressure references and ensures tracking of fast pressure references

1st HYCON PhD School on Hybrid Systems 1st HYCON PhD School on Hybrid Systems

• A. Balluchi
• A. Balluchi -
• PARADES Siena, 19

PARADES Siena, 19-

• 22 July 2005 p.

22 July 2005 p.54

Bibliography Bibliography

E. E.-

• H.
• H. Azibi

Azibi, J. , J.-

• C.
• C. Sardas

Sardas. . Computer Computer-

• aided process engineering and transformation of the process desi

aided process engineering and transformation of the process design gn activity in automotive industry activity in automotive industry. In Proc. 2002 IEEE Inter. Conf. on Sys.s, Man and . In Proc. 2002 IEEE Inter. Conf. on Sys.s, Man and Cyb Cyb.s, 2002. .s, 2002. A.

• A. Agostini

Agostini, A. Balluchi, A. , A. Balluchi, A. Bicchi Bicchi, B. , B. Piccoli Piccoli, A. L. , A. L. Sangiovanni Sangiovanni-

• Vincentelli

Vincentelli, K. , K. Zadarnowska Zadarnowska. . Randomized Randomized algorithms for platform algorithms for platform– –based design based design. In Proc. 44th IEEE Conf. on Dec. and Cont., CDC2005, 2005. . In Proc. 44th IEEE Conf. on Dec. and Cont., CDC2005, 2005. M.

• M. Antoniotti

Antoniotti, A. Balluchi, L. , A. Balluchi, L. Benvenuti Benvenuti, A. Ferrari, C. , A. Ferrari, C. Pinello Pinello, A. L. , A. L. Sangiovanni Sangiovanni-

• Vincentelli

Vincentelli, R. Flora, W. , R. Flora, W. Nesci Nesci, C. Rossi, G. , C. Rossi, G. Serra Serra, M. , M. Tabaro Tabaro. . A top A top-

• down constraints

down constraints-

• driven design methodology for

driven design methodology for powertrain powertrain control system control system. In Proc. GPC98, Global . In Proc. GPC98, Global Powertrain Powertrain Congress, Detroit, MI, October 1998. Congress, Detroit, MI, October 1998. H.

• H. Heinecke

Heinecke, K. , K.-

• P.
• P. Schnelle

Schnelle, H. Fennel, J. , H. Fennel, J. Bortolazzi Bortolazzi, L. , L. Lundh Lundh, J. , J. Leflour Leflour, J. , J.-

• L. Mate’, K. Nishikawa, T.
• L. Mate’, K. Nishikawa, T.

Scharnhorst Scharnhorst. . Automotive open system architecture Automotive open system architecture -

• an industry

an industry-

• wide initiative to manage the

wide initiative to manage the complexity of emerging automotive e/e complexity of emerging automotive e/e-

• architectures
• architectures. In Proc. Convergence 2004, n. 2004

. In Proc. Convergence 2004, n. 2004-

• 21

21-

• 0042,

0042, Detroit, MI, 2004. AUTOSAR. www. Detroit, MI, 2004. AUTOSAR. www.autosar autosar.org. .org. B.S. Heck, L.M. Wills, G.J. B.S. Heck, L.M. Wills, G.J. Vachtsevanos Vachtsevanos. . Software technology for implementing reusable, distributed Software technology for implementing reusable, distributed control systems control systems. IEEE Control Systems Magazine, pages 21 . IEEE Control Systems Magazine, pages 21– –35, February 2003. 35, February 2003.

• A. Balluchi, L.
• A. Balluchi, L. Benvenuti

Benvenuti, M. D. , M. D. Di Benedetto Di Benedetto, C. , C. Pinello Pinello, A. L. , A. L. Sangiovanni Sangiovanni-

• Vincentelli

Vincentelli. . Automotive engine Automotive engine control and hybrid systems: Challenges and opportunities control and hybrid systems: Challenges and opportunities. Proc. of the IEEE, 88, (7):888 . Proc. of the IEEE, 88, (7):888– –912, 2000. 912, 2000.

• A. Balluchi, L.
• A. Balluchi, L. Benvenuti

Benvenuti, M. D. , M. D. Di Benedetto Di Benedetto, A. L. , A. L. Sangiovanni Sangiovanni-

• Vincentelli

Vincentelli. . Design of observers for hybrid Design of observers for hybrid systems

• systems. In Hybrid Systems: Computation and Control, Proc. HSCC2002, v.

. In Hybrid Systems: Computation and Control, Proc. HSCC2002, v. 2289 of LNCS, 76 2289 of LNCS, 76– –89, 2002. 89, 2002.

• A. Balluchi, L.
• A. Balluchi, L. Benvenuti

Benvenuti, C. Lemma, A. L. , C. Lemma, A. L. Sangiovanni Sangiovanni-

• Vincentelli

Vincentelli, G. , G. Serra Serra. . Actual engaged gear Actual engaged gear identification identification: a hybrid observer approach : a hybrid observer approach. In Proc. 16th IFAC World Congress, Prague (CZ), July 2005. . In Proc. 16th IFAC World Congress, Prague (CZ), July 2005.