Modelingg Of

Mechatronic Systems
NIZAR CHATTI

E-MAIL : [email protected]

OFFICE E38, ISTIA, UNIVERSITY OF ANGERS, FRANCE

TEL: +33(0)244687573
+33(
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

OUTLINE

1 ‐ Introduction
2 ‐ Bond graph methodology
3 ‐ Bond Graph model designing procedures
4 ‐ Causality
5 ‐ Mathematical models deduced from a Bond Graph
6 ‐ Conclusion

N. CHATTI 2
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

MECHATRONIC SYSTEM

• Controlled multiphysics system with information and power

transfers
Control supervision
Control‐supervision

references computer sensors

Actuating chain
Power
Power Sources Sources
actuators System

Information bond Power bond

N. CHATTI 3
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

DESIGNING NEW TECHNOLOGICAL

ENGINEERING SYSTEMS :
A MORE AND MORE COMPLEX TASK
Difficulties linked to the system: Teams with specialists
 Multiphysics, multienergy of different physical domains
(mechanical, hydraulic, electrical, … domains)
Mixing power and information « Mechatronic approach »
 Constraints :
E
Energy consumption,
ti SSafety
f t andd reliability
li bilit « Integrated
I t t d design
d i »

Difficulties
Diffi lti lilinked
k d tto the
th iindustrial
d t i l context:
t t
Economical constraints:
Costs for research, production Study on virtual
 Industrial strategy : prototypes first
To be the first
To guarantee the « just necessary » quality
asked for by the market

N. CHATTI 4
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

INTEGRATED DESIGN

Modeling Control

Simulation Analysis for control

Analysis
l i Control designing
Model simplification Actuator sizing

Supervision

Fault Detection
and Isolation

Virtual Prototype
Physical system
Virtual test bench

N. CHATTI 5
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion
The virtual prototyping desk

Simulation Software
CAD (geometric) (CATIA, Solidworks…)
Meshing/3D solid solver (IDEAS, NASTRAN, ABAQUS, LMS)
M hi / 3D fluid
Meshing fl id solver
l (GAMBIT FLUENT,
(GAMBIT, FLUENT Star
St CD,
CD …))
Multibody/ rigid‐elastic (ADAMS, LMS …)
3D Electromagnetics (Flux 2D, Flux 3D, ….)

1D power transmission (AMESim, SimX, Easy5, Saber, …)

Control design (Matlab/Simulink, …)

D i process scenario
Design i
Optimization
Data analysis (Optimus, OptiY, Mode frontier …)
Co‐simulation / real time ((Cosimate, …))
ANSYS, Comsol: multidomain distributed parameters simulation

Wide offer of dedicated simulation tools Lower offer for multi‐domain

multi‐domain, multi‐scale
covering most of the need for domain simulation for system engineers
specialists

N. CHATTI 6
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

SKILLS AND OBJECTIVES

• Systematic approach for global analysis of complex multiphysical
systems
• Finding innovative solutions
• Transversal skills on dynamic modeling of Engineering systems
independently of their physical nature
• Deduction in systematic way state equations and their simulation of
industrial systems
• Training with new software’s tools for integrated design and
simulation of industrial systems
• Managing of multidsciplinary teams

Keywords: Mechatronics,
Mechatronics Integrated design
design, Bond Graphs
Graphs, Simulation
Simulation, Dynamic Modeling
Modeling,
Automatic control

N. CHATTI 7
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

MOTIVATIONS
 Complexity of systems are due of coupling of multi energies
(mechanical, electrical, thermal, hydraulic…).
 For example: Power station

 Why dynamic modeling?

 Design,
Design Analysis,
Analysis Decision,
Decision Control,
Control Diagnosis…
Diagnosis
 Which skills for this task
 Multidisciplinary project management
 Which kind of tool is needed?
 Structured, unified, generic

N. CHATTI 8
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

WHAT IS MECHATRONIC SYSTEM

 Mechatronics («Meca» + «Tronics»)

 Engineering systems putting in evidence multiple skills

 Mechanics: Hydraulics, thermal engineering, Mechnism, pneumatic…

 Electronics: power electronics, Networks, converters, AN/NA, Micro controllers...
 A t
Automatic
ti control:
t l Linear
Li andd nonlinear
li control,
t l Advanced
Ad d control,
t l Stability…
St bilit
 Computer engineering: Real‐Time implementation

 Why
Wh Mechatronics?
M h t i ?

 Integrating
g g harmoniouslyy those technologies,
g , mechatronics enables to design
g new
and innovative industrial products simpler, more economical, reliable and versatile
(flexible) systems.

N. CHATTI 9
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

MECHATRONICS: SYNERGETIC EFFECTS

Information
Electronics
Power electronics technology
Networks
Converters AN/NA MECHATRONICS System theory
Automatic control
Micro controllers Computer engineering
Actuators Diagnosis
Sensors Artificial intelligence
Software

Mechanics
Hydraulics
Thermal engineering
Mechanism
Pneumatic
Mechanical elements
Precision mechanics

N. CHATTI 10
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLES OF MECHATRONIC SYSTEMS

Examples of Mechatronic systems include:

 Remotely controlled vehicles such as the Mars Rover

 A rover is a space exploration vehicle designed to move across the surface of
a planet or other astronomical body.

o Control of take‐off and up to exploration of Mars planet

o Remote control
o Embedded supervision, network communication
o Virtual simulation

 Automation systems:
 Vehicle stability control
 Precision control of robots
 D i off hybrid
Design h b id vehicle
hi l
 …

N. CHATTI 11
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

PEM Fuel Cell

for electrical vehicle

Different physical domains

( P, n ou m ) : Hydraulic
(T , H ou Q ) : Thermal
(  , n ) : Chemical
( E, i) : Electrical
l l

N. CHATTI 12
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

PEM Fuel Cell

for electrical vehicle

e‐ e‐
Supply
H2 Tank Load manifold

Cooler &
Active humidifier
H idifi
Humidifier e‐ zone e‐
H2 Compressor
H+ O2 +
O2 Motor
H2 H+

ode
(CA)
Catho
Anode

Membrane
(AN)

H+

H2O

Water
Heat Heat

Return
manifold

N. CHATTI 13
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

MECHATRONIC SYSTEMS
 Before 1950
 Complex systems are studied as electromechanical sub systems
 Around 1950
 Emergence of semi conductors, electronic control and power
electronics.
l t i
 1960‐1970
 Design of microcontrollers because of appearance of computer
engineering Possibility to design embedded control systems more
engineering.
efficient.
 1970‐
 « Mechatronics » was first introduced in Japan Yaskawa Electric
Corporation

Definition given by Rolf Isermann:

The new integrated systems changed from electro‐mechanical systems with

discrete electrical and mechanical parts to integrated electronic‐mechanical
systems with sensors, actuators and digital microelectronics.

N. CHATTI 14
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

• Examples of mechatronic components in a car:

• Engine Control : injection, turbocompressor,…

• Suspension Control : piloted damper, hydractive suspension, …
• Braking Control : ABS, ESP,…
• Steeringg Control
• Gear box Control : automatic, robotized,…

N. CHATTI 15
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

METHODOLOGY FOR TESTING

Development of generic models and control
algorithms

Validation using MiL

Test

Validation using SiL

Test

Validation using HiL

Test

I d
Industrial
i l validation
lid i

N. CHATTI 16
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

TESTS IN MECHATRONIC SYSTEMS

 Tests can be executed using:

• Dynamic models (Model‐in‐the‐Loop, MiL),

• Existing function (Software‐in‐the‐loop, SiL),
• A real industrial computer (Hardware‐in‐the‐loop, HiL)

 MiL (Model in the Loop)

• Test object: model

• Input signals are simulated
• Output signal values are saved to be compared to the expected values
• A t
Automatic
ti test
t t execution
ti through:
th h

 The development environment used for modeling

 Specific software’s
software s (MATLAB/Simulink)
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Tests in mechatronic systems

 SiL (Software in the Loop)
 Test object: generated code
 Environment is simulated
 The inputs and outputs of the test object are connected to the test system
 The generated code is executed on a PC or on an evaluation board
 Automatic test execution through:

o Use of MATLAB/Simulink with Real‐time workshop

o Interfaces
I t f tto external
t l ttools
l

 HiL (Hardware in the Loop)

 TTestt object:
bj t reall ECU
 Environment simulation through environment models (e.g.: MATLAB/Simulink)
 Inputs and Outputs are connected to the HiL‐Simulator
 Comparison of the ECU output values to the expected values
 Automatic test execution through the control software of the HiL‐Simulator

N. CHATTI 18
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Tests in mechatronic systems

N. CHATTI 19
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BOND GRAPHS: TOOL FOR INTEGRATED DESIGN

 Bond Graphs:

 Bond Graph is a unified graphical language used for any kind of physical
d
domain.
i The
Th tool
t l is
i confirmed
fi d as a structured
t t d approach h for
f modeling
d li and d
simulation of multidisciplinary systems.

 Bond Graphs for modelling and more

more…

 Because of its architecture representation, causal and structural

properties,
i Bond
B d graph
h modeling
d li isi used
d not only
l for
f modeling
d li but
b for:
f

 Control analysis, diagnosis, supervision, alarm filtering

 Automatic
A t ti generation
ti off d
dynamici modeling
d li and d supervision
ii
algorithms
 Sizing
 Used today by industrial companies (PSA,
(PSA Renault,
Renault IFP,
IFP CEA,
CEA
Airbus…)
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Level of modeling

N. CHATTI 21
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Bond Graph for integrated supervision design

N. CHATTI 22
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Dedicated software

N. CHATTI 23
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Graphical User Interface

N. CHATTI 24
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Graphical User Interface

N. CHATTI 25
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Simulation Interface

N. CHATTI 26
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

History of Bond Graph Modeling

 Founder of BG: Henry Paynter (MIT Boston)
 Th
The Bond
B d Graph
G h tooll was first
fi developed
d l d since
i 1961 at MIT,
MIT BOSTON,
BOSTON
USA by Paynter (April 24, 1959).
 Symbolism and rules development:
 Karnopp (university of California),
California) Rosenberg (Michigan university),
university) Jean
Thoma (Waterloo)

 Introduced in Europe only since 1971

 Netherland and France (Alstom)

 Teaching in Europe, USA…

 France: Univ
Univ. Lyon1,
Lyon1 INSA LYON,
LYON EC Lille,
Lille ESE Rennes,
Rennes Univ.
Univ Mulhouse,
Mulhouse
Polytech’Lille
 LNMIIT (Jaipur, INDIA), University of London, University of Enshede (The
Netherlands)
 Companies using this tool
 Automobile company: PSA, Renault
 Nuclear company: EDF, CEA, Alstom
 Electronic: Thomson, Aerospace
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

PEM Fuel Cell

for electrical vehicle
1
O2  2 H   2e   H 2O Reduction
e‐
2
e‐
Supply
H2 Tank Load manifold

Cooler &
Active humidifier
Humidifier e‐ zone e‐
H2 Compressor
H+ O2 +
O2 Motor
H2 H+

Cathode
(CA)
Anode

Membrane
(AN)

H+

H2O
1
H 2  O2  H 2O
H 2  2 H   2e : Oxidation
Water
Heat Heat 2
Return
manifold

N. CHATTI 28
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

PEM Fuel Cell

for electrical vehicle

 The model is complex and the numerical values are not

available

 The heart of FC is poorly instrumented.

 This is whyy structural model ((based on existence or not of

the links between variables and the relations) for
diagnosability analysis (which fault can be detected and
isolated with given instrumentation architecture) is well
suited

N. CHATTI 29
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Application : Ekofisk oil rigs (1/7)

6 rigs (1 hostel rig for 280 workers)

Oil and ggas p
production started in 1970 ‐ Phillips
p Petroleum company
p y ((Norway)
y)

N. CHATTI 30
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Application : Ekofisk oil rigs (2/7)

In 1984, subsidence phenomena detected

Measurements by satellites show a 4 m driving in of the set of rigs, 35
cm more per year

risks of huge sea waves (25 m height possible in winter) and wind (200 km/h)

 safety no more ensured

Norwegian government imposes to the company to act to protect the installations

and the workers

 Different solutions proposed:

‐ build a huge dyke around the set : not possible (70m depth of the sea)
‐ cast old tankers as breakwater : too difficult to maintain
‐ raise the set of rigs using jacks : chosen solution

N. CHATTI 31
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Application : Ekofisk oil rigs (4/7)

Implementation during August 1985, still in operation

1. Positioning of
jacks around 2. The jacks bear the 3. Fixing of the
the rig leg and cutting rig during the rising extension bars
off the
h leg
l of the rig of 6.5m
6 5m

before after

N. CHATTI 32
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Application : Ekofisk oil rigs (5/7)

N. CHATTI 33
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Application : Ekofisk oil rigs (6/7)

N. CHATTI 34
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Application
pp : Ekofisk oil rigs
g ((7/7)
/ )

• No means to do in an other way than using dynamic modeling

and simulation
• Scale factor too huge to use reduce scale hard models

• Impossible to fix the controller gains using check‐and‐error procedure,

because of the veryy hard constraints

• BG methodology helpful because of the ability of coupling 3D (flexible

structure) and 1D models

N. CHATTI 35
Introduction BG elements Design procedures Extensions Causality Mathematical models Conclusion

! crucial factor :

choice of a model usable during all the steps of this design procedure

‐ a knowledge model

* with real physical insight

* for all the domains of physics
* for
f dynamic
d i study
t d
* for energetic approach

‐ a representation model for controller design.

design

 the bond graph tool is well adapted for that purpose

N. CHATTI 36
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

MODELING APPROACH
1) Functional Analysis
Decomposition in sub‐systems exchanging power
 Word Bond graph

2) Phenomenological Analysis
Id ifi i off the
Identification h components and d physical
h i l phenomena
h
transforming the power they get in stored or dissipated energy
 Detailed Bond graph

3) Causal Analysis
Visualisation of the cause‐to‐effect relations and analysis of
causality conflicts
 Causal Bond graph

4) Structural Analysis
Causal paths in the bond graph
 actuators and sensors choice for control and monitoring

5) Building of associated mathematical models

N. CHATTI 37
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Orifice

DC motor

Shaft

Pinion Cylinder
Pi t
Piston

Rack

DC Pinion Cylinder
Shaft Piston
Motor
oo + +
Rack Orifice

Electrical Rotation Rotation Translation Pneumatic

+Rotation Mechanical +translation Mechanical
Mechanical Mechanical Pneumatic

N. CHATTI 38
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

DC MOTOR

ua DC T1
motor « a » for armature
1
ia « f » for field

uf
if
( Ra , La ) ia

T1
Tem
ua
e Load
1
( J , b)

Electrical Part Mechanical Part

N. CHATTI 39
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

ua DC T1 T2 Pinion Fpist P l Cylinder Penv

motor Piston Pcyl
Shaft + +
1 2 qv
ia Rack Vm Orifice qvenv

u T1 T2 pinion F P Cylinder Penv

DC Piston
shaft + +
Motor 1 2 V qv
i rack Orifice qv‐
env

Modeling hypothesis? Equations?

Questions : is it complete?
Q p Can these equations
q be simulated?

N. CHATTI 40
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

WORD BOND GRAPH OF THE PEM FUEL CELL

PRVH PVMH
2 2
 RVH
m  VMH
m
Valve
MF
2
H2 tank TRVH 2
TVMH
2 KV1 2 H2

H H VMH n MFH
RVH2
2 2

PHMO
U FC

Manifold

FC stack
2

DC/AC
PCHO
2
 HMO
m i FC
 CHO THM O 2
m 2

H HMO
2

Hu
umidifier
TCH O
2
PMHH O  MF
compressorr

2 O2
2
m MHH O iCL U CL
H CHO TMHH O n MFO
Air

2
2 2
2

H MHH O
PHTH O m
 2
2 HTHO
2
Load

H HTH O T
2 HTH2O  TME
H  TFE
H
ME FE
 THE
HHE

Tank H2O
Environment

N. CHATTI 41
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

DEFINITION REPRESENTATION
DEFINITION,

N. CHATTI 42
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

GENERALIZED VARIABLES

1- Power variables

power P(t )  e(t ). f (t )

e(t) « effort » f(t) « flow»

2 – Energy variables

energy  E(t)   P( )d

 e( )d
p (t ) " generalized momentum "

   f ( )d
q(t )  " generalized displacement "

 
4 generalized variables e, f, p, q

N. CHATTI 43
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BOND AS POWER CONNECTION

N. CHATTI 44
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BOND ACTIVATION

N. CHATTI 45
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BOND GRAPH MODEL IN BLOCK DIAGRAM

N. CHATTI 46
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLE

N. CHATTI 47
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

POWER VARIABLE FOR SEVERAL DOMAINS

DOMAIN EFFORT (e) FLOW (f)

Electrical VOLTAGE Current
u [V] i [A]
Mechanical FORCE VELOCITY
(translation) F [N] v [m/s]
Mechanical (rotation) TORQUE ANGULAR VELOCITY
 [Nm]  [rad/s]
H d li
Hydraulic PRESSURE VOLUME FLOW
P [pa] dV/dt m3 / s
Chemical CHEMICAL POTENTAL MOLAR FLOW
  J / mole  dn / dt  mole / s 

Thermal TEMPERATURE ENTROPY FLOW

T K  ds / dt  J / s 

N. CHATTI 48
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

WHY ENERGY VARIABLES

N. CHATTI 49
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

ENERGY VARIABLES FOR SEVERAL DOMAINS

DOMAIN DISPLACEMENT (q) IMPULSE (p)

Electrical CHARGE FLUX
q [Coulomb]  [wb]
Mechanical DISPLACEMENT MOMENT
(translation) x [m] J [Ns]
Mechanical (rotation) ANGLE ANGULAR MOMENTUM
 [rad] [Nms]
H d li
Hydraulic VOLUME MOMENTUM
V [m3] Ns / m 2
Chemical NBRE OF MOLE ?
  J / mole 

Thermal ENTROPY ?
T K 

N. CHATTI 50
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Energy variables: analogy

N. CHATTI 51
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

ELECTRICAL INDUCTION MOTOR

N. CHATTI 52
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

POWER STATION

N. CHATTI 53
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BOND GRAPH ELEMENTS

N. CHATTI 54
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BG ELEMENTS

Sources 3 passive elements (reçeive power)

Se, Sf
R : energy dissipation
C , I : energy storage

2 active éléments (supply power)

Se , Sf : effort source, flow source
Energy Junction Energy
gy
storage Structure Dissipation
I, C (O, 1, TF, GY) R 4 Junction elements (power
conservative)
0 1
0, 1, TF,
TF GY

2 detectors supposed ideal (no

Detectors power consumption)
De, Df
De, Df

N. CHATTI 55
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BOND GRAPH WELL SUITED AUTOMATED

ELEMENTS

N. CHATTI 56
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

PASSIVE ELEMENTS

 Representation

 Definition

The Bond Graph elements are called passive because they transform received
power into dissipated poser (R‐element), stored under potential energy (C‐
element) or kinetic (I‐element).

N. CHATTI 57
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

R ELEMENT

N. CHATTI 58
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLES OF R ELEMENTS

N. CHATTI 59
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BUFFERS

N. CHATTI 60
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLES OF C ELEMENTS

N. CHATTI 61
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

I ELEMENT

N. CHATTI 62
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

TETRAHEDRON OF STATE

N. CHATTI 63
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

TRANSFORMERS
 Convert
C t energy in
i one physical
h i l domain
d i as wellll as between
b t one physical
h i l
domain and another
 Examples: lever, pulley stem, gear pair, electrical transformer, change of
physical
p y domain…

N. CHATTI 64
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLES OF TRANSFORMERS

N. CHATTI 65
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

GYRATORS
 Convert
C t energy in
i one physical
h i l domain
d i as wellll as between
b t one physical
h i l
domain and another
 Examples: Gyroscope, Hall effect sensor, change of physical domain…

N. CHATTI 66
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLE OF GYRATOR: DC MOTOR

N. CHATTI 67
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

ACTIVATED ELEMENTS
 EFFORT AND FLOW SOURCES Se, Sf
 A source maintains one of power variabes constant or a specific function of time no matter how
large the other variable may be.

1. EFFORT SOURCE Se:

Examples: Generator of voltage, gravity force, pump, battery…

2. FLOW SOURCE Sf:

Examples: Current generator,
generator applied velocity…
velocity

N. CHATTI 68
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLES OF 0‐JUNCTION

N. CHATTI 69
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

1‐JUNCTION

N. CHATTI 70
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLES OF 1‐JUNCTION

N. CHATTI 71
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Summary

N. CHATTI 72
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BUILDING ELECTRICAL MODEL

1
1. Fix a reference direction for the current,
current it will be used as power direction
2. For each node in circuit with a distinct potential create a 0‐junction
3. Insert 1‐junction between 0‐junctions, attach all Bond Graph elements
submitted
b itt d to
t the
th potential
t ti l diff
difference (R,
(R I,
I C,
C Se,
S Sf elements)
l t ) to
t this
thi 1‐
1
junction
4. Assign power directions to all Bonds
5. For explicit ground potential, delete corresponding 0‐junction and its
adjacent Bonds. If non explicit ground potential is shown, choose any 0‐
junction and delete it.
6
6. Si lif resulting
Simplify lti Bond
B d Graph
G h (remove
( extaneous
t j ti
junctions. F example:
For l
1 0 1 is replaced by 1 1

Note:
Hydraulic, thermal systems similar, but mechanical different

N. CHATTI 73
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Simplifications of Bond Graph

N. CHATTI 74
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

ELECTRICAL CIRCUIT: EXAMPLE 1

N. CHATTI 75
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLE OF AN ELECTRICAL SYSTEM

R2
a R1 b c e R3

f
L1 C1
g
E
L2
d h

N. CHATTI 76
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLE OF AN ELECTRICAL SYSTEM

N. CHATTI 77
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

ELECTRICAL CIRCUIT: EXAMPLE 2

N. CHATTI 78
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Electrical circuit: Example 3

N. CHATTI 79
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Building Mechanical model

1
1. Fix a reference axis for velocities
2. Consider all different velocities (absolute veclocities for mass and inertia
and relative velocities for others).
3
3. FFor each
h distinct
di ti t velocity,
l it establish
t bli h a 1‐junction.
1 j ti Att h to
Attach t the
th 1‐junction
1 j ti
corresponding Bond Graph elements.
4. Express the relationships between velocities. Add 0‐junction (used to
represent those relationships) for each relationship between 1‐junctions.
1‐junctions
5. Place sources
6. Link all junctions taking into account the power direction
7. Eliminate any zero velocity 1‐junctions and their bonds.
8. Simplify Bond Graph by condesing 2 ports 0 and 1‐junctions into bonds: for
example:
1 0 1 is replaced by 1 1

N. CHATTI 80
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

MECHANICAL SYSTEM: EXAMPLE

N. CHATTI 81
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

SUSPENSION (1/4
/ OF A VEHICLE) : EXAMPLE

N. CHATTI 82
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

MECHANICAL SYSTEM: EXAMPLE

N. CHATTI 83
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

ELECTRO‐MECHANICAL SYSTEM

N. CHATTI 84
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXERCICE

N. CHATTI 85
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Building hydraulic models

1
1. Fix for the fluid a power direction
2. For each distinct pressure establish a 0‐junction (usually there are tank,
compressibility…)
3
3. Place a 1‐junction
Pl 1 j ti b t
between t
two 0 j ti
0‐junctions andd attach
tt h to
t this
thi junction
j ti
components submitted to the pressure difference
4. Add pressure and flow sources
5. Assign power directions
6. Define all pressures relative to reference (usually atmospheric) pressure,
and eliminate the reference 0‐junction and its bonds
7. Simplify the Bond Graph

N. CHATTI 86
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Hydraulic system: Example

N. CHATTI 87
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

THERMAL SYSTEM
T
a

Heat Source
Ts
Q S

N. CHATTI 88
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Orifice

DC Motor
M

shaft
Hypothesis :
Cylinder Jpinion neglected
Pinion
Piston

Rack

N. CHATTI 89
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Causality

 Causal analysis is the determination of the direction of the efforts and

flows in a BG model. The result is a causal BG which can be
considered as a compact block diagram. From causal BG we can
directlyy derive an equivalent
q block diagram.
g It is algorithmic
g level of
the modeling.

 Main outcomes of causal properties:

 Simulation
 Alarm flitering
 Monitorability
 Controllability
 Observability

N. CHATTI 90
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CONVENTIONS

N. CHATTI 91
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CONVENTIONS

e
A B P=e. f
f

A B A B

e f
A B A B

f e

Convention : the causal stroke is placed CLOSE TO (FAR FROM) the element
for which the EFFORT (the FLOW) is known

N. CHATTI 92
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

PRINCIPLE

N. CHATTI 93
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CAUSALITY

N. CHATTI 94
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CAUSALITY

N. CHATTI 95
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

INTEGRAL AND DERIVATIVE CAUSALITY

N. CHATTI 96
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CAUSALITY FOR 1‐JUNCTION

N. CHATTI 97
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CAUSALITY FOR 0‐JUNCTION

N. CHATTI 98
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CAUSALITY FOR TF‐JUNCTION

N. CHATTI 99
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CAUSALITY FOR GY‐JUNCTION

N. CHATTI 100
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CAUSALITY FOR GY‐JUNCTION

N. CHATTI 101
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

SEQUENTIAL CAUSALITY ASSIGNMENT PROCEDURE (SCAP)

 Apply a fixed causality to the source elements Se and Sf

 Apply a preferred causality to C and I elements

 With simulation, we prefer to avoid differentiation. In other words, with
the C‐element the effort‐out is prefered and with I –element the effort in
causality is preferred.

 Extend the causality through the nearly junction 0, 1, TF and GY

 Assign a causality to R‐element which have indifferent causality.

If these operations give a derivative causality on one element, it is

usually better to add other elements in order to avoid causal conflicts.
This elements must have a p physical
y meaningg ((thermal losses,,
resistance…)
N. CHATTI 102
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLE: HOW TO AVOID DERIVATIVE CAUSALITY

N. CHATTI 103
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

SUMMARY: RULES FOR ASSIGNING CAUSALITY

Causality Symbol Rule

Se e imposed by Se
required
Sf p
f imposed by
y Sf

C e C ( f() d

integral
I f  I ( e()
f ( )d )
d)


C f C ( de/dt 
derivative 
I 
e I ( df/dt )

R eRf
arbitrary
( li
linear case ) R f = 1/R
/ . e

R e R ( f
non arbitrary 
( non linear ) R 
f R ( e 

N. CHATTI 104
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

SUMMARY : RULES FOR ASSIGNING CAUSALITY

Causality Symbol Rule

2 i e1 = ei
e2 = ei
1 0 n .....
en = ei
aifi = - a1 . f1 - a2 . f2 -
.... - an . fn
2 i
Causality f1 = fi
restrictions f2 = fi
1 1 n .....
fn = fi
aiei = - a1 . e1 - a2 . e2
- .... - an . en

TF e1 = m . e2
m f 2 = m . f1

TF e2 = 1/m . e1
m f1 = 1/m . f2

GY e 1 = r . f2
r e 2 = r . f1

GY f1 = 1/r . e2
r f2 = 1/r . e1
105
N. CHATTI
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

PROCEDURE FOR ASSIGNEMENT OF CAUSALITY

1 – Assign
Assi n the req
required
ired ca
causality
salit to the so
sources
rces and e
extend
tend the implications thro
through
h
the graph as far as possible, using the constraint elements (0, 1, TF, GY)
2 – Assign a preferred integral causality to the I and C –elements and extend the
implications through the graph
3 – Assign a causality on the R‐elements
4 ‐ In case of causality conflict at a junction, search for the I‐ or C‐element
responsible of the conflict and assign it with a derivative causality
causality. Repeat step 3
3.

Se …. Se ….

R R

N. CHATTI 106
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLE 1

R C

Se

N. CHATTI 107
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLE 2

C : C1 I : I1 C : C2

Se 1 TF 0 GY 0 I : I2

R : R1 R : R2 R : R3

N. CHATTI 108
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLE 2

Other choice

C : C1 I : I1 C : C2

Se 1 TF 0 GY 0 I : I2

R : R1 R : R2 R : R3

N. CHATTI 109
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

EXAMPLE 3

I : I1 I : I2

Se 1 TF 1 C : C1

R : R1 R : R2

N. CHATTI 110
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

4 INFORMATION GIVEN BY THE BOND GRAPH

N. CHATTI 111
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CAUSAL PATH AND CAUSALITY

N. CHATTI 112
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

CAUSALITY  STRUCTURED WRITING OF

EQUATIONS
I:L C:C
L R1
2 5

E Se : E 1 0
R2 1 4
C 6
3
e
R : R1 R: R2
A B
f * Junction structure * Elements
1-Junction
P  e. f

B f : g B (e)

A e : g A ( f ) 0-Junction

N. CHATTI 113
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Jpinion

I:Jpin
I : LA I : J mot
I : mcrem C : Ccyll R : Roriff

u T1 T2 Fm
F
P P  Penv
1 GY 1 0 1 TF 1 TF 0 1
i  V
1 2 rp Vm Ap q vo

R : RA R : bmot
C : 1/ ksh

No causal singularity
No announced If the pinion inertia taken Derivative Causality!
simulation p
problem into account Acceptable?
cceptab e?
Go back to the model?
N. CHATTI 114
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

I : LA I: J I : mcrem C : Ccyl R : Rorifif

u T2 Fm
F
Pcyl Pcyl  Penv
T1 C2
1 GY 1 0 1 TF 1 TF 0 1
i 1 2 V qv
: rp V Ap

C : 1/ ksh
R : RA R :b

u DC T1 T2 Pinion F P Cylinder ‐Penv

Piston
motor shaft + +
1 2 V qv Orifice qv‐env
i rack

N. CHATTI 115
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

u T1 T2 pinion F P Cylinder
C li d Penv
DC Piston
shaft + +
Motor 1 2 V qv
i rack Orifice qv‐
env

u i Libraries of models : models with

T1
DC motor 1  a defined energetic environment
 a coherent choice of coupling variables
 a systematic determination of the
input‐output
input output variables

N. CHATTI 116
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

FROM BOND GRAPH TO BLOCK DIAGRAM

N. CHATTI 117
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

FROM BOND GRAPH TO BLOCK DIAGRAM

N. CHATTI 118
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

APPLICATION TO ELECTRICAL SYSTEM: BG MODEL

N. CHATTI 119
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

HOW TO OBTAIN STATE EQUATION

 Write structural laws associated with Junction (0,

(0 1,
1 TF,
TF GY)

 Write constitutive equations of each element (R,C,I)

 Combine those different laws to obtain state equation

N. CHATTI 120
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

APPLICATION TO ELECTRICAL SYSTEM: STATE EQUATION

N. CHATTI 121
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

APPLICATION TO ELECTRICAL SYSTEM: BLOCK DIAGRAM

N. CHATTI 122
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

APPLICATION TO ELECTRICAL SYSTEM: BLOCK DIAGRAM

N. CHATTI 123
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

NON DETERMINISTIC CAUSALITY

R1 C1 C : C1 I : L1

Se 1 0
E R2 L1

R : R1 R : R2

Price to pay:
‐ value of l ?
‐veryy small time constant

N. CHATTI 124
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Jpinion

No causal singularity
No announced If the pinion inertia taken Derivative Causality!
simulation p
problem into account Acceptable?
cceptab e?
Go back to the model?
N. CHATTI 125
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

u DC T1 T2 Pinion F P Cylinder ‐Penv

Piston
motor shaft + +
1 2 V qv Orifice qv‐env
i rack

N. CHATTI 126
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

u T1 T2 pinion F P Cylinder
C li d Penv
DC Piston
shaft + +
Motor 1 2 V qv
i rack Orifice qv‐
env

u i Libraries of models : models with

T1
DC motor 1  a defined energetic environment
 a coherent choice of coupling variables
 a systematic determination of the
input‐output
input output variables

N. CHATTI 127
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

I : La I : J mot
I : mcrem C : Ccyl R : Rorif

ua e Tem T1 T2 Fpiston Pcyl Pcyl  Penv

T2 Fpignon
1 GY 1 0 1 TF 1 TF 0 1
ia 1 2 rp Ap qvenv
Vm qv

C : 1/ ksh
R : Ra R : bmot

128
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Application

N. CHATTI 129
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

STATE EQUATION FROM BOND GRAPH

N. CHATTI 130
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

SIMULATION

N. CHATTI 131
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Hydraulic system

N. CHATTI 132
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

ELECTRIC SYSTEM

N. CHATTI 133
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

MECHANICAL SYSTEM

N. CHATTI 134
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

BG for Modeling
 Unified representation language
 Makes it easier the dialog between specialists of different physical
domains
 Simplifies the building of multidisciplinary models
 Shows up explicitely the power transfers
 Shows up explicitely the causality
 Allows a systematic writing of the mathematical models associated with
tthe
e BG
G

! Need a good knowledge of physics

Identification
 No « black box » models
 identification of unknown parameters of known physical phenomena
 Physical insight of the deduced mathematical model

N. CHATTI 135
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Analysis
 Showing up of the causality problems and thus of the ability for the model to be
simulated
i l d
 Stability Study (Ljapunov functions)
 estimation of the modes of the model and identification of the slow and fast variables
 Choice
Ch i off the
h simulation
i l i step
 Study of system architecture properties:
 Help for the designing of the actuator and sensor system (choice and placement)

Control
 Physical meaning of the variables
 Designing of observers from the state equations deduced from the BG
 Control law designing from reduced models

N. CHATTI 136
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Monitoring
 Graphical determination
• Of monitorability properties
• Of the sensor number, type, placement for the model to be monitorable
 Buidingg of analytical
y redundancyy relations

Simulation
 BG dedicated sofware (20sim, CAMPG+ACSL, ENPORT, MS1, ARCHER , ...)
 a priori knowledge of numerical problems (differential‐algebraic equations, implicite
equations) thanks to causality
 Study of the functioning in faulty mode
 To be done after a careful analysis of the model

N. CHATTI 137
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

«Integrated Design», « System Approach», « Mechatronic Approach»

+
Virtual test benchs

More and more used in industry

 Transverse
T andd simultaneous
i l engineering
i i
 Storage of engineering know‐how
 Reusability of validated solutions
 Designed product « good at the first attempt» (almost!)

Minimize COSTS
Respect DEADLINES

N. CHATTI 138
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Interest of using bond graph tool for new product designing

or improvement of existing products

D i wheel
Deming h l ‐ PDCA cycle
l : TTooll for
f quality
lit managementt : W.
W Edwards
Ed d DDeming
i (1950)

A P

C D

Plan
Act
Identify objectives
Criticize simulation
Results Act Plan Formulate hypothesis
Define validation criteria
Identify further needs

Check
Check Do
Causal analysis Do
Structural analysis Build model
Simulation

N. CHATTI 139
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Satisfaction of the 3 criteria

fundamental for companies

QUALITY ‐ COSTS ‐ DEADLINES

N. CHATTI 140
 REFERENCES

• H. Paynter "Analysis and design of engineering systems", MIT Press, 1961

• D. Karnopp, R. Rosenberg "Systems dynamics : a unified approach", John Wiley
& sons, 1975, 1991 (2nde édition)
• N. Chatti et al. "Model-based approach for fault diagnosis using set-membership
formulation" International journal of Engineering Applications of Artificial
Intelligence pages 307-319,
Intelligence, 307 319 vol.
vol 55,
55 2016.
2016
• N. Chatti et al. "Signed Bond Graph for multiple faults diagnosis", International
journal of Engineering Applications of Artificial Intelligence, pages 134-147,
2014.
• N. Chatti et al. "Functional and behavioural models for the supervision of an
intelligent and autonomous system" IEEE Transactions on Automation Science
and Engineering, pages 431-445, issue2, vol.10, 2013.
• R. Rosenberg, D. Karnopp "Introduction to physical system dynamics", series in
mechanical engineering, Mac Graw Hill, 1983
• J. Thoma "Introduction to bond ggraphs
p and their applications",
pp , Pergamon
g Press,,
1975
• B. Ould-Bouamama, N. Chatti and A.-L. Gehin "SBG for health Monitoring of
Fuel Cell System" ICREGA’14-Renewable Energy: Generation and Applications,
S i
Springer International
I i l Publishing,
P bli hi pages 73-85,
73 85 2014.
2014

N. CHATTI 141
 Fault diagnosis

Hydraulic system

N. CHATTI 142
 Fault diagnosis

Hydraulic system

N. CHATTI 143
 Fault diagnosis

Hydraulic system

N. CHATTI 144
 Fault diagnosis

Hydraulic system

N. CHATTI 145
 Fault diagnosis

Hydraulic system

N. CHATTI 146
Introduction BG methodology Model Designing Extensions Causality Mathematical models Conclusion

Hydraulic system

N. CHATTI 147
APPENDIX: PHYSICAL
SYSTEMS MODELING

148
149
150
151
152
153
154
155
156
157
159
160
161
162
163
164
165
166
167
168
169
170
171
SUPERVISION GUI
SUPERVISION GUI
FDI: MEDICAL INTERPRETATION
FDI STEPS IN TECHNOLOGICAL
PROCESS SUPERVISION
MODEL BASED FDI
FDI BASED ON IDENTIFICATION AND
OBSERVERS
NO MODEL BASED
DUPLEX REDUNDANCY
TRIPLEX REDUNDANCY
ARRS AND RESIDUALS
GENERAL PRINCIPLE
DETECTABILITY AND ISOLABILITY
DETECTABILITY AND ISOLABILITY