Académique Documents
Professionnel Documents
Culture Documents
advantage by technology
Dr.-Ing. Carsten Schnfelder, Bertrandt Ingenieurbro GmbH, Ingolstadt
Automotive Colloquium HTW Dresden, 31.05.2011, 3 pm, L 211
www.seriouswheels.com
Ingolstadt
Nikolaus Otto
Quelle: www.wissen.de, u.a.
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
Ingolstadt
Ingolstadt
Content
Introduction
Presentation Bertrandt AG
What is calibration?
Calibration processes
Calibration on the engine test stand
Model based calibration with DoE
(design of experiments)
Alternative calibration processes
Automobile calibration
Loading and oxidation of DPF systems
Misfire detection diagnosis of gasoline engines
Summary and outlook
Quelle: www.audi.de
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
Ingolstadt
Inhalt
Introduction
Presentation Bertrandt AG
What is calibration?
Calibration processes
Calibration on the engine test stand
Model based calibration with DoE
(design of experiments)
Alternative calibration processes
Automobile calibration
Loading and oxidation of DPF systems
Misfire detection diagnosis of gasoline engines
Ingolstadt
Bertrandt Services
Bertrandt
Competency
body
electronics
simulation
interior
powertrain
test division
Ingolstadt
434
6.523
6.080
429
385
5.431
340
4.708
employees Powertrain
547
GJ 06/07
GJ 07/08
GJ 08/09
450
GJ 09/10
126 *
650
529
GJ 10/11
GJ 06/07
GJ 07/08
GJ 08/09
GJ 09/10
GJ 10/11
300
Ingolstadt
Structure of shareholder
Ingolstadt
History
1974
1990
since 2004
Germany
Ingolstadt
Neckarsulm
Mnchen
Ehningen
Wolfsburg
Kln
Rsselsheim
France
Paris (Bivres)
Great Britain
Dunton
Spain
Barcelona
Ingolstadt
Ingolstadt
10
Ingolstadt
Simulation development of
powertrain management
quantitative estimation of different drive
concepts
interpretation of individual components of
different powertrain configurations
integration of additional information
signals (GPS, traffic, environment, )
in the drive management
estimates and analysis of fuel
consumption and emission potentials
development of operational strategy
11
Ingolstadt
Engine calibration
calibration, thermodynamics and exhaust aftertreatment on the
engine test stand and in the vehicle
combustion process development
(gasoline, diesel, HCCI, PCCI, CAI)
exhaust aftertreatment (DPF, SCR, )
gas exchange analysis, pressure analysis
functional development and testing
Adaptation of calibration to fuels
vehicle start-up / exhaust measurements
endurance run monitoring
DoE (Design of Experiments)
OBD development
measuring data acquisition and processing
12
Ingolstadt
Workshop
reconstruction / rebuilding of test vehicles and prototypes
(components fittings, electrification, start-up, )
measurement engineering installation in test vehicles
engine replacement and accordingly change of components on functional vehicles
prototype assembly of components
wiring harness production for function vehicles
ads
22. Juni 2011 / Dr.-Ing. Carsten Schnfelder
13
Ingolstadt
Tools
CAD systems:
CATIA V4/ V5
ICEM Surf
Pro/ENGINEER
I-DEAS
Unigraphics
Measuring engineering
tools:
DIAdem
Canon
LINoe
INCA
LabVIEW
simulation tools:
MSC/NASTRAN
ABAQUS
OptiStruct
STAR-CD
Flowmaster
Dymola
Kuli
ADAMS
developement tools:
MATLAB/Simulink
TargetLink
ASCET
AVL CAMEO
calibration tools:
INCA
DiagRA
LABCAR
Codes
AVL PUMA
AVL CONCERTO
14
Ingolstadt
15
Ingolstadt
16
Ingolstadt
Quelle: Audi AG
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
17
Ingolstadt
Content
Introduction
Presentation Bertrandt AG
What is calibration?
Calibration processes
Calibration on the engine test stand
Model based calibration with DoE
(design of experiments)
Alternative calibration processes
Automobile calibration
Loading and oxidation of DPF systems
Misfire detection diagnosis of gasoline engines
Summary and outlook
Quelle: www.audi.de
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
18
Ingolstadt
Electronics:
Control unit (ECU)
Mechanics:
Combustion engine
Actuators
Sensors
Driver
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007, Renault / Bosch / VKA / TU Dresden (IVK)
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
19
Ingolstadt
AD
C
control unit
actuators
Function
Calculator
RAM
FlashEPROM
EEPROM
Observer
20
Ingolstadt
data set
adapted by calibration engineer
access to control unit using special software
torque
4
0.15
parameter
5
injection
rate
2
actual size
Control parameter
torque
map
100
injection rate
low pass
3
nominal size
1
activation
switch
integrator
saturation
1
I_rate
control model
software function
contains structural information, inputs and outputs, calculation blocks
compiled by control equipment manufacturers
no access by calibration engineer
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007 Vieweg Dieselmotor-Management / Bosch
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
21
Ingolstadt
22
Ingolstadt
23
Ingolstadt
24
Ingolstadt
25
Ingolstadt
Content
Introduction
Presentation Bertrandt AG
What is calibration?
Calibration processes
Calibration on the engine test stand
Model based calibration with DoE
(design of experiments)
Alternative calibration processes
Automobile calibration
Loading and oxidation of DPF systems
Misfire detection diagnosis of gasoline engines
Summary and outlook
Quelle: www.seriouswheels.com
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
26
Ingolstadt
electronics
Cause and
consequence
functionality
complexity
understanding
quality
reliability
trust
27
Ingolstadt
requirement
management of increasing
complexity
(understanding, data amount)
ensuring quality and reliability
reduction of development time
and costs (e.g. prototype
vehicles)
Quelle: Fast ECU Access via EtherCAT, Schnellbacher, ETAS GmbH
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
28
Ingolstadt
requirement
expanded use of simulation
methods
standardized processes
and automated procedures
repeatability, portability
tool-assisted optimization
Quelle: ETAS
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
29
Ingolstadt
requirements
system calculation
and design
integrated model-based
process of software
development and calibration
HiL
*
system spec.
component calculation and
design
HiL*
Components spec.
MiL*
SiL*
design of control
software
software spec.
model as "executable
specification"
MiL: Model-in-the-loop
SiL: Software-in-the-loop
HiL: Hardware-in-the-loop
data exchange by
offline calibration
product
endurance
test
ECU vehicle
electr.integr.
vehicle
calibration
ECU Motor
requirement and advantage
continuous
engine calibration
Softwaremodel-based
development process
integration (test bed)
models as graphical "living"
ECU Code specification
early understanding (including
functional development
calibration)
early testing, early correction
automated, repeatable and
complete testing
documented process
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007
30
Ingolstadt
(hardware)
Model-in-theloop (MiL)
validation
control
model
control
model
(Hardware)
Hardware-inthe-loop (HiL)
(hardware)
rapid control
prototyping (RCP)
control
software
control
software
process
process models
control models
calibration
process
model
process
model
control
model
real
prototyping
physical models
formula relationships
physical effective direction
non-physical models
characteristic fields from
measurements
analytical (polynomial models, etc.)
process
physical models
Formula relationships (inverse)
model-based predictive control
non-physical models
characteristic maps / curves
analytical (e.g. neural networks )
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007
31
Ingolstadt
Content
Introduction
Presentation Bertrandt AG
What is calibration?
Calibration processes
Calibration on the engine test stand
Model based calibration with DoE
(design of experiments)
Alternative calibration processes
Automobile calibration
Loading and oxidation of DPF systems
Misfire detection diagnosis of gasoline engines
Summary and outlook
Quelle: Bertrandt
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
32
Ingolstadt
=> 5 measurements
4 parameter
start of injection
rail pressure
EGR
boost pressure
=> 25 measurements
3 parameter
start of injection
rail pressure
EGR
5 Parameter
start of injection
rail pressure
EGR
boost pressure
SL mass
33
Ingolstadt
34
Ingolstadt
x1
x2
Process model
b1, b2, bp
xn
Coefficients
(to be calcul.)
p
Linear model:
y = bi fi (x1,..,xn )
y
Output value,
measured
(e.g. soot emission SZBO*)
*SZBO: Schwarzrauch, gemessen
ber Filtertrbung, korrelierbar zu
Partikelmassenemissionen
i =1
Polynominal model:
y = b1 + b2 x1 + b3 x2 + b4 x1 x2 + b5 x12 + b6 x22
(Bsp.)
35
Ingolstadt
Neuronal network:
Calculation units with weights
Learning by connections,
weights, thresholds
Interpretation difficult
y = bi ( x ci
i =1
(univariant functions):
1
= 2
r + w2
mit
p: Number of terms, b:
Weigthing, c: Center,
||..|| euclidic distance, w: Width
x1
x2
x3
Linear
transfer
function
x4
Dendriten Weight
Transfer
(Synapsen) function
36
Ingolstadt
2
(yi yi ) = Minimum
Linear regression
y1 f1(x1) L fp (x1) b1 1
=
M
M
O
M
M + M
y f (x ) L f (x ) b
{
n
p n p n
114n44244
4
3 { {
y
y = X
2 !
2
2
(yi yi ) = (y y) = y X = Minimum
( )
Model
y
(AG)
Model value
yi
Residuum
(yi yi )
y
Meas. point i
Design value
x (EG)
1
= XT X XT y
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
37
Ingolstadt
Err.
Res.
No LOF
Modell i.O.
Err.
Res.
No LOF
Modell i.O.
x
LOF
Modell wrong
x
Err.: Pure Error
Res.: Residuum between model and measurement
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
Err..
Res.
Err.
Res.
Overfit
Modell wrong
x
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007
22. Juni 2011 / Dr.-Ing. Carsten Schnfelder
39
Ingolstadt
Design space
Input values, 3-8 typical for Diesel engines (injection
quantity, timings, rail pressure, air mass / EGR)
Definition of the boundaries
Measurement points
Optimization
Map calculation
After the optimization for each operational point
Calculation with regard to map smoothness
Calculation of models
Optimization using models
Final check
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007
40
Ingolstadt
variables
Optimization regarding
NOx minization
NOx
EGR rate
Consumption
Noise
Optimization leads to lower NOx and soot emissions, but also slightly increasing fuel
consumption
Additional reduction of particulate emissions by aftertreatment
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007, FEV PROcal
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
Ingolstadt
Content
Introduction
Presentation Bertrandt AG
What is calibration?
Calibration processes
Calibration on the engine test stand
Model based calibration with DoE
(design of experiments)
Alternative calibration processes
Automobile calibration
Loading and oxidation of DPF systems
Misfire detection diagnosis of gasoline engines
Summary and outlook
Quelle: Bertrandt, TU Darmstadt
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
42
Ingolstadt
Data Acquisition
Data Analysis
Data Optimization
43
Ingolstadt
Measurements
Evaluation
Plausibility (limits, elimination of oszillations)
Weight factors (stationary optimum)
Limitation of maximum gradients of actuator
signals (transient optimum)
Automated map calculation and optimization
with regard to map smoothness
44
Ingolstadt
Spezifischer
Verbrauch [g/kWh]
Test plan
Detection of oszillations
and automized correction
Datenzeile
Measurements
Evaluation
Plausibility (limits, elimination of oszillations)
Weight factors (stationary optimum)
Limitation of maximum gradients of actuator
signals (transient optimum)
Automated map calculation and optimization
with regard to map smoothness
45
Ingolstadt
Measurements
Evaluation
Plausibility (limits, elimination of oszillations)
Weight factors (stationary optimum)
Limitation of maximum gradients of actuator
signals (transient optimum)
Automated map calculation and optimization
with regard to map smoothness
46
Ingolstadt
Content
Introduction
Presentation Bertrandt AG
What is calibration?
Calibration processes
Calibration on the engine test stand
Model based calibration with DoE
(design of experiments)
Alternative calibration processes
Automobile calibration
Loading and oxidation of DPF systems
Misfire detection diagnosis of gasoline engines
Summary and outlook
Quelle: Opel / Auto-Reporter.net, u.a.
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
47
Ingolstadt
Exhaust
gases
Temp. sensor
With NO2
NO2 + C CO + NO
2 NO2 + C CO2 + 2 NO
48
Ingolstadt
In case of problems:
Regeneration/
State
Adsorption/
Transition
Error
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
49
Ingolstadt
Soot emission
from engine
Light off/
Adsorption/
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
50
Ingolstadt
NOx
NOx
Massenstrom
mass flow
Soot
Russmassenmassflowstrom
Without
ohne
EGR
AGR
++
Soot
RussmassenmassflowWith
stromEGR
mit AGR
dynamical
correction
Einspritzinjection0
quantity
menge
2000
4000
Drehzahl
speed
NO2 effect
Speed
Injectionquant.
Temperature
Soot mass
NOx
NOx
Massenmas strom
flow engine
Motor
NO2
mass flow (-)
NO2 Konvertieru.
conversion .
akt.Russbelad.
Act,.soot load .
Integration
Soot mass
In filter
[g] or [g/l]
Oxidation
Temperaturemodel DPF
f(u)
Oxidation
(-)
Volume flow.
Lambda
Soot mass
z.B. reaction kinetics
52
Ingolstadt
Conventional
Many test runs
Time for development
High costs
Vehicle
First measurement
Soot emissions
NO2 effect
Filter load
Regeneration
Model supported
Decoupling from test
bench and vehicle
Fewer prototype vehicles
Reduction of time and
costs
Increasing quality
Automation of process
Simulation
Offline with PC
Load model
Regeneration
Vehicle
Loading test runs
Filter load
Regeneration
Optimization
calibration
Requirement
fulfilled?
Requirement?
Vehicle
Validation test run
54
Ingolstadt
Optimized calibration
of the loading model
by offline simulation
Quelle: Promotionsvortrag Schnfelder, RWTH Aachen 2007, FEV TOPexpert vehicle suite
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
55
Ingolstadt
Content
Introduction
Presentation Bertrandt AG
What is calibration?
Calibration processes
Calibration on the engine test stand
Model based calibration with DoE
(design of experiments)
Alternative calibration processes
Automobile calibration
Loading and oxidation of DPF systems
Misfire detection diagnosis of gasoline engines
Summary and outlook
Quelle: div.
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
56
Ingolstadt
Consequence:
Unburned fuel into exhaust
Raised exhaust emissions
Catalyst destruction by
overheating
Lost Power
Disturbed engine running
Quelle: Diplomarbeit Jan Brycki, Bertrandt Ingolstadt, KFZ-Technik SEUBERT
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
57
Ingolstadt
58
Ingolstadt
59
Ingolstadt
Calculation of
Thresholds
Calculation of
engine roughness
Rating of
engine roughness
60
Ingolstadt
Calculation of
engine roughness
Calculation of
Thresholds
Rating of
engine roughness
Engine roughness = Segment time (n +1) - Segment Time (s) - Compensation time
The difference between the stored segment time and the current segment time taking
into account of dynamic processes (accelerating or braking) results in rough running
for each cycle. The rough running is normally zero.
61
Ingolstadt
Calculation of
engine roughness
Calculation of
Thresholds
Rating of
engine roughness
differences between
running-period
Averange delay during the last
engine revolution (trend)
62
Ingolstadt
Calculation of
engine roughness
engine speed
load
Calculation of
Thresholds
Rating of
engine roughness
Reference
value to engine
roughness
Adaptation of
flywheel signal
temperature
engine speed
63
Ingolstadt
Calculation of
engine roughness
Calculation of
Thresholds
Rating of
engine roughness
Engine roughness
Detected bit
Misfire detected
Misfire detected
Quelle: Diplomarbeit Jan Brycki, Bertrandt Ingolstadt
64
Ingolstadt
Segment
period
Flywheel signal
correction
Segment
period
Flywheel signal
correction
Segment
period
Flywheel signal
correction
Engine roughness
calculation > 0,2% - 8%
Misfire detectionalgorithm
Error memory
Management
Malfunction
Indicator
Lamp
Injection blank-out
Pre-/
debouncing
diagnostic tester/
PC
65
Ingolstadt
1. Measurement
Filtering of data
n- iterations
n - optimal
datasets
Tasks:
Legislation
EOBD
CARB-tables
Map
calculation
automated
Simulation
Map
optimization
detection rate
calculation
DCM
Format /
MCD3
Advantages:
Reduction of measurement effort, increase in quality, reproducibility, automated calibration
Quelle: Diplomarbeit Jan Brycki, Bertrandt Ingolstadt
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
66
Ingolstadt
67
Ingolstadt
INCA
interface
ECU
DMD - Tool
Write calibration
data
- values
- curves
- maps
68
Ingolstadt
Content
Introduction
Presentation Bertrandt AG
What is calibration?
Calibration processes
Calibration on the engine test stand
Model based calibration with DoE
(design of experiments)
Alternative calibration processes
Automobile calibration
Loading and oxidation of DPF systems
Misfire detection diagnosis of gasoline engines
Summary and outlook
Quelle: www.seriouswheels.com
Applikation moderner Verbrennungsmotoren Vorsprung durch Technologie
69
Ingolstadt
The drivetrain of future vehicles is one of the most interesting development task in
the automobile world
Calibration is the central position in the drivetrain development
Thermodynamics, mechanics, E/E, software, methods
Using intelligent technologies like the model-based calibration its possible to
realize more functionalities
reduce additional engineering effort
decrease emissions and consumption
increase fun to drive
last but not least: earn money
70
Ingolstadt
71