Sie sind auf Seite 1von 7

originalarbeiten

Elektrotechnik & Informationstechnik (2009) 126/5: 173179. DOI 10.1007/s00502-009-0639-z

Engine state monitoring and fault diagnosis


of large marine diesel engines
D. Watzenig, M. S. Sommer, G. Steiner

The reliable detection of engine malfunctions in order to predict and to plan maintenance intervals is of major importance in
various fields of industry. For instance, occurring faults of marine diesel engines which are on the high seas for months may lead to expensive
holding times. In this context, condition monitoring systems (CMS) should be able to assess engine health, to predict developing failures,
i.e. engine state degradation, and to diagnose failure modes at a low price. In this article, two different thermodynamical model-based
approaches to detect two common failure modes increased blow-by and compression ratio failures of large diesel engines given cylinder
pressure traces with low sampling rate are discussed and compared. Special focus is put on estimation robustness and reliability by excluding
the combustion phase and signal parts with high noise level. The proposed algorithms are validated with experimental data.
Keywords: parameter identification; blow-by; compression failure; elimination of offset terms

berwachung und Fehlerdiagnose von Schiffsdieselmotoren.


Zustandsu
In vielen Industriezweigen ist die zuverlassige Erkennung von Fehlfunktionen im Motor fur die Vorhersage und Planung von
Wartungsintervallen unabkommlich. Auf Hochseeschiffen, die sich oft mehrere Monate auf offener See befinden, kann ein Ausfall des
Motors zu teuren Standzeiten fuhren. Zustandsdiagnosesysteme (ZDS) sollten daher in der Lage sein, kostengunstig sowohl den
Gesundheitszustand des Motors abzuschatzen als auch Abnutzungserscheinungen rechtzeitig erkennen und identifizieren zu konnen.
In diesem Artikel werden zwei verschiedene auf thermodynamischen Modellen basierende Ansatze fur die Erkennung von zwei
haufig in Gromotoren auftretenden Fehlerursachen erhohtes blow-by und Kompressionsverluste unter Verwendung der
Zylinderinnendruckverlaufe mit geringer Abtastrate diskutiert. Besonderes Augenmerk liegt dabei auf der Robustheit und Zuverlassigkeit
der Parameterschatzung durch Ausblenden der Verbrennungsphase und von Signalteilen mit hohem Rauschpegel. Die vorgestellten
Algorithmen werden anhand von realen Messdaten validiert.
Schlusselworter: Parameteridentifikation; blow-by; Kompressionsfehler; Eliminierung von Offsetgroen

Received December 2, 2008, accepted January 30, 2009


Springer-Verlag 2009

1. Introduction
Although the history of diesel engines extends back to the end of
the nineteenth century and in spite of the predominant position
such engines now hold in various applications, they are still subject
of intensive research and development. Economic pressure, safety
critical aspects, compulsory onboard diagnosis as well as the reduction of emission limits lead to continuous advances in the development of combustion engines.
Condition monitoring and fault diagnosis is a valuable set of
methods designed to ensure that the engine stays in good condition
during its lifecycle. Diagnosis in the context of diesel engines is not
new and various approaches have been proposed in the past years,
however, recent technical and computational advances and environmental legislation have stimulated the development of more efficient and robust techniques. In addition, the number of electronic
components like sensors or actuators and the complexity of engine
control units (ECUs) are steadily increasing. Meanwhile most of the
software running on the main ECU is responsible for condition
monitoring of sensor signals, monitoring parameter ranges, detection of short/open circuits and verifying control deviations. However,
these kinds of CMS are not designed to detect and to identify
different engine failures, sensor drifts and to predict developing
failures, i.e. to asses degradation of certain components right in
time. Especially the reliable detection and separation of engine malfunctions is of major importance in various fields of industry in order
to predict and to plan maintenance intervals.

Mai 2009 | 126. Jahrgang

Diesel engines usually consist of a fuel injection system, pistons,


rings, liners, an inlet and exhaust system, heat exchangers, a lubrication system, bearings and an ECU. For the design of an efficient
CMS it is essential to know as much as possible about the underlying
thermodynamical processes and about possible faults and malfunctions. This information can be seen as a-priori knowledge and can be
used to increase the robustness of fault detection algorithms.
Common diesel engine faults and fault mechanisms, and their
causes are
"
"

"

"

power loss caused by misfire, blow-by and incorrect timing.


emission change caused by e.g. loss of compression, turbocharger malfunction, fuel filter blocked, incorrect injector timing,
poor diesel fuel, over-fuelling, air intake filter blocked, incorrect
piston topping, or ECU malfunction,
lubricating system fault due to incorrect oil pressure and oil
deterioration.
thermal overload as a result of one or a combination of leaking
injection valves, piston ring-cylinder wear or failure, eroded injec-

Watzenig, Daniel, Dipl.-lng. Univ.-Doz. Dr., Institute of Electrical Measurement


and Measurement Signal Processing, Graz University of Technology, Kopernikusgasse 24/4,
8010 Graz, Austria, and, The Virtual Vehicle Competence Center (ViF), Inffeldgasse 21 A,
8010 Graz, Austria; Sommer, Martin S., Dipl.-lng., Steiner, Gerald, Dipl.-lng. Dr.,
Institute of Electrical Measurement and Measurement Signal Processing, Graz University
of Technology, Kopernikusgasse 24/4, 8010 Graz, Austria
(E-mail: daniel.watzenig.at)

heft 5.2009

173

originalarbeiten
D. Watzenig et al. Engine state monitoring and fault diagnosis of large marine diesel engines

"
"
"

tor holes, too low injection pressure, high engine friction, misfire,
incorrect timing, leaking intake or exhaust manifold/valves, high
coolant or lubricant temperature, . . .
leaks in the fuel injection system, lubrication system, or air valves.
wear caused by either corrosion or abrasion, or both.
noise and vibration caused by the impact of one engine part
against another (mechanical noise), vibrations resulting from
combustion, intake and exhaust noise.

From the list above the main challenge in engine fault detection
can be derived the ambiguity between faults and causes. Certain
engine faults may be caused by a combination of causes (with
different weights) and certain causes may end up in different engine
faults. The assessment of engine states from sparse measurement
data and a reliable assignment of failure effects and causes are an
active research field.
The problems relating to marine diesel engines, especially medium- and high-speed engines, are due mainly to their large size and
there high operating speed. Occurring faults of marine diesel
engines which are on the high seas for several months may lead
to expensive holding times. On the other hand, additional sensors
and measurement equipment for condition monitoring are undesirable since engines have to be modified to place those additional
sensors.
A topical review on different fault diagnosis methods for condition monitoring can be found in (Jones, Li, 2000). Both standard
methods (Fourier analysis of pressure, torque, power, crankshaft
speed and vibration signals) and advanced methods (neural networks, fuzzy techniques) are encountered and briefly described.
Pontoppidan et al. discuss the detection of a single fault in a statistical framework (hypotheses testing) by measuring acoustic emission
energy signals and applying independent component analysis
(Pontoppidan, Sigurdsson, Larsen, 2005). However, most methods
usually rely on heuristic knowledge and on data training phase as
well as on the specification of threshold levels in order to assign
states as faulty or non-faulty. In the last decade, a paradigm shift
from classical signal processing and feature extraction to computational expensive model-based CMS can be observed. In contrast to
classical condition monitoring, model-based methods can manage
distributed and multiple correlated parameters (Woud, Boot, 1993).
They cover a wide variety of states since the engine behavior is
described in terms of physical relationships and hence, parameters
that influence certain parts of the first principles equations can be
isolated or at least correlations can be determined. Three different
methods to estimate the compression ratio from simulated cylinder
pressure traces are presented in (Klein, Eriksson, 2006) and compared in terms of estimation accuracy and computation time. By
reconstruction of only one single failure based on polytropic compression and expansion of the cylinder pressure remarkable results
have been reported. However, the detection of multiple failures
from in-cylinder pressure measurements is still an open issue.
In this work main focus is on a robust model-based identification
and separation of two common failure modes of large marine diesel
engines that cause very similar changes in the cylinder pressure by
accurately modeling the underlying thermodynamic process,
"

"

changes in the compression ratio primarily leading to emission


changes
increased blow-by mainly resulting in a loss of power

Following a model-based approach allows to identify the above


mentioned failures and to clearly separate them given uncertain
measurement data with low sampling rate (1 of crank angle). By
only measuring cylinder pressure traces of every cylinder, the symptoms due to faults are determined.

174

heft 5.2009

Two different approaches ratiometric and nonlinear parameter


estimation are investigated, validated with measured data and
compared to each other in terms of performance, accuracy, and
robustness given sensor drift and uncertain measurements.
2. Modeling of the thermodynamic process
Various approaches to model diesel engines have been proposed in
the literature, however, the main focus is on small-size engines that
are commonly used in the automotive industry. The typical differential equations that represent the thermodynamic processes, i.e. the
interrelations between system pressure, temperature and mass can
be found e.g. in (Heywood, 1988; Liu, Chalhoub, Henein, 2001;
Kouremenos, Hountalas, 1997).
Since in this work, identification of blow-by and compression ratio
is of primary interest a simplified thermodynamical model capable of
running in real-time is developed. The main reason for compression
losses are referred to damages of the piston crown during the
combustion phase leading to an increase in volume V0 in the top
dead center (TDC) of the piston. In the equation for the volume V in
the cylinder the constant volume V0 is considered by the term h0 A
with h0 representing the compression parameter and A the crosssectional area of the cylinder. In the time-varying fraction of the
volume equation ! denotes the instantaneous angular velocity of
the crankshaft. By also taking into account the ratio  of the crank
radius to the length of the connecting rod regarding to the equation
of a standard crank mechanism the equation for the volume and its
time derivative can be summarized in the form:

V
1  cos !t
V h0 A
2

q
  1  1  2 sin2 !t
!
dV V
3 cos !t

! sin !t 1 p
dt
2
1  2 sin2 !t

The time derivative of the mass fraction passing by the piston is


described by:
dm ~ 1
k p p
dt
T

with k~ denoting the parameter for blow-by. For simplicity it is


assumed that for the healthy state of the cylinder the effect of
blow-by can be neglected. Because of the fact that blow-by is
rapidly increasing when it comes to a tear-off of the oil film between
piston and liner due to the loss of the sealing function of the oil the
simple model of k~ as a constant is not sufficient. To model this
nonlinear behavior a sigmoid function described by:
k~p

k~max
1 eapb

is used where b describes the pressure when 50% of the maximum


blow-by is reached and a denotes the ascending slope of the sigmoid
function (Watzenig, Steiner, Sommer, 2008). For the complete thermodynamical description also the equations for the temperature T
dT
p dV kV 1 T 2 p3
Tin 

dt
mcv dT
mcv
and the pressure in the cylinder p


dp
dm
dT
dV 1
RT
R mp
dt
dt
dt
dt V

e&i elektrotechnik und informationstechnik

originalarbeiten
D. Watzenig et al. Engine state monitoring and fault diagnosis of large marine diesel engines

containing the isochore heat capacitance cv and the ideal gas constant R respectively are needed. Tin in Eq. (5) denotes the rapid
increase of the temperature in the cylinder during the combustion
phase and can therefore be neglected for the investigation of the
failure parameters during compression. For the healthy state of the
diesel engine the pair h0 k~ 0:150 for compression and blow-by
was identified.
3. Measurement noise
To achieve the goals of reliability and estimation robustness common perturbations of the pressure signal like detection uncertainties
of the TDC, pressure offset p0 and measurement noise nk have to be
analyzed. While the TDC-offset is corrected by the manufacturer
and the pressure offset can be included in the nonlinear parameter
estimation approach the task lies in finding the probability density
function (PDF) of the measurement noise. In the following the
assumption of a Gaussian PDF represented by:
"
#
1
x  2
; 1 < x < 1
7
px p exp
22
22
where  is the mean and 2 the variance of the random variable x is
verified. As can be seen in Fig. 1 the noise data extracted from
several measurements of the cylinder pressure signal is distributed
according to N 0; 2 . Therefore there exists no additional offset in
the pressure signal due to measurement noise. The range of the
analysis window of 90; 40 degrees to the TDC for the determination of the noise PDF was selected according to the reasonable
signal to noise ratio (SNR) in this area.
1.2

Increased
compression

Pmax_ref
Pmax
P2

P1
Pmin

100 .()
e1

e2

TDC

Fig. 2. Typical cylinder pressure traces representing a healthy


state (dashed) and a cylinder with increased compression ratio
(solid). The analysis window d(c) is applied within the well-defined
compression phase in order to avoid the influence of combustion
effects as well as measurement noise at low signal levels
state whereas the solid curve reflects a cylinder state with increased
compression ratio. The ratiometric parameter q allows to find
dependencies between the error parameters h0 and k~ and the position of the analysis window "1 ; "2  within the compression phase.
Due to the fact that blow-by has a strong nonlinear behavior causing
its main influence only at high pressures near the TDC, two analysis
windows are used with the lower window being placed before and
the upper window after the point of inflection of the cylinder pressure trace. To gain additional information the pressure trace within
the two windows is approximated by polynomials of the form

Noise data
Gaussian distribution

P P1 a1   a2 2 a3 3 :

1.0

Relative occurence

Healthy state

The healthy state (0% error) of the engine is described by the pair
0:150 whereas the maximum error (100% error) is represented by


0:16  2  105 . The procedure is described by a case study with
simulated data with 70% compression and 10% blow-by error in
the Figs. 35.

0.8

0.6

0.4

Analysis window [60, 30], = 0.047354 bar


3.35
Increased compression
Increased blowby

0
0.4

0.3

0.2

0.1

0.1

0.2

0.3

0.4

Noise amplitude

Fig. 1. Histogram of the measurement noise (repeated measurements)pcompared


to a Gaussian PDF (both curves are normalized
by 1== 2pr2 )

4. Condition monitoring algorithms

Pmax  Pmin
P2  P1

are displayed together with two typical traces of the cylinder pressure of the diesel engine. The dashed curve represents a healthy

Mai 2009 | 126. Jahrgang

3.30

3.25

3.20

3.15

3.10
0

4.1 Ratiometric approach


The main advantages of using a ratiometric approach lie in the
independency of a pressure offset in the measurement data, the
ease of implementation of the method, and the calculation speed. In
Fig. 2 the entire set of parameters for the determination of the
ratiometric variable q
q

Ratiometric parameter q

0.2

10

20

30

40

50

60

70

80

90

100

Error in %

Fig. 3. Ratiometric parameter q for the lower analysis window


[2608, 2308] to TDC (q 5 3.1857). The lines for compression and
blow-by error are too close to each other for a failure separation
The first parameter to be evaluated is the ratiometric parameter q.
Figure 3 illustrates that only monitoring q is not sufficient to distinguish between the two failure modes. Therefore the additional
parameters slope a1 and curvature a2 have to be evaluated for
confident failure separation. Figure 4 depicts that the compression
failure is overestimated by 10% and a separation of the two failure

heft 5.2009

175

originalarbeiten
D. Watzenig et al. Engine state monitoring and fault diagnosis of large marine diesel engines

Analysis window [60, 30], = 0.047354 bar

Because of the small values of a2 the disturbance of the curve by


measurement noise with  0:047354 bar becomes visible. This
sensitivity to measurement noise is the main drawback of this
method.
Consequently, in order to apply the ratiometric principle to measured data, Eq. (8) for the calculation of the ratiometric parameter
has to be modified in terms of:

0.0328
Increased compression
Increased blowby

0.0326

Slope a1

0.0324
0.0322
0.0320
0.0318

qmod

0.0316
0.0314

10

20

30

40

50

60

70

80

90

100

Error in %

Fig. 4. Slope parameter a1 for the lower analysis window [2608,


2308] to TDC (a15 3.180  1022). The lines for compression and
blow-by error are still too close together for a separation of the
failure modes

2.55

10

Analysis window [60, 30], = 0.047354 bar

Increased compression
Increased blowby

Curvature a2

2.50

2.45

Pdefect
Phealthy

10

with Pdefect representing a cylinder with either blow-by or compression failure. In Fig. 6 the different sign of curvature of qmod can be
determined very easily. As can be seen the greatest differences occur
at crank angles close to the TDC which are partly outside of the
observation window limited by the upper bound of 8 to the TDC.
In Table 1 the coefficients according to Eq. (9) are summarized for
three cylinders with known failure sources of two different engines
excluding the pressure offset. The different signs of the coefficients
a1 and a3 for blow-by and compression failure reflect the different
curvatures. However, along with an increasing measurement noise
variance this method becomes unusable in terms of reliability.
Due to the limitations mentioned above and the fact that the
ratiometric approach only allows for a qualitative statement, a more
robust and reliable method has to be developed the nonlinear
parameter estimation approach.

2.40

2.35

2.30

10

20

30

40

50

60

70

80

90

100

Error in %

Fig. 5. Curvature parameter a2 for the lower analysis window


[2608, 2308] to TDC (a2 5 2.447  1025). Because there is only
one failure mode in the allowed range a2 allows the separation
between blow-by and compression error
modes is not possible. The evaluation of the curvature information
a2 shown in Fig. 5 allows the distinction between compression and
blow-by error but the compression error is still overestimated.

1.05

4.2 Nonlinear parameter estimation


The proposed approach aims at finding a parameter vector 
h0 k~ p0 T which is comprised of the compression ratio h0 , the
blow-by parameter k~ and the pressure offset p0 by minimizing the
L2 -norm of the error kek22 ! min between measured data and computed cylinder pressure in a nonlinear least squares sense for each
cycle. The block diagram is shown in Fig. 7. The disturbance of the
data y~k due to measurement noise nk is considered by an additional
summation node with the output y k representing the corrupted
data. The thermodynamic model is calibrated for a measured healthy
state prior to the parameter identification by adapting the parameter
vector u 1 2 3 Rcv kT . The parameter identification problem
consists of finding the set of parameters  2 Rn that minimize the
target function f() at a single point. The inequality constraints

Blowby error measured


Blowby error fit
Compression error measured
Compression error fit

1.04
1.03

Pdefect/Phealthy

1.02
1.01
1.00
0.99
0.98
0.97
0.96
0.95
60

50

40

30

20

10

Crank angle to TDC in

Fig. 6. Comparison of the ratio Pdefect= Phealthy for an engine showing blow-by error and compression error, respectively, showing different
curvature in their slopes especially in the interval [2408, 08]

176

heft 5.2009

e&i elektrotechnik und informationstechnik

originalarbeiten
D. Watzenig et al. Engine state monitoring and fault diagnosis of large marine diesel engines

a2

a3

Occurred failures

a1

Increased blow-by

3:99104 3:57106 1:09108

Changed compression
ratio

5:38103 4:99104
3:32103 3:98104

Real
system

nk

yk

yk

uk

Model

1:62105
1:29105

^
yk

ek

|| . ||22

Fig. 7. Block diagram of the parameter identification of large diesel


engines. The index k indicates the iterative nature of the opitimization procedure. By minimizing the residual error between measured
and calculated cylinder pressure, the optimal parameter configuration for blow-by and compression ratio is found. Based on the a
priori known limits of the parameters, the engine state can be
assessed and monitored
simultaneously have to be satisfied at this single point where both
the target and constraint functions depend on the parameter vector.
The objective is to find a parameter configuration that satisfies

5. Condition monitoring results


In the following, results for two measured data sets of different engines
containing single blow-by and single compression ratio failure are presented. For the evaluation of the source of defect the engines were
disassembled by the manufacturer. The reason for lower compression
ratios was identified as burn-of of the piston crown whereas increased
blow-by occurred due to defects of the crankcase cover gasket. Figure 8
shows of the pressure traces of a five cylinder (top) and a seven cylinder
(bottom) diesel engine respectively. In both cases the sources of defect
were known. Because blow-by errors often lead to severe damages of
the engine most of the time the crankcase cover gaskets are replaced
before the error occurs and therefore there exist only a few data sets
where blow-by is documented. Figure 8 bottom shows such a case for
one cylinder of a seven cylinder diesel engine. As can be seen the single
pressure traces are close together up to the TDC. As the observation
window is limited by 8 to the TDC, the area with the greatest change
in the cylinder pressure cant be used which makes the detection and
separation of the interesting failures a challenging task. For quantification the model limits for the parameters to be estimated are again
h0 0:15 0:16 for compression and k~ 0;  2105  for blowby corresponding to 0%100% of failure.
120

min f  min k y~k  y^k k22


s:t:

Healthy state
Increased compression
Errorfree state
Increased compression
Increased compression

100

Cylinder pressure in bar

Table 1. Coefficients of the fitting polynomial

80

60

40

20

bl    bu
0

zk  y k



150

200

250

300

350

400

Increased blow by
Healthy state

100

80

60

40

20

12
13

where 2 0 360 denotes the crank angle in degrees. The modelbased estimation of  is based on the windowed signal zk by solving
the constrained nonlinear optimization problem (11). Signal parts
with low signal magnitude as well as the signal part that corresponds to the combustion phase depicted in Fig. 2 are cut off for
the estimation procedure. The dashed curve representing the
healthy state is used to calibrate the thermodynamical model by
adapting the model parameter vector u.

Mai 2009 | 126. Jahrgang

100

120

l if "1   "2
0 else

50

Crank angle in

Cylinder pressure in bar

where y~k denotes the measured cylinder pressure and y^k  represents the estimated measurements based on the thermodynamic
model. The bounds bl and bu are the lower and the upper bound
for the unknown parameter vector, i.e. the imposed constraints on
the parameters to be reconstructed from measured data.
In order to mask out undesired effects of the starting combustion
close to the TDC and the low SNR at small cylinder pressures, an
analysis window  2 "1 ; "2 is applied to the measured cylinder
pressure y k . The proposed rectangular window is mainly restricted to
the compression phase. If the entire signal y k is provided to the
parameter identification problem, a robust detection and identification of blow-by and compression ratio failures is impossible
since various other effects influence the cylinder pressure during
combustion.

50

100

150

200

250

300

350

400

Crank angle in

Fig. 8. top: Measured cylinder pressure traces representing a


healthy state and cylinders with increased compression of one
specific engine; bottom: Measured cylinder pressure representing a
healthy state and one cylinder with increased blow-by of a different engine. The sources of defect were in both cases documented by
the manufacturer after disassembly of the machine
Table 2 summarizes the results of the estimated parameter vector
 with varying upper bound "2 of the analysis window. The first

heft 5.2009

177

originalarbeiten
D. Watzenig et al. Engine state monitoring and fault diagnosis of large marine diesel engines

Occurred failures

-TDC

h0

k~

Acknowledgements
This research was founded as part of the Christian Doppler Laboratory for Automotive Measurement Research.

Increased blow-by

8
9
10

0.150
0.150
0.150

1:84105
1:67105
1:99105

Appendix

Changed compression
ratio

8
9
10

0.154
0.154
0.154

3:011015
2:721015
1:721015

Table 2. Results for estimated blow-by and compression ratio

block indicates increased blow-by given the desired compression


ratio while in the second block a changed compression ratio is clearly
identified. The blow-by remains very small denoting that blow-by
has not increased.
6. Conclusions
This paper addresses two different methods for robust detection of
increased blow-by and compression faults from measured cylinder
pressure traces of large marine diesel engines. By accurately modeling the underlying thermodynamic process, including prior knowledge about the system, and characterizing the measurement noise,
faults can be detected and isolated from each other even in the
presence of sensor drift. The ratiometric approach allows only for
qualitative statements and is not capable of clearly distinguishing
between the two failure modes. Main drawbacks are the sensitivity
to measurement noise and the fact that crank angles close to the
TDC are required where one has typically to deal with effects of
starting combustion. On the other hand the method is very fast due
to its simplicity and independency to a global pressure offset in the
measurement signal.
The nonlinear parameter estimation approach overcomes the
drawbacks of the ratiometric method. The robustness is investigated by examining the influence of the upper limit of the analysis
window close to the TDC. The applicability of both model-based
approaches is verified by measurement data given information
about the sources of defect of the engine. Due to the low sampling interval of 1 of the crank angle both condition monitoring
systems (CMS) exhibit real-time performance. The detection of
these failures can be used in order to predict maintenance
intervals.

Nomenclature
h0
A
V
m
T
P
R
cv
k~
k
1
2
3

compression parameter
cylinder cross-sectional area
cylinder volume
mass of the mixture
temperature of the mixture
cylinder pressure
ideal gas constant
isochore heat capacitance
blow-by parameter
constant
power of volume
power of temperature
power of pressure

m
m2
m3
kg

K
bar
J/(molK)
J/(kgK)

References
Heywood, J. B. (1988): Internal combustion engine fundamentals, McGrawHill.
Jones, N. B., Li, Y.-H. (2000): A review of condition monitoring and fault diagnosis for
diesel engines. Tribotest J., 6: 267291.
Klein, M., Eriksson, L. (2006): Methods for cylinder pressure based compression ratio
estimation, SAE Technical Paper Series, 2006010185, SAE World Congress, Detroit,
USA, April 36.
Kouremenos, D. A., Hountalas, D. T. (1997): Diagnosis and condition monitoring of
mediumspeed marine diesel engines. Tribotest J., 4: 6391.
Liu, H.-Q., Chalhoub, N. G., Henein, N. (2001): Simulation of a single cylinder diesel engine
under cold start conditions using Simulink. J. Eng. Gas Turbines Power,
123: 117124.
Pontoppidan, N. H., Sigurdsson, S., Larsen, J. (2005): Condition monitoring with
mean field independent component analysis. J. Mech. Syst. Signal. Pr., 19:
13371347.
Watzenig, D., Steiner, G., Sommer, M. S. (2008): Robust estimation of blow-by and
compression ratio for large diesel engines based on cylinder pressure traces, IEEE
Instrumentation and Measurement Technology Conference (IMTC), Vancouver,
Canada, May 1215, 974978.
Woud, J. K., Boot, P. (1993): Diesel engine condition monitoring and fault diagnosis based
on process models. 20th Int. Congress on Combustion Engines, London.

Authors
Daniel Watzenig
received his M.Sc. and his Ph.D. degree in
Electrical Engineering from Graz University
of Technology in 2002 and 2006, respectively.
In 2009, he received the venia docendi on
electrical measurement and measurement signal processing. He is currently heading the
Vehicle Electrics and Electronics Group of
the Virtual Vehicle Competence Center (K2
Mobility SVT Competence Center of Excellent
Technologies) in Graz. His research interests include statistical inverse
problems, automotive electronics and probabilistic design. He is distinguished lecturer at the Institute of Electrical Measurement and
Measurement Signal Processing at Graz University of Technology.

178

heft 5.2009

Martin Sommer
received his M.Sc. degree in Mechatronics in
2007 from the Johannes Kepler University
Linz. He is now working on his Ph.D. in the
field of model based measurement at the
Institute of Measurement and Measurement
Signal Processing at Graz University of Technology. Among his research interests are signal processing and mathematical modeling
of mechatronic systems.

e&i elektrotechnik und informationstechnik

originalarbeiten
D. Watzenig et al. Engine state monitoring and fault diagnosis of large marine diesel engines

Gerald Steiner
received the Dipl.-Ing. and Dr.techn. degrees in
Electrical Engineering from the Graz University
of Technology in 2002 and 2006, respectively.
He is currently with Anton Paar GmbH, Graz, as
R&D project coordinator for process instrumentation and serves as external lecturer at
the Graz University of Technology. His research
interests include industrial inverse problems,
sensor fusion, and process instrumentation.

Mai 2009 | 126. Jahrgang

heft 5.2009

179