[Zurück]


Zeitschriftenartikel:

D. Erdogan, S. Jakubek, C. Mayr, C. Hametner:
"Combustion Engine Test Bed System Identification Under the Presence of Cyclic Disturbances";
IEEE Transactions on Industrial Electronics, 1 (2020).



Kurzfassung deutsch:
In this paper, a new approach for combus-
tion engine test bed system identification is presented.
Modern engine test beds are used to reduce development
time by utilizing them for Hardware-in-the-Loop (HIL) tests.
For highly accurate and close-to-reality tests, model-based
controllers are deployed, which rely on precise models of
the test bed. Hereby, the main challenge is the unknown,
cyclic disturbance of the engine due to compression and
combustion of the air-fuel mixture. These disturbances also
occur when the throttle is kept at zero and have a big impact
on the measurements. Therefore, parameter estimation is
strongly impeded. A method for dealing with these distur-
bances and embed them into the identification procedure
in a two-step fashion is presented in this contribution.
Further advantages of this new approach are that a physical
model is obtained, but prior knowledge of parameters is
not necessary. Moreover, disassembling of the test bed
to perform separate parameter estimation is not needed
and hence, the time consumption of the procedure is kept
low. Experimental results on a 4-cylinder diesel engine test
bed prove the feasibility and effectiveness of the proposed
concepts.

Kurzfassung englisch:
In this paper, a new approach for combus-
tion engine test bed system identification is presented.
Modern engine test beds are used to reduce development
time by utilizing them for Hardware-in-the-Loop (HIL) tests.
For highly accurate and close-to-reality tests, model-based
controllers are deployed, which rely on precise models of
the test bed. Hereby, the main challenge is the unknown,
cyclic disturbance of the engine due to compression and
combustion of the air-fuel mixture. These disturbances also
occur when the throttle is kept at zero and have a big impact
on the measurements. Therefore, parameter estimation is
strongly impeded. A method for dealing with these distur-
bances and embed them into the identification procedure
in a two-step fashion is presented in this contribution.
Further advantages of this new approach are that a physical
model is obtained, but prior knowledge of parameters is
not necessary. Moreover, disassembling of the test bed
to perform separate parameter estimation is not needed
and hence, the time consumption of the procedure is kept
low. Experimental results on a 4-cylinder diesel engine test
bed prove the feasibility and effectiveness of the proposed
concepts.

Schlagworte:
Engine torque modeling, system identification, test bed control.


"Offizielle" elektronische Version der Publikation (entsprechend ihrem Digital Object Identifier - DOI)
http://dx.doi.org/10.1109/TIE.2020.3016262



Zugeordnete Projekte:
Projektleitung Christoph Hametner:
CDL Powertrain Control and Monitoring


Erstellt aus der Publikationsdatenbank der Technischen Universitšt Wien.