TU Berlin

Chair of Electronic Measurement and
Diagnostic Technology
Nonlinear System Identification (NoSI)

Page Content

to Navigation

Nonlinear System Identification (NoSI)

Struktur des NoSI Modellansatzes

Due to the shortening in developement time in the automotive industry the requirements for the methods and tools used increase. Especially in engine developement and application methods are sought-after that require little time to get insight into the characteristics of the engine. Model-based methods have proven to be of good use due to their ability to extrapolate.
This stands in stark contrast to the methods like the engine mapping which requires time consuming scans. Another obvious adavantage is the posibility to incorporate the model into the design of the driving cycles, e.g. "Design of Experiments" (DoE).
This reduces the time in the testbed considerably.

The project Nonlinear System Identification (NoSI) focusses its research on the application of data based model structures, e.g. digital filters and neuronal networks, to engine modeling. One area of interest is the question which modeling approaches are suited best for specific parts of the engine. A matter of particular interest is the ability of simulation in real time to be able to put the models to use in 'Hardware-in-the -Loop' simulators. To be able to interpret the results physical structures will be accounted for and integrated into the structure of the model.
Further points of interest are stability and error propagation in the model especially with regard to the combination of the dynamics of the modes of the engine parts.


Quick Access

Schnellnavigation zur Seite über Nummerneingabe