************************************
* Identification Toolbox 1.1 beta **
************************************
Identification Toolbox is used to construct mathematical models of LTI systems
from measured input-output data sequences. There is also possible to combine the
identification based on measured input-output data with first principles
approaches.
Whereas, the tool is developed at the educational and research institution, it
serves primarily the new approaches and the development is focused particularly
in the innovations and testing.
The tool allows:
- Preprocessing of signals (resampling, filtering, detrending,...)
- Identification of LTI systems
- Validating of constructed models
Supported Models:
- ARX
- ARMAX
- State-space
Supported Methods:
- Prediction Error Method (PEM)
- Subspace State-Space System IDentification (4SID)
- Combining 4SID + PEM
- Errors in Variables (EIV) identification
New features:
- Use of Nonnegative Garrote Method (NNG) for order selection (reduction) of
ARX/ARMAX models
- Handling Certain Structure Information in 4SID Identification
- Handling Structure Information in Black-Box (BB) Modeling
- Estimating Physical Parameters of a System by BB
Future extensions:
- Graphical User Interface
- Other LTI models
- Other identification methods
