Next: 4.5.4 Example of a
Up: 4.5 Integration into a
Previous: 4.5.2 Fault Recovery
4.5.3 SIESTA Optimizer Module
The SIESTA framework includes a module for optimization tasks.
In this module a complete optimization loop with model evaluation can
be performed.
For definition of an optimization the following data is required:
- Model:
The model function is a series of various simulators and data
manipulations. The result of the model is the target
function4.8.
- Parameter ranges:
For a successful optimization a good start value and the minimum and
maximum values of the parameters have to be supplied to the
optimizer.
- Settings for the optimizer:
For the optimizer the information of the termination condition is
defined.
All these data are given to the SIESTA framework by the experiment
file. An example of a calibration task is given in
Section 4.5.4 and for an optimization task with nonlinear
constraints in Section 4.5.5.
Footnotes
- ... function4.8
- For the nonlinear optimizer also the inequality
constraints and equality constraints are results of the model
R. Plasun