POST PROCESSING

Once data has been obtained, whether from an optimization or DOE, or from data importation, the user can turn to the extensive post-processing features in modeFRONTIER to analyze the results. The software offers wide-ranging toolbox, allowing the user to perform sophisticated statistical analysis and data visualization.

WIZARD TOOLS:

NEW ► CAP (Computer Aided Principle)

DOWNLOAD ► modeFRONTIER technical brochure | Post Processing

activeVIEWER is a new module able to open modeFRONTIER project files, view charts and check tables and functions from modeFRONTIER DesignSpace without having a modeFRONTIER's license installed locally.

Statistical analysis

Statistical tools can be used to analyze distributions, coming from experiments or from DOE, to analyze how stable is a solution in the event of uncertain fluctuations of inputs (Robust Design).

Statistical tools can be used to find correlations, in particular which input variables have most influences in the system outputs; these results can be obtained from an available database, or from DOE.

Is the DOE set up correctly for the analysis we need?
Which are the most important design variables?
Can we reduce the space between variables?
What is the best design space region to address for the optimization process?
What is a reasonable number of objectives or constraints to be defined?
What is the effect of the variations of input variable on the outputs?
Are there interaction effects between variables? Are local behaviours different from the global ones?
Is the solution stable in the event of small fluctuations of input variables?

Best practice for Statistical Analysis

Define a proper DOE: correlations between input variables should be as low as possible.
When DOE is run, use Statistical Analysis to find the most significant variables, reducing variable space or the number of objectives
In the reduced variable space, optimization could be run.