2.7 Design of Experiment (DOE)

To run an optimization, we need a first set of models (designs), from which optimization algorithms start the optimization process.
modeFRONTIER supplies different types of algorithms to generate an effective Design of Experiments (DOE).

The DOE component has been introduced in the Work Flow desktop in section 2.4, as the starting point of the logic flow of the process.
To enter the DOE Properties window, double click on its icon or RMB click and select Edit Node Properties, as usual.

There is a list of possible DOE, providing different ways to develop the DOE and they are:

• User Sequence: based on a custom user sequence, the user must specify the number of designs and provide the values for each design.
• Random Sequence: random values will be generated for the number of required designs.
• Sobol: based on a pseudo-random Sobol sequence, it assures a uniform distribution of the starting values.
• Monte Carlo: based on a Gaussian distribution process.
• Full Factorial: this is a systematic series of experiments. In case of large numbers of variables, a lot of cases are generated: nlevel^ninputs.
• Reduced Factorial: based on a 2 level reduced factorial sequence.
• Cubic Face Centered: based on the Cubic Face Centered method, is equivalent to a full factorial with 2 levels plus the center of each hyper face.
• Box-Behnken: based on the Box-Benhken method.
• Latin Square: based on the Graeco-Latin square method.
• Taguchi Matrix based on Taguchi orthogonal array.
• Plackett Burman based on Plackett-Burman orthogonal array.
• D-Optimal based on D-Optimality criteria.
• Cross Validation based on the Cross Validation method applied to the Kriging algorithm.

Once created, the table can be manually modified by editing the values. Operations as cut , copy , paste are possible, besides the values can be imported from a file as well.

Fig. T2.7.1 shows a case with 10 initial designs generated using the Sobol algorithm.

Fig. T2.7.1: a Sobol generated DOE.

The rows outlined in yellow are not feasible designs, because they do not fulfil the constraint BH > 0: to point out this fact, the pertinent cell is highlighted in orange

When you want to import the values from a file (usually named *.des), be sure it is an ASCII file and that the data are arranged in a table as the one below:

          <ID>   H                    L   S                    T     B    F  Wc     Mc     Young
        000000   8  2.652352352352e+001  10  3.002002002002e+001     8  150   1  0.001  2.1e+005
        000001   4  1.476176176176e+001  10  2.001001001001e+001   4.5  150   1  0.001  2.1e+005
        000002  12  3.828528528529e+001  10  4.003003003003e+001  11.5  150   1  0.001  2.1e+005