3.1.17 Transformations

To model the system behavior accurately, the Design of Experiments module make use of transformations of the parameter space to linearize the dependence of the output variables on the transformed input parameters.

Figure 3.5 shows how transformation of the controls are performed in the Design of Experiments module. The control parameters and ranges are mapped in the transformed space where the core algorithm of the DoE module chooses the sample points. Afterwards, the values of the experimental points are transformed back in the real parameter space.

Figure 3.5: Structure of the Design of Experiments program with transformations.

For each input parameter a transformation function can be selected from a set of transformations. This list consists of logarithmic [18], square root, inverse [4], exponential and other special transformations [8]. A complete list of transformations is given in Table D.2.

Some transformation function requires parameters (transformation parameters) which have to be specified in the input file.

These parameters may either be specified explicitly, for example, in the case where a physical formula has been established, or they may be determined from a set of sample points automatically. This mechanism is described in Section 3.2.4.