
Often the actual parameters for the different dispersive models (Debye, Lorentz, Drude, and the combination of Drude-Lorentz) are not known for a given material. However, the values of Relative Permittivity (εr) and Electrical Conductivity (σ) might be known at different frequencies. Due to the complexity of the dispersive models it is necessary to use an iterative optimization process to fit the parameters of a given model to the εr(f) and σ(f) curves of a material.

| Based on Genetic Algorithm |
| Known values entered directly by the user or imported from a text file |
| Plot entered data |
| Four different disersive models |
| First pole initial aproximation by linear distribution |
| Once the optimization has been completed the ε_stat, ε_inf and σ_E values are directly available to be used in SEMCAD X. |