Authors: M. Anton, S. Bechtold, O. Nüssen and R. Laur
Affilation: University of Bremen, Germany
Pages: 362 - 367
Keywords: behavioral modeling, curve fitting, system identification, nonlinear least squares, model synthesis.
When creating mathematical models of microsystem components or microelectronic devices, analytical and experimental modeling concepts can be distinguished. Analytical modeling, often performed on process or device level, can only be applied if the equations and geometries describing the components' behaviour are known. If there is none or insufficient information about the component to be modeled, tir:e ansfer characteristics may be determined by using methods of system identification and curve fitting from given input and output characteristics. These methods can be applied to typical domains found in microsystem technology. In this paper a software package is presented which enables the user to generate behavioural models from transfer characteristics given by measurements, simulations or data sheets.