Program system SIDYS: Simulation and parameter identification of dynamic systems

J. Benz, J. Polster, R. Bär, G. Gauglitz

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

We describe a program SIDYS for simulation and parameter identification of complex dynamic systems. SIDYS is characterized by a combination of modeling, data handling, simulation and identification. Either continuous or time discrete processes can be simulated by SIDYS; a choice is offered among Euler, Runge-Kutta-Merson and Gear for the integration of systems of differential equations. Least squares or weighted least squares can be used for parameter evaluation. The nonlinear minimization can be carried out by random search, Rosenbrock, quasi-Newton, or lattice search. An optimal analysis of the residuals allows the quality of the parameter identification to be judged, thereby revealing specification errors in the model programming. The basic flow diagram of SIDYS is described. The use of SIDYS is illustrated by a spectroscopic-kinetic analysis of a reversible photoreaction of dihydroindolizines accompanied by thermal reaction. The evaluation shows that a differential absorbance equation can be used with a noisy experimental data set only if the rate constant of the thermal reaction is taken as known. An analysis of the residuals is provided and the significance of the parameters is discussed.

Original languageEnglish
Pages (from-to)41-48
Number of pages8
JournalComputers and Chemistry
Volume11
Issue number1
DOIs
StatePublished - 1987

Fingerprint

Dive into the research topics of 'Program system SIDYS: Simulation and parameter identification of dynamic systems'. Together they form a unique fingerprint.

Cite this