Identification of model structure via qualitative simulation

Different approaches and programs for qualitative simulation are briefly reviewed to illustrate the characteristic features of QUALSIM, an object-oriented package for supporting the identification of the structure of dynamic systems. The package is based on a qualitative interface and a stochastic calculus using variables´ ranges defined in the real number space. The intuitive way of defining variables and connections and the possibility of avoiding the definitions of equations linking the variables make it possible to use the package for fast model prototyping and as an educational tool. The identification of a model of lake eutrophication is shown as an example