Using QR to solve quantitative modeling problems : an application to intracellular thiamine kinetic s

Recent works carried out within the Qualitativ e Reasoning (QR) research framework are centred on th e exploitation of QR techniques to address the proble m of quantitative system Identification (SI) with the goa l to enhance the overall process, namely the selection o f a proper model identifier and the parameter estimatio n procedure . Traditional Si, both parametric and nonparametric, may be really problematic for those applica tion domains, such as the medical/physiological one, o f which either the available knowledge is incomplete o r the structural model is not identifiable or the observe d data are poor in number and in quality . This paper deal s with the application of an hybrid method, which build s a fuzzy system identifier upon a qualitative structura l model, to solve identification problems of the intracel lular kinetics of Thiamine (vitamin Br ) . The model ob tained is not as much informative as a purely structura l one but robust enough to be used as a simulator, an d then to provide physiologists with a deeper understand ing of the Thiamine metabolism in the cells .