(369b) Simultanuous State Reconstruction and Parameter Identification Using Wavelets

Preisig, H. A., Norwegian University of Science and Technology (NTNU)

Wavelets can be used to obtain multi-scale information about the process from measured input/output signals. Beta-spline multi-wavelets provide not only scale-filtered signals, but also their derivatives to a certain order. The maximal order is defined by the parameters of the used beta-spline.

A gas-phase reaction is taken as and example discussing: - reconstruction of the state being the partial pressure or the component concentration based on total pressure measurement. - identification of the kinetic parameter - arguments to choose or not to choose input signals - effect of unmodelled sensor dynamics - filtering properties - the mathematical problem and structure of simultaneous observer and identification problem, in particular what is linear and what is nonlinear - the separation of linear and nonlinear parts is closely related to the sequential identification problem in reactive systems (reconstruction of the reaction rate from the experiments, regularisation and parameter identification.