P Vanrolleghem

Practical Identifiability of a Biokinetic Model of Activated-sludge Respiration

This paper deals with the estimation of the parameters of the Monod model for the activated sludge process on the basis of oxygen uptake rate data only. The objective of the paper is to concentrate on the practical identifiability properties of the model and on the design of informative experiments for parameter estimation. The results are illustrated by experimental data. Improvements in parameter estimation accuracy with a factor 2 can be obtained by small extensions of the respirometric experiments. The optimal experimental design procedure for parameter estimation (OED/PE) can be implemented in the respirographic sensor.

Structural Identifiability of Biokinetic Models of Activated-sludge Respiration

This paper deals with the identifiability of parameters of kinetic models describing the activated sludge process. The main concern of the paper is to present important aspects of the structural identifiability properties. The identifiability analysis is based on the availability of only on-line oxygen uptake rate data (given by a novel respirographic biosensor). Four model candidates (exponential, Monod, double Monod and modified IAWQ No. 1) are considered. Two different methods (Taylor series expansion, and transformation of the nonlinear model into a model linear-in-the-parameters) are considered, their advantages and drawbacks are illustrated with the four kinetic models. For each model it is found that only a smaller set of the original parameters are structurally identifiable on the basis of oxygen uptake rate data only.