Model Reduction of Activated Sludge Model no. 1 and Bioprocess Models for Identification and Control

Abstract

This paper treats model reduction of the important Activated Sludge Model No. 1 (ASM1) for dynamic modelling of wastewater treatment plants. Section 1 motivates the need for model reduction and gives an overview of reduction approaches in the literature. Section 2 reviews model reduction of activated sludge models, especially ASM1. The review shows that model reduction is often performed quite heuristically and that more insight in the timescale properties of ASM1 is desired. Singular perturbation is a systematic technique for order reduction. Section 3 reviews application of singular perturbation to bioprocess systems to reveal if methods exist to obtain or detect the so-called standard form, which is the difficult part of reduction by singular perturbation. Because existing methods appear insufficiently straightforward, Section 4 develops a method for this task. Three procedures are proposed and tested on a simple continuous general bioprocess model, which yields valuable insight into properties of bioprocess models. A proposed timescale estimation procedure appears a helpful tool in model reduction through timescale separation and provides a good basis for further reduction studies.