Use this URL to cite or link to this record in EThOS:
Title: Nonlinear dynamic modelling and identification of a three-way catalytic converter
Author: Soumelidis, Michail I.
ISNI:       0000 0001 3472 366X
Awarding Body: University of Sussex
Current Institution: University of Sussex
Date of Award: 2004
Availability of Full Text:
Access from EThOS:
Increasingly stringent automotive exhaust emissions legislation demands both advanced catalyst control for super ultra low emission vehicle (SULEV) requirements, and close monitoring of catalyst performance. Modem monitoring and control strategies for three-way catalysts (TWC) are usually based on simplified oxygen storage models. Because such control-oriented models need to be simple in structure, they often fail to capture significant dynamic phenomena that take place inside the TWC. Although several extended oxygen storage models have been recently proposed, often compromising their original simple structure, most of them still appear incomplete. In the study presented in this thesis, the limitations of a conventional control-oriented TWC model, based on the dynamics of oxygen storage, were first investigated. Model validation tests, based on an extensive range of experimental data, suggest that inclusion of the basic kinetics of the governing TWC reactions in some form can be beneficial during the modelling process. A simple chemical model was developed to assist the identification process. The model proved to be accurate and robust in estimating the oxygen storage process and predicting the post-catalyst concentrations of some of the main exhaust gas components. Nonlinear modelling and identification techniques were then employed in order to explore alternative means of modelling the strongly nonlinear dynamic behaviour of TWCs. The polynomial NARMAX model was selected to describe the TWC dynamic system and both its structure and parameters were identified in real-time. Validation tests demonstrated the superior performance of the NARMAX model compared to the oxygen storage model, in predicting the post-catalyst air-fuel ratio (AFR) over a wide range of operating conditions. A nonlinear catalyst model was finally proposed, incorporating four NARMAX models, each optimised for local prediction. The chemical model was ' used to identify the current operating region and select the appropriate local model for prediction. The proposed dynamic model is simple in structure, only requires knowledge of the upstream/downstream AFR values, and could form the basis of an on-board catalyst monitoring and control system
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available