Prediction of Gas Concentrations in a Three-Way Catalyst for On-Board Diagnostic Applications

The process of controlling tailpipe emissions leads to the need to understand the dynamic behaviour of the after-treatment devices. The model provides the basis for design prediction, on-line diagnosis and real time control. Although a number of models have been presented in the literature, their efficient performance continues to require further development and validation to meet increasingly demanding requirements.

Models have been developed that use the basic physical framework including thermal behaviour, fluid mechanics and basic chemistries. As more demands are placed on models, more phenomena need to be taken into account and in particular, progressively more of the chemistry of the Three-Way Catalyst (TWC) itself.

In this paper we present a black-box model for a three-way catalytic converter that has been developed and tested using real experimental data. The objective of the modelling exercise is a real time prediction of emission concentrations for both control and On-Board Diagnostics (OBD). The model has two components with similar structures. The first component is used for post-catalyst air-to-fuel ratio (AFR) prediction, whereas the second uses pre-catalyst AFR measurement and the predicted post-catalyst AFR to estimate individual gas concentrations.