State-of-the-art control strategies for spark ignition engines take into account the oxygen storage level of the indispensable three-way catalyst (TWC) by a model-based approach. Owing to the limited onboard processing power, the models used are simplified compared to the actual system dynamics. Their accuracy can be improved by an additional direct state observation, e. g., with microwave measurement systems in which the catalyst state is inferred from measured cavity resonance curves. However, the accuracy of time-invariant models deteriorates with time because of catalyst aging. We have presented a promising concept to obtain more reliable TWC state information by post-processing of various sensor output data (including microwave measurement data). This post-processing yields a corrected air-fuel ratio value and the oxygen storage capacity. Based on this information, a classical TWC model is able to generate more accurate information about the TWC state and in particular the relative oxygen storage level. The tests performed have shown that this approach is able to reduce measurement errors. Furthermore, the processing of the output values of the common upstream lambda probe and the microwave measurement system might also pave the wave for a novel on-board diagnostics approach. It would allow to detect TWC aging or damage without the need of further lambda probes. It is emphasized that the signal processing approach discussed is on-board and real-time capable.
Real-time capable parameter tuning for exhaust gas aftertreatment models in automotive applications
2013
6 Seiten, 7 Bilder, 1 Tabelle, 9 Quellen
Aufsatz (Konferenz)
Datenträger
Englisch
SAE Technical Papers | 2011
|Thermomechanical Behavior of an Automotive Exhaust Aftertreatment Application
British Library Conference Proceedings | 2022
|