Model-Based Control and Diagnostic Monitoring for Automotive Catalyst Systems

An integrated model-based automotive catalyst control and diagnostic monitoring system is presented. This system incorporates a simplified dynamic catalyst model that describes oxygen storage and release in the catalyst and predicts the post-catalyst exhaust gas oxygen sensor signal. The model-based controller maintains the catalyst stored oxygen at a desired operating condition to prevent post-catalyst emissions breakthrough by ultimately adjusting the engine fuelling. Diagnostic monitoring is performed by detecting changes in a test statistic derived from the post-catalyst sensor response. These changes, which are sensitive to both long-term catalyst deactivation effects and short-term emission control device failures, are detected based on the result of nonparametric statistical tests on a moving horizon of past test statistic values.