Observer-based estimation of selective catalytic reduction catalyst ammonia storage

Abstract

This paper presents two non-linear observer designs with different robustness to estimate engine selective catalytic reduction (SCR) ammonia coverage ratio. The observers estimate the SCR catalyst ammonia storage (or coverage ratio) based on the available measurements of NO x, NH3, and temperature. An extended Kalman filter is proposed to eliminate the NO x sensor cross-sensitivity to ammonia. A sliding mode observer technique was employed for the design. The robustness of the observers with respect to sensor measurement uncertainties was theoretically analysed. FTP75 test cycle simulation results show the robustness and estimation accuracies of the different observers in the presence of different measurement errors/uncertainties.