Robust Mixed H_2/H_∞ Gain-Scheduling Observer Design for Removal of NO_X Sensor Ammonia Cross Sensitivity in Selective Catalytic Reduction Systems

摘要

This paper focuses on the robust mixed H_2/H_∞ observer design problem for removal of NO_x sensor ammonia cross-sensitivity in selective catalytic reduction (SCR) systems. Due to the ammonia cross-sensitivity, the reading of the NO_x sensor is contaminated severely. In order to utilize the concentration of the NO_x in the downstream tailpipe, it is desired to design an observer to estimate the actual value of the NO_x concentration and the time-varying cross-sensitivity factor. It is assumed that the variation of the cross-sensitivity is bounded. A design approach of the gain-scheduling robust mixed H_2/H_∞ observer is proposed, in which the gain-scheduling strategy is adaptive to compensate for the nonlinearity and the time-varying parameters in the SCR model. Simulations and comparisons are carried out to show the efficacy and the advantages of the proposed approach over the existing extended Kalman filter on a Diesel engine powertrain.