Smart Sensing and Decomposition of NO_X and NH_3 Components from Production NO_X Sensor Signals

摘要

Production NO_X sensors have a strong cross-sensitivity to ammonia which limits their use for closed-loop SCR control and diagnostics since increases in sensor output can be caused by either gas component. Recently, Ammonia / NO_X Ratio (ANR) perturbation methods have been proposed for determining the dominant component in the post-SCR exhaust as part of the overall SCR control strategy, but these methods or the issue of sensor cross-sensitivity have not been critically evaluated or studied in their own right. In this paper the dynamic sensor direct- and cross-sensitivities are estimated from experimental FTIR data (after compensating for the dynamics of the gas sampling system) and compared to nominal values provided by the manufacturer. The ANR perturbation method, and the use of different input excitations are then discussed within an analytical framework, and applied to experimental data from a large diesel engine. The method successfully discriminates between NO_X and NH_3, in the post-SCR exhaust, and is extended to provide an estimate of the composition in situations where simultaneous NO_X and MH_3 emissions occur. The results are validated against the previously processed and independent FTIR data.