Design of Engine-Out Virtual NO _x Sensor Using Neural Networks and Dynamic System Identification

Fuel economy improvement and stringent emission regulations worldwide require advanced air charging and combustion technologies, such as low temperature combustion, PCCI or HCCI combustion. Furthermore, NO_x aftertreatment systems, like Selective Catalyst Reduction (SCR) or lean NO_x trap (LNT), are needed to reduce vehicle tailpipe emissions. The information on engine-out NO_x emissions is essential for engine combustion optimization, for engine and aftertreatment system development, especially for those involving combustion optimization, system integration, control strategies, and for on-board diagnosis (OBD).

A physical NO_x sensor involves additional cost and requires on-board diagnostic algorithms to monitor the performance of the NO_x sensor. A control-oriented engine-out NO_x model that meets throughput constraints is desirable for on-board applications as a virtual NO_x sensor, which could be used in replacement of the physical sensor or run parallel as an analytical redundancy to detect potential in-range failure for the physical sensor.

This paper presents the design of a virtual NO_x sensor using neural networks combined with nonlinear dynamic system. NO_x emission estimation with or without cylinder pressure information is investigated, and a correction to a baseline NO_x model is proposed in order to compensate transient EGR responses. The developed virtual NO_x sensor demonstrates a higher level of fidelity estimation when compared with actual NO_x measurements.