Reduction of transient NOx emissions based on set-point adaptation of real-time combustion control for light-duty diesel engines

Abstract

This paper proposes a set-point adaptation algorithm for combustion control to reduce transient NOx emissions. The set-point adaptation algorithm is activated only when an NOx emission spike is detected. For the detection of NOx emission spikes, an empirical NOx prediction model was developed. The NOx prediction model was derived based on three key parameters associated with the mechanistic formation of NO: the combustion temperature, mixed gas composition, and combustion duration. The predicted results of the developed NOx model showed strong linear relation with the measured NOx emissions, with a root mean square error of 12.92 ppm and an R2 of 0.98. Using the proposed NOx prediction model, transient NOx emission spikes were detected and the set-point adaptation algorithm was activated to reduce transient NOx emissions. The set-point adaptation algorithm retarded the combustion phase under such conditions, and the experimental results showed that transient NOx emission spikes were reduced by the proposed set-point adaptation algorithm by about 25%.