Optimal Sensor Placement for High Pressure and Low Pressure EGR Estimation

Abstract

Low pressure exhaust gases recirculation (LP-EGR) is becoming a state-of-the-art technique for Nitrogen oxides (NOx) reduction in compression ignited (CI) engines. However, despite the pollutant reduction benefits, LP-EGR suffers from strong non-linearities and delays which are difficult to handle, resulting in reduced engine performance under certain conditions. Measurement and observation of oxygen concentration at the intake have been a research topic over the past few years, and it may be critical for transition phases (from low pressure to high pressure EGR). Here, an adequate selection of models and sensors is essential to obtain a precise and fast measurement for control purposes. The present paper analyses different sensor configurations, with oxygen concentration measurements at the intake and exhaust manifold and combines observation techniques with sensor models to determine the potential of each configuration. Experimental results from a 2.2 l. diesel engine are used to validate the presented techniques.

Predicting instantaneous engine-out NOx emissions in a real-driving vehicle data scenario

“…Over the six training calibration, the resulting model parameters for the delays and sensor dynamics were comprised of t LP = 0:6060:01s, a NO x = 0:982, and t NO x = 0:3660:01s which agrees with what can be found in literature. 70,76 The original goal of considering different training datasets was to observe the effect of the ambient conditions on the training of the model and its performance once applied on the same validation data (see Figure 6). For each batch in the validation data, the error between the sensor and the model output was calculated: e NO x = z NO x À y NO x .…”

“…Over the six training calibration, the resulting model parameters for the delays and sensor dynamics were comprised of τ LP = 0 . 60 ± 0 . 01 s , a N O x = 0 . 982 , and τ N O x = 0 . 36 ± 0 . 01 s which agrees with what can be found in literature. 70,76…”

“…Accordingly, each EGR flow is made of a different fuel-air ratio, due to the EGR line transport delay, that will affect the product of r egr ϕ in (1). Here, it was decided to assume that the high-pressure recirculation (HP-EGR) was instantaneous, 70 that is, the path between the exhaust and the intake manifolds is expected to be sufficiently short to consider the fuel-air ratio levels at the time. However, the path taken by the LP-EGR gases is traditionally longer, as illustrated in blue in Figure 2.…”