Katie Vantine

ANALYSIS OF AN EVENT-BASED DIESEL ENGINE MODEL FOR CONTROL PURPOSES

Abstract A diesel engine model with strokes (combustion events) as the independent variable is compared to the more traditional time-based model. The structural differences between the models result in different open- and closed-loop behavior. The variations and expected robustness for several transfer functions is investigated in the frequency domain. It is concluded that for SISO engine speed controllers, the time-based model should be used while for applicationsinvolving flow, e.g., EGR and VGT controllers, or if the controller is to be implemented with a constant sampling angle rather than sampling time, the event-based model is preferred.

A wavelet-based combustion noise meter

The combustion noise of diesel engines can be assessed by filtering in-cylinder pressure traces. In this paper, two wavelet-based methods are discussed. It is shown that the statistical properties of a noise meter based on shift-invariant wavelet transforms are improved compared to one that is based on traditional (shift-variant) wavelet transforms. The application of noise metering for pilot injection detection is also presented