Authors: Baitao Xiao; Robert Prucka
Addresses: Clemson University, International Center for Automotive Research, 4 Research Drive, Greenville SC 29607, USA ' Clemson University, International Center for Automotive Research, 4 Research Drive, Greenville SC 29607, USA
Abstract: This research develops a virtual laminar flame speed sensor that can be used for model-based control of multi-fuel adaptive spark-ignition engines. Laminar flame speed is emphasised because it is a primary physical parameter related to ignition timing prediction, and it will change if a different fuel source is used. The observer (virtual sensor) outputs an estimate of laminar flame speed for an arbitrary fuel based on the burn-rate difference to a baseline case, using an inverse quasi-dimensional turbulent flame entrainment model with in-cylinder pressure feedback. A real-time model re-calibration method for laminar flame speed prediction is also proposed. The process is intended to support model-based feed-forward ignition timing control. The method is described in detail and validated experimentally with both E85 and gasoline.
Keywords: fuel properties; virtual sensors; nonlinear observers; spark-ignition engines; flame entrainment modelling; laminar flame speed; multi-fuel engines; spark timing control; closed-loop systems; cylinder pressure feedback; calibration; model-based control.
International Journal of Powertrains, 2014 Vol.3 No.2, pp.154 - 172
Received: 31 Mar 2012
Accepted: 06 Feb 2013
Published online: 21 Jun 2014 *