On-line Identification of Fuel Dynamics for a Model-based Injection Control

Abstract

Literature showed quite clearly that the efficiency of Air to Fuel Ratio (AFR) control for Spark Ignition (SI) Internal Combustion Engines (ICE) strongly depends on its capacity to deal with the fuel-flow phenomena inside intake manifolds. Moreover, engine performances (such as power output, specific fuel consumption, and exhaust gas emissions) are directly related to the efficiency of the combustion process, which, on its turn, can be affected substantially by the air/fuel ratio variations related to the fuel-film dynamics. In this work a comprehensive model-based air/fuel ratio control technique is proposed: this is based on a dynamical model of the air dynamics inside inlet manifolds and on the online identification of the fuel-film parameters. Here the identification procedure is illustrated in detail and validated basing on experimental data regarding a single-cylinder engine. In order to demonstrate the method validity, in fact, a single-cylinder research engine (type AVL 5401) equipped with a port- fuel injection system has been experimented using a dynamic test bench. The first-order model of Aquino has been used as a basis to form the required algorithm for the identification process. The required input variables are the amount of fuel injected and that of the air inducted in the cylinder. This latter value is not obtained (as usual) by a look-up table stored in the ECU, but is calculated within each engine cycle by a mathematical model of the air dynamics developed by the same authors. The model used is named Model Of Interconnected Capacities (MOIC) and has been previously presented and separately validated.