Experimental validation of a dynamic model for mixture formation in a multi-point injection SI engine

A mean value dynamic model for spark ignition engines with electronic control systems is presented, for the prediction of fuel consumption and emissions during driving cycles, considering air and fuel flow dynamics in the intake manifold and unsteady cylinder wall temperature effects on exhaust emissions. The major results obtained on mixture strength excursions in transient operation and on the possibility of designing model-based compensation strategies are reviewed, as well as the results of optimization analysis for both deterministic and stochastic cases. The results of an experimental investigation on a multi-point injection engine are presented, in order to validate air and fuel flow sub-models, using a step response technique for throttle opening and injection time. For the air model, based on a mean value filling and emptying approach, a good agreement between predicted and measured data is obtained. A least square technique has been used to identify the parameters of the fuel submodel, which predicts the observed values with a good accuracy, consistent with the objectives of the whole engine dynamic model.