Search Results

Among the existing concepts that help to improve the efficiency of spark-ignition engines at part load, Controlled Auto-Ignition™ (CAI™) is an effective way to lower both fuel consumption and pollutant emissions. This combustion concept is based on the auto-ignition of an air-fuel-mixture highly diluted with hot burnt gases to achieve high indicated efficiency and low pollutant emissions through low temperature combustion. To minimize the costs of conversion of a standard spark-ignition engine into a CAI engine, the present study is restricted to a Port Fuel Injection engine with a cam-profile switching system and a cam phaser on both intake and exhaust sides. In a 4-stroke engine, a large amount of burnt gases can be trapped in the cylinder via early closure of the exhaust valves. This so-called Negative Valve Overlap (NVO) strategy has a key parameter to control the amount of trapped burnt gases and consequently the combustion: the exhaust valve-lift profile.

A local Gaussian process regression approach is presented, which allows to model nonlinearities of internal combustion engines more accurate than global Gaussian process regression. By building smaller models, the prediction of local system behavior improves significantly. In order to predict a value, the algorithm chooses the nearest training points. The number of chosen training points depends on the intensity of estimated nonlinearity. After determining the training points, a model is built, the prediction performed and the model discarded. The approach is demonstrated with a benchmark system and air charge test bed measurements. The measurements are taken from a turbocharged SI gasoline engine with both variable inlet valve lift and variable inlet and exhaust valve opening angle. The results show how local Gaussian process regression outmatches global Gaussian process regression concerning model quality and nonlinearities in particular.

Air charge calibration of turbocharged SI gasoline engines with both variable inlet valve lift and variable inlet and exhaust valve opening angle has to be very accurate and needs a high number of measurements. In particular, the modeling of the transition area from unthrottled, inlet valve controlled resp. throttled mode to turbocharged mode, suffers from small number of measurements (e.g. when applying Design of Experiments (DoE)). This is due to the strong impact of residual gas respectively scavenging dominating locally in this area. In this article, a virtual residual gas sensor in order to enable black-box-modeling of the air charge is presented. The sensor is a multilayer perceptron artificial neural network. Amongst others, the physically calculated air mass is used as training data for the artificial neural network.

The optimization of fuel efficiency and the minimization of the residual gas fraction require individual cylinder control of the amounts of inducted air mass and injected fuel mass. Determination of an individual cylinder air/fuel ratio (AFR) regulator is based on the measured AFR for each cylinder, using 4 proportional UEGO sensors. The innovative character of this study describes a unified and robust individual cylinder AFR estimator, using a single measuring point: a proportional oxygen sensor located in the exhaust manifold. The model used for the estimator is a state model such that the dimension of the state and measurement matrices are unique, whatever the manifold configuration and the sensor position (confluence point or exhaust manifold: unified model), the engine speed (robust model).