Abstract
The paper presents the main steps of formulating a semi-empirical computational model of the working cycle of an engine. Parameters of the model are identified on the basis of experimental measurements carried out for a CI engine supercharged by a turbo compressor with direct injection equipped with an electronically controlled Common Rail system. Measurements were performed for different engine load and possible recirculation degrees under various operating conditions. The identification process is formulated as a dynamic optimisation task. Genetic algorithms are used in order to obtain the values of model parameters which ensure the minimal difference between the measured and calculated pressure courses in the cylinder. Next, in order to generalize the results of identification, an approximation task was formulated and solved by using an artificial neural network. In order to verify the accuracy of the model, some additional experimental measurements were carried out. Results obtained form the model in comparison with the validation measurements show that an acceptable accuracy of the model was achieved.
Keywords: CI engine modelling, semi-empirical model, optimisation, experimental tests