AN ESTIMATION OF IN-CYLINDER PRESSURE BASED ON LAMBDA AND ENGINE SPEED IN HCCI ENGINE USING ARTIFICIAL NEURAL NETWORKS

Abstract

In this study, the in-cylinder pressure predicted based on lambda and engine speed with the ANN method for HCCI engine. In-cylinder pressures obtained at different lambda and engine speeds, constant inlet air temperature (80 degrees C), RON40 (40% iso-octane/60% n-heptane) fuel in a single cylinder, four-stroke, naturally aspirated, port injection HCCI engine. MATLAB ANN program was used for training, validation and testing ofinputs. The crank angle, engine speed, and lambda were used as input values and the in-cylinder pressure was used as the target value. The Levenberg-Marquardt training algorithm was used for the training of inputs. Also, three layers and 10 neurons were used for the training process. The best validation performance was obtained at epoch 535 as 0.000043691 MSE value. The correlation factor of training, validation, and testing between the targets to outputs were obtained at 0.99912, 0.99905 and 0.99893 respectively. The total correlation factor was found at 0.99908. It is observed that there is a high degree of accuracy between the estimation of results and experimental data using the developed ANN model.