Predictive modelling for enhances petrol engine reliability with petroleum and hydrogen fuels mixture

Abstract

Internal combustion engines that use a mixture of hydrogen and gasoline are designed to take advantage of the advantages of both fuels. These hybrid systems can reduce emissions compared to conventional gasoline-only engines. The integration of hydrogen can improve the overall efficiency and performance of the combustion process. In this research a comprehensive statistical modeling approach for ecological and energy gaining parameters was performed. Statistical models were trained using 11 input parameters such as fuel mixture level, petrol mass, petrol energy ratio, ethanol mass, ethanol energy ratio, hydrogen mass, hydrogen energy ratio, total input energy, ratio of pure carbon, oxygen, and hydrogen. Output parameters were n, CO, CO2, lambda, O2, HC, NOx, pi, ISFC, and BTE. The best accuracy for the training set was achieved for n, CO2, lambda, Pi, BTE, and ISFC with MAPE of 0.45%, 1.29%, 0.55%, 0.85%, 2.02%, and 2.48%, respectively. Developed prognostic models were also validated using test dataset and reached valuable accuracy for n, CO2, lambda, Pi, BTE, and ISFC with MAPE of 1.67%, 0.97%, 2.66%, 2.56%, 2.61%, and 5.39%, respectively. In this study the developed predictive statistical models achieved high accuracy in predicting the ecological and energetic parameters of engine emissions generated by engines powered by a petrol-hydrogen mixture.