Performance & emission analysis of HHO enriched dual-fuelled diesel engine with artificial neural network prediction approaches

Abstract

Most of the studies on conventional fuel types that can be used in internal combustion engines have been made in order to improve performance values. Nowadays environmental problems have shown that emission values are more important and interest in low carbon alternative fuels has highly increased in recent years. In this study, performance and emission values of soybean biodiesel (B25) fuel mixture used in diesel engine were investigated in detail by making different ratios of hydroxy (HHO) enrichment (3, 5 and 7 L/min). HHO enrichments increased brake torque and power outputs with direct correlation to flow rate amount; at the same time brake specific fuel consumption has decreased. Also, one of the main objectives of this study is to predict the optimum hydrogen requirement against performance reductions and NOx formations among test fuels (3, 5, and 7 L/min HHO enriched B25), too by using artificial intelligence. For developing the ANN structure, Levenberg-Marquardt (LM) learning algorithm was used to adjust the weights in the cascade forward network. The results show that the ANN model has 95,82%, 96,07%, and 92,35% estimation accuracies for motor torque, motor power, and NOx emission, respectively.