Prediction of performace parameters using artificial neural network for 4-stroke C I engine fueled with esterified neem oil and biogas

Experiments on the technical feasibility of the use of esterified neem oil and biogas under dual fuel mode has been carried out in a 3.7 kW diesel engine. Esterified neem oil (ENO) blends and bottled biogas was used for the experimentation. Biogas is directly mixed with the inlet air in the induction manifold and tests were carried out for varying loads and pressures. The performance parameters obtained from test were within the acceptable range for ENO and biogas combination. Mechanical efficiency was improved and ENO consumption was reduced due to higher biogas introduction; however, a distinguishable noise was noticed when further increase in supply of biogas. Artificial neural network (ANN) technique was also applied for prediction of thermal efficiency and BSFC of the engine. The input parameters for ANN method were speed, gas flow rate, temperature, total fuel consumption, and load. It was seen that ANN predicted performance values were close to experimentally obtained data when 90% of data were in training set. (paper)