Application of artificial intelligence to analyze the performance-emission characteristics of a single cylinder, four-stroke spark ignition engine fuelled with alcohol gasoline blends: An ANN approach.

This study explains the application of Artificial Neural Network (ANN) for analysing emission and performance attributes of a four-stroke air-cooled single-cylinder, spark ignition (S.I.) engine operating with ethanol and higher alcohols gasoline blends. These Blends were assessed according to their physical and chemical properties, including calorific value (CV), Research Octane Number (RON), density, and oxygen concentration, etc. Experimental tests were performed to measure Brake Specific Fuel Consumption (BSFC) and Brake Thermal Efficiency (BTE) along with emission of Unburnt Hydrocarbon (UHC), Carbon Monoxide(CO), Carbon Dioxide (CO2), and Nitrogen Oxide (NOx). This ANN model demonstrated high predictive accuracy with a correlation coefficient within a range from 0.93 to 0.99, with low values of the Root Mean Square Error (RMSE). These ANN predictions closely matched the experimental results. This study concludes that ANN modeling can be a robust tool for optimizing fuel blends in S.I. engines, offering a sustainable approach to improving fuel efficiency and lowering emissions.