Modelling of SI Engine Performance and Emissions with Ether Additives and Low-Octane Gasoline Using Artificial Neural Network

Fuel quality is an important indicator for the suitability of certain fuel to operate SI engine efficiently. Additives are one of the most viable methods to improve low-octane gasoline fuel properties. In the current study, Methyl Tert-Butyl Ether (MTBE) has been used with low-octane gasoline fuel as an octane booster at 5%, 10%, and 15% additive ratio to evaluate the engine performance and emissions using Artificial Neural Network (ANN) Model. Based on a widely used back-propagation learning method, the engine model was created. Engine speed, engine load, and blend ratio have been chosen as the three inputs for the ANN that would forecast the output during the internal building of the back-propagation method. The main component of the ANN is the activation function, and each input has a distinct function depending on the kind of data population (linear or non-linear). The model's output, which is controlled by the prediction process, includes power, torque, and emissions. For each anticipated value, there is a desirable set of input variables that will make the engine function optimally. The findings of the best ANN model show that the created model can accurately predict multiple engine parameters under various circumstances at the appropriate R-value of speed, load, and blend ratio for training, validation, and testing, respectively. As a result, the ANN model's performance and accuracy are adequate.