Comparative Analysis of Engine Performance, Noise Emissions, and Energy Efficiency of Euro Diesel and Safflower Methyl Ester Fuels Using Artificial Neural Networks

In this study, energy analysis, engine performance, and noise emission tests of Euro diesel and safflower methyl ester fuels were conducted on a diesel engine. The experiments were conducted independently for each fuel type across engine speeds ranging from 1000 to 2400 rpm, and the physicochemical properties of the fuels were characterized and evaluated through engine testing. Noise emission values were recorded from four different points around the engine at a distance of one meter and were compared with those of reference diesel fuel. According to the test results, the most suitable fuel type was determined based on engine performance. noise emission and energy analysis. In this study, modeling was performed using artificial neural networks (ANNs) based on experimentally obtained data, and the noise emission characteristics of B100 and D100 fuels were analyzed. Both raw and normalized datasets were evaluated to assess the predictive accuracy of the models. It was concluded that the predictive success was closely associated with the choice of training algorithms and transfer functions utilized. The findings highlight that selecting suitable models and algorithms tailored to the structure of the dataset plays a critical role in enhancing prediction accuracy.