The growing demand for clean energy has intensified interest in alternative fuels like biodiesel. In this work, waste cooking oil (WCO) was utilized as a low-cost, sustainable feedstock for biodiesel production. The produced biodiesel was tested in a two-cylinder diesel engine operating at 1500 rpm under various load conditions, focusing on the B30 blend. To enhance performance, magnesium oxide nanoparticles (nano-MgO) were added to the B30 blend at concentrations of 10, 30, and 60 ppm. Experimental results demonstrated that the B30 blend provided balanced engine performance. The incorporation of nano-MgO enhanced combustion characteristics, leading to improved brake thermal efficiency by (3.52% increase) and lower brake-specific fuel consumption by (11% reduction) with the use of B30 + 60 ppm MgO. compared with untreated biodiesel. Emission analysis revealed a noticeable decline in major regulated pollutants (including a 30% reduction in CO, 26% reduction in HC, 24.3% reduction in NOx, and 33% reduction in smoke opacity with B30 + 60 ppm MgO) driven by the catalytic influence of nano-MgO, which promotes oxygen availability for complete combustion. Additionally, a multi-output artificial neural network (ANN) based on Bayesian regularization was developed to predict engine performance and emissions. The model achieved high predictive accuracy, with an average test coefficient of determination (R²) of approximately 0.9439 across all outputs. Parity and sample-based assessments confirmed its robustness under different operating conditions, validating the developed ANN as an efficient tool for performance prediction and optimization.
