Machine learning-driven performance and emission optimization of cerium oxide nano-enhanced tobacco biodiesel in CI engines

As fossil fuel reserves continue to diminish and environmental concerns intensify, the search for renewable, sustainable, and cleaner alternatives has gained momentum. Biodiesel derived from tobacco seed oil offers an eco-friendly and non-edible option for compression ignition (CI) engines. In this study, tobacco seed oil methyl ester (TSOME) was blended with diesel at B10, B20, B30, and B40 ratios, each supplemented with 0.5 wt.% cerium oxide (CeO 2 ) nanoparticles to enhance combustion efficiency and emission behavior. A combined framework of experimental evaluation, machine learning (ML) prediction, and response surface methodology (RSM) optimization was employed. Performance and emission characteristics were assessed using a single-cylinder, four-stroke, water-cooled Kirloskar CI engine under varying loads (25–100%). Standard instrumentation measured brake thermal efficiency (BTE), brake specific fuel consumption (BSFC), oxides of nitrogen (NOx), carbon monoxide (CO), hydrocarbons (HC), smoke opacity, and exhaust gas temperature (EGT). Results showed that B10 and B20 blends achieved efficiencies comparable to diesel at higher loads, with