Enhancing engine performance and sustainability: gold nanoparticles and machine learning for biodiesel optimization in compression ignition systems

Nanotechnology and machine learning are transforming energy systems by enhancing engine efficiency and sustainability. The integration of advanced nanomaterials, such as gold nanoparticles (AuNPs), with predictive modeling offers new opportunities for optimizing biodiesel performance in compression ignition (CI) engines. This study aims to develop and evaluate a novel biodiesel blend using 50% waste cooking oil and 50% Simarouba oil, enhanced with gold nanoparticles, and to apply machine learning for performance prediction and optimization in CI engines. Gold nanoparticles were synthesized from a plant extract and characterized by UV–visible spectrophotometry, confirming their presence at an absorption peak of 650 nm. Biodiesel blends (B20, B40, B60, and B80) were prepared and tested in a single-cylinder CI engine at various compression ratios (14:1, 16:1, and 18:1) and loads. Key performance parameters, including brake thermal efficiency (BTE) and brake specific fuel consumption (BSFC), were measured. An Extreme Gradient Boosting (XGBoost) machine learning model was trained on the experimental data to predict engine performance. The addition of AuNPs to biodiesel blends resulted in significant performance improvements: BTE increased by up to 6.57% for B20 at a 14:1 compression ratio, and BSFC decreased by up to 9.17% for B40 at a 16:1 compression ratio. The XGBoost model accurately predicted BTE and BSFC, with maximum errors of 4.17% and 3.53%, respectively. Gold nanoparticle-enhanced biodiesel blends offer improved CI engine performance and fuel efficiency. The use of XGBoost enables the reliable prediction of key performance metrics, reducing experimental costs and accelerating optimization. This integrated approach supports the development of sustainable, high-performance biofuels and advances the application of machine learning in energy systems.