Process Optimization and LSTM Modeling of Nano‐Enhanced Waste Fat Biodiesel Combustion

ABSTRACT Waste‐derived biodiesel offers a sustainable alternative to diesel fuel but faces challenges in combustion and emissions. This study evaluates a CRDI diesel engine fueled with B50 duck waste fat biodiesel blended with 50 ppm Al(NO 3 ) 3 nanoparticles under varying injection timings (21–25°CA bTDC) and injection pressures (500–600 bar). A long short‐term memory (LSTM) model was developed to predict performance and emissions. Under optimal conditions (25°CA bTDC and 600 bar), brake thermal efficiency (BTE) reached 30.85%, whereas smoke, hydrocarbon (HC), carbon monoxide (CO), and nitrogen oxide (NOx) emissions decreased by 17.04%, 20.14%, 27.17%, and 16.8%, respectively, compared with diesel. The LSTM model achieved R 2 values above 0.92 for all outputs. The novelty lies in combining Al(NO 3 ) 3 nano‐additives, injection optimization, and LSTM modeling to improve the combustion and emission characteristics of waste‐fat biodiesel in CRDI engines.