Experimental investigation and artificial intelligence-based prediction of combustion and emission characteristics of DBM15 diesel fuel at variable injection timings in a compression ignition engine

Abstract

This study comprehensively investigates the effects of 15% dibutyl maleate (DBM)-added diesel fuel under different injection timings to reduce environmental impacts and optimize combustion performance in diesel engines. In the experimental studies conducted under constant engine speed (2400 rpm) and 50% engine load, the injection angle was varied between 15 degrees and 27 degrees before Top Dead Center (bTDC). In-cylinder pressure, heat release rate (HRR), and in-cylinder temperature values were analyzed both experimentally and using three different machine learning models: an Artificial Neural Network (ANN), a Convolutional Neural Network (CNN), and a Long Short-Term Memory (LSTM) network. The DBM additive caused a 2.42% reduction in maximum incylinder pressure and an 11.43% reduction in maximum HRR value. At the earliest injection angle of 27 degrees bTDC, in-cylinder pressure increased by 28.7%, while Carbon Monoxide (CO) emissions decreased by 5.9%, Hydrocarbons (HC) emissions by 20%, and smoke opacity by 10.7%. In contrast, Nitrogen Oxides (NOx) emissions increased by up to 79% with advanced injection. Among the models evaluated, the LSTM architecture yielded the most successful results. The Coefficient of Determination (R2) values achieved for in-cylinder pressure, HRR, and in-cylinder temperature were 0.9971, 0.9310, and 0.9970, respectively. The study fills an important gap in terms of both alternative fuels and artificial intelligence-assisted combustion modeling. The findings obtained enable the development of decision support systems for additive use in engine optimization processes.