Prediction of RCCI combustion fueled with CNG and algal biodiesel to sustain efficient diesel engines using machine learning techniques

Abstract

This study used microalgae biodiesel as a high-reactive fuel directly injected along with various Compressed Natural Gas (CNG) energy shares (10, 20, 30, and 40%) as low-reactive fuel injected into the intake system. The experiments are performed in a single-cylinder, water-cooled, 1500 rpm, 3.5 kW power Compression Ignition (CI) engine under various loading conditions to examine the effects of CNG energy share on performance and emissions in Reactivity Controlled Compression Ignition (RCCI) combustion mode. The study found that the 30%CNG share decreased Nitrogen oxides (NOx) and smoke by 25 and 31%, as well as an increase in thermal efficiency of 4.35% in comparison to traditional biodiesel combustion. Finally, two machine learning (ML) models, namely the Gradient Boosting Regressor (GBR) and LASSO (Least Absolute Shrinkage and Selection Operator) Regression, were developed for predicting the dependent variables individually from the independent variables. Both the LASSO and GBR models achieved high accuracy with R2 values of 0.98–0.99 and relatively low Root Mean Square Error (RMSE) values.

Keywords

• CNG fuel
• Microalgae biodiesel
• Energy fraction
• RCCI combustion
• Machine learning techniques