Establishing optimal engine output characteristics by tuning fuel injection parameter utilizing machine learning: operated by mahua biodiesel-diesel blend with antioxidants

This research aims at exploring the effects of antioxidants namely, butylated hydroxy anisole (BHA) and butylated hydroxytoluene (BHT) in reducing nitrogen oxide (NOx), smoke emissions and also optimizing engine input parameters for achieving improved engine output characteristics for hybrid powertrain application. Experiments are performed on a compression ignition(CI) engine fuelled with 30 % Mahua biodiesel-diesel blend which was established as optimal fuel blend in our earlier study. Appropriate injection timing (IT), engine load and fitting antioxidant were established by predictive and optimization modelling approach of engine characteristics using artificial neural networks (ANN) and response surface methodology (RSM). The engine input settings of M30B1000 blend and 26.4˚CA bTDC IT at 75.9 % load has been established for hybrid operation. A simultaneous reduction on carbon monoxide (CO), hydrocarbon (HC), NOx, and smoke by 15.3 % , 14.4 %, 4.5 %, and 15.03 % were realised compared to neat diesel. The ANN model has shown to have a greater correlation and a comparatively smaller prediction error than the RSM.