Gene expression programming-based model prediction of performance and emission characteristics of a diesel engine fueled with linseed oil biodiesel/diesel blends: An artificial intelligence approach

Abstract

ABSTRACT In the present study, Artificial Intelligence (AI) based Gene Expression Programming (GEP) is used to develop a model to predict the performance and emission characteristics of a single-cylinder diesel engine fueled with linseed oil methyl ester (LOME) blended with mineral diesel. The data to be used for GEP were obtained experimentally by varying the biodiesel/mineral diesel blending ratio, engine load, fuel injection pressure, and fuel injection timing. The GEP-based model was developed to predict the brake thermal efficiency (BTE), brake specific fuel consumption (BSFC), NOx, and unburned hydrocarbon (UHC) emission. A major part (70%) of the collected data was used for training and remaining (30%) was used for model validation. The developed GEP model was robust enough to provide a high degree of accuracy in the prediction of engine performance and emission parameters. The statistical measure of model robustness such as coefficient of correlation (R) was in the range of 0.9926–0.9999 and the coefficient of determination (R2) was 0.9854–0.9998 for the output prediction. The root mean square error (RMSE) in the GEP model predicted results were in the range of 0.0048–2.597 and mean absolute error (MSE) was 0.0037–4.386. Abbreviations: AI: Artificial Intelligence; BSFC: Brake specific fuel consumption; bTDC: Before top dead center; BTE: Brake thermal efficiency; B0: (MOME 0% + Mineral diesel 100%); B10: (MOME 10% + Mineral diesel 90%); B20: (MOME 20% + Mineral diesel 80%); CI: Compression ignition; CO: Carbon monoxide; ET: Expression tree; FIP: Fuel injection pressure; FIT: Fuel injection timing; GEP: Gene expression programming; ICE: Internal combustion engine; ID: Ignition delay; MAE: Mean absolute error; MIT: Machine identical tool; NOx: Nitrogen oxides; ppm: Parts per million; R: Regression coefficient; R2: Coefficient of determination; RMSE: Root mean square error; UHC: Unburned hydrocarbon