Experimental Investigation and Prediction of Performance, Combustion, and Emission Features of a Diesel Engine Fuelled with Pumpkin-Maize Biodiesel using Different Machine Learning Algorithms

Abstract

The current study examines the usage of biodiesel as a diesel substitute that is produced through the transesterification of pumpkin and maize with the addition of a diethyl ether (DEE) additive. Pumpkin-maize (PM) biodiesel and addition of diethyl ether (DEE), as well as their blends of 10%, 20%, 30%, 40%, and 50% with diesel, were used in performance, combustion, and emission examinations under various load conditions. According to the experimental findings, adding 5 ml of the DEE boosts BTE by 31.91 percent (B20 blend) compared to diesel. While the diesel equivalent of BSFC decreases by 9.519%. NO emission dropped by 34.91 percent at peak loads, HC emissions were augmented by 32.43%, and smoke opacity improved by 27.24%. To enhance the engine performance, combustion, and emission features of the substitute biodiesel diesel, the study emphasises the precise mix proportions of PM biodiesel with DEE addition. Using ANN, BTE, and NO were predicted with R2 values of 0.93 and 0.95, respectively. As can be observed, the R2 value for the ANN model was almost one, suggesting that the ANN models had better predictive power than other machine learning (ML) models.