Emission and performance investigation of mango seed oil biodiesel supplied with n-pentanol and n-hexanol additives and optimization of fuel blends using modified deep neural network

Abstract

In this study, the emission and performance characteristics of single-cylinder diesel engines were tested using various biodiesel blends prepared by mixing diesel with mango seed oil biodiesel (MSOB). Furthermore, the effect of n-amyl and n-hexanol alcohol additions on the performance and emission results of manufactured biodiesel blends is investigated and compared with diesel fuel. On the other hand, a hybrid deep neural network (DNN) based on the manta ray foraging optimization (MRFO) method is developed to forecast ideal biodiesel blends in order to reduce emissions from diesel engines while improving performance. The optimal brake thermal efficiency (BTE) and brake specific fuel consumption (BSFC) for this study were 32.3916 % for 75 % diesel + 20 % MSOB + 5 % n-hexanol fuel and 0.0453 kg/kWh for 75 % diesel + 20 % MSOB + 5 % n-amyl fuel, respectively. The optimal emissions from the test engine were 0.1034 % CO from 60 % diesel + 20 % MSOB + 20 % n-hexanol and 28.886 ppm HC from 75 % diesel + 20 % MSOB + 5 % n-hexanol fuel. The optimal smoke and NO x levels are achieved with a blend of 60 % diesel, 20 % MSOB, 5 % n-amyl, and 5 % n-hexane. Moreover, the developed DNN-MRFO achieved 0.9979, 0.9992 and 0.9975 overall regression coefficients during training, validation and testing. The root mean square error (RMSE) of DNN-MRFO also ranges from 0.019 to 0.032.