Modeling of a dual fueled diesel engine operated by a novel fuel containing glycerol triacetate additive and biodiesel using artificial neural network tuned by genetic algorithm to reduce engine emissions

In this study, a diesel engine was modified to operate in dual fuel mode with natural gas as the main fuel and a novel fuel mixture of biodiesel and glycerol triacetate additive as pilot fuel. Regarding to experimental tests results, engine emissions were modeled using a combination of artificial neural network and genetic algorithm to determine the appropriate ratio of pilot fuel to gaseous fuel, biodiesel and additive to reduce engine emissions. The algorithm inputs were engine torque, pilot fuel and natural gas consumption, biodiesel and additive proportions in pilot fuel, while outputs were exhaust emissions including NOx, PM, CO, and UHC. Overall, the results of the modeling were consistent well with experimental data. Simulations were performed for a variety of biodiesel and additive compositions and it was accordingly concluded that by using biodiesel and additive, NOx and CO emissions were reduced by up to 63% and 42%, respectively, while PM was reduced substantially by 27 times in comparison with neat diesel fuel in the diesel operation mode. In the dual fuel mode, 24% reduction in NOx emission. However, under these circumstances, UHC emission was 10% higher than that of the diesel operation mode. (C) 2018 Published by Elsevier Ltd.