Predication of performance and emissions characteristics adual fuel engine energized with liquid and gaseous materials by Artificial neural network

Abstract

The current study investigates the use of Artificial Neural Network (ANN) modelling for prediction of performance and emission characteristics biogas fuelled compression ignition engine under dual fuel mode. The biogas flow rate was varied from 1.2 to 3.2 kg/h while diesel fuel was injected as ignition source. The ANN model was developed to predict Brake Power (BP), Brake Specific Energy Consumption (BSEC), Brake Thermal Efficiency (BTE), Brake Specific Fuel Consumption (BSFC), Carbon Monoxide (CO), Carbon dioxide (CO2), and Hydrocarbons (HC) based on experimentation performed by varying load and biogas flow rates. In this study, ANN based models have been developed in the MATLAB2012a environment using the Neural Network toolbox, Back propagation algorithm, Levenberg- Marquardt Back propagation (TRAINLM), was used for the ANN structure for prediction of performance and emissions characteristics. There was a high correlation between the predicted values by the ANN model and the measured values resulted from experimental tests. The correlation coefficient was 0.99956 in the analysis of whole network when biogas and diesel were used, which implies that the model succeeded in prediction of the generator performance. It is found that the ANN modelling are good tools for prediction of performance and emission characteristics of biogas operated dual fuel diesel engine with varying engine operating loads and flow rates.