Artificial intelligence for prediction of performance and emission parameters of CI engine using bio-fuel

• Main Authors: Kolhe, Ajay V., Malwe, Prateek D., Wahile, Ganesh S.
• Format: Conference Proceeding
• Language: English
• Subjects: Air-fuel ratio; Artificial intelligence; Artificial neural networks; Biodiesel fuels; Brakes; Calorific value; Carbon monoxide; Correlation coefficients; Diesel fuels; Emissions control; Exhaust emission; Exhaust gases; Fuel consumption; Gas temperature; Mathematical models; Mixtures; Neural networks; Nitrogen oxides; Nitrous oxide; Parameters; Performance prediction; Thermodynamic efficiency
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0060862

The objective of this work is to find the performance and emission parameter of different blends of Karanja biodiesel with diesel and compare these parameters with pure diesel. This study investigates the potential of Karanja oil as a source of biodiesel. The objective of this work is to find the performance and emission parameters of 10 %, 20 %, 30 %, 40 %, and 50 % of blends with biodiesel and compared various parameters with diesel. The results showed that Brake Thermal Efficiency (BTE) decreases with an increase in the percent of biodiesel and Brake Specific Fuel Consumption (BSFC) decreases with an increase in the percent of biodiesel. Hydrocarbon (HC) and carbon monoxide (CO) emission reduces with an increase in blend percent whereas Nitrous oxide (NOx) emission increases with an increase in blend percent. Neural networks obviate the need to use complex mathematically explicit formulas, computer models, and impractical and costly physical models. In this work we use Neurosolution software for prediction of performance and emission parameters, separate models were developed for performance parameters as well as emission parameters. To train network, load, blend percentage, calorific value, the viscosity of fuel & air-fuel ratio was used as input value whereas engine performance parameters like brake thermal efficiency, brake specific fuel consumption & exhaust gas temperature were used as output value for performance model and engine exhaust emission such as NOx, CO, and HC values were used as the output parameters for emission model. Artificial Neural Network (ANN) results showed that there is a good correlation between the ANN predicted values and the experimental values for various engine performance and exhaust emission parameters. It is observed that the ANN model can predict the engine BTE, BSFC with a correlation coefficient of about 0.998435668, 0.990616392, and 0.993346689 respectively for performance model and emission model CO, HC and NOx predict with a correlation coefficient of 0.986098699, 0.991243454 & 0.9855593.