سامانه همایش های دانشگاه آزاد اسلامی واحد نجف آباد

صفحه اصلی / International Conference on Artificial Intelligence; City, Industry and Health

Predictive Modeling of Pollutant Emissions from Biodiesel-diesel Fuel Blends in a Diesel Engine Using Artificial Neural Networks

نویسندگان :

Alireza Shirneshan¹

1- Department of Mechanical Engineering, Na.C., Islamic Azad University, Najafabad, Iran

کلمات کلیدی :

Biodiesel،ANN،Emission،Diesel Engine

چکیده :

The increasing significance of alternative and renewable energy sources, particularly biofuels, is driven by the depletion of petroleum reserves and the environmental issues associated with fossil fuel combustion. This study investigates the application of artificial intelligence (AI) to predict the impact of biodiesel-diesel fuel mixtures and engine operating parameters—specifically engine speed and load—on the emission characteristics of diesel engines. An artificial neural network (ANN) model was utilized to establish relationships between these independent variables and the emissions of hydrocarbons (HC), carbon monoxide (CO), and nitrogen oxides (NOx). The results indicate that CO emissions can be predicted with high accuracy, reflecting a consistent response to variations in biodiesel content and engine load. Similarly, NOx emissions show a strong correlation with engine load and fuel blend. In contrast, HC emissions exhibit greater variability, likely influenced by combustion stability, which may not be fully captured by the selected input features. The ANN models demonstrated robust predictive capabilities for CO and NOx emissions, affirming the viability of data-driven approaches in biodiesel combustion research. This methodology can be expanded to incorporate larger datasets and more complex architectures to enhance predictive accuracy, particularly for HC emissions.