Data-based Sensing of Composition and Quality of Product in Biodiesel Production

Abstract

Fossil fuels are finite source of energy which are depleting due to anthropogenic activities. Therefore, there is a need of renewable alternative to fossil fuels. Biodiesel production is a point of great interest for researchers because of its renewable nature and little to no overall CO 2 emissions in the environment. Efficient operation of biodiesel production process is vital in realizing high product quality and less consumption of raw material. Data-based sensors are replacing the conventional, i.e., hardware based, sensing system in realizing stable and efficient operation. In this study, a data-based sensor is developed using ensemble learning method, i.e., boosting, for prediction of product composition and quality of product in biodiesel production process from vegetable oil. Data for the model development is generated through interfacing of MATLAB®, Excel ® and Aspen PLUS® environment. The proposed framework is highly accurate in prediction and is suitable for real time applications.