Artificial Neural Network Modelling for Production of Biodiesel from Waste Cooking Oil

Abstract

Artificial neural networks (ANN) are bioinspired algorithms used in various engineering applications. The objective of this present study is to create a model algorithm for biodiesel synthesis from the collected waste cooking oil utilizing artificial neural networks (ANN). The factors consider to be influencing the biodiesel production are concentrations of solutions, time and temperature, pH of the solution, and agitation speed. In previous trials, methanol to oil ratio, sodium hydroxide to oil ratio, in addition to reaction temperature were taken as in terms of independent type of variables and % biodiesel yield inulin as a dependent variable. The results reveal that the 3-10-1 architecture of ANN provides goodness of fit to predict the percentage yield of biodiesel. The prediction ability of ANN is assessed by the coefficient of determination (R^2). The resultant R^2 value shows that the ANN predicted values fitted well to the percentage yield of biodiesel.