ABSTRACT
In this work, the experiments of the transesterification process were carried out on jatropha-algae oil blend and the prediction of the synthesized biodiesel was investigated. The study was divided into two parts. In the first part, a series of experiments were employed practically and in the second part, the prediction is made with the artificial neural network (ANN). The ANN with Levenberg–Marquardt (LM) algorithm was trained with topology 4–10-1. The estimated results were compared with the experimental results. An ANN model was developed based on a back-propagation learning algorithm. An R-square value of the model from ANN was 0.9976. The results confirmed that the use of an ANN technique is quite suitable. The artificial neural network gave acceptable results.
Abbreviations
ANN | = | Artificial neural network |
FAME | = | Fatty acid methyl esters |
FFA | = | Free fatty acid |
RSM | = | Response surface methodology |
GA | = | Genetic algorithm |
BBD | = | Box-Behnken Design |
KOH | = | potassium hydroxide |
LM | = | Levenberg–Marquardt |
RP | = | Resilient backpropagation |
SCG | = | Scaled conjugate gradient |
BSO | = | Biofuel sustainability ordinance |
Linear Transfer Function | = | purelin |
Log-Sigmoid Transfer Function | = | logsig |
Tan-Sigmoid Transfer Function | = | tansig |
potassium hydroxide | = | KOH |