The Application of Artificial Neural Network on Predicting Brazil Nut Effect Phenomena
Muhammad Iqbal Rahmadhan Putra, Nadhira Azzahra, Sparisoma Viridi

Nuclear Physics and Biophysics Research Division, Department of Physics, Institut Teknologi Bandung.

Abstract

Brazil nut effect is a phenomenon in which large-size particle lift up when vibrated vertically on granular system. This phenomenon has been studied both through experiments and computational simulations. One parameter that can be extracted from this phenomenon is the final height of the large-size particle. However, due to the large number of input parameters that affect the final results, the way to predict is still the subject of study. In this study, we used artificial neural network (ANN) as a method to predict the final results of experiments through molecular dynamic simulations with different input parameters. The different input parameters were created by generating random numbers of the appropriate range. After the simulation was carried out and the final result was obtained, the input and final parameter data were used as a database for training ANN. To get optimal results, a comparison is made of various ANN model architectures.

Keywords: granular, simulation, molecular dynamic, artifical neural network,

Topic: Simulation in general, etc

Link: https://ifory.id/abstract-plain/MwzaNcKDykmf

Web Format | Corresponding Author (Muhammad Iqbal Rahmadhan Putra)