#TITLE_ALTERNATIVE#

• Main Author: FAISAL FATTAH (NIM 12203034), ESHA
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/10370
• Institution: Institut Teknologi Bandung

This time, Indonesia is facing a crucial dilemma related to decreasing of national energy that source from oil and gas. Supply disturbance will affect the industry output and others sector, and will cause the decreasing of national capacity in the global level. Because of the disruption of gas supply and the rising of oil price at several places in Indonesia, some of industries are forced to operate under their capacity. The prediction of oil and gas demand for all kind of industry is needed due to the short and long time supply planning. Therefore, it is important to do a study to arrange the model of oil and gas demand with dynamic characteristic and high accuracy level.<p>Artificial Neural Network is recommended for prediction, generalization, and classification. Neural network is not depend on formula and rules, it work by just using data sample. Therefore, it is suitable to be applied on problems with unknown formula, but the related variables are known. This study is using a feed forward ANN model with three layers and using back propagation training algorithm. The model is developed by using energy demand theory in industry sector, energy demand is the product of industry intensity and industry activity.<p>The prediction results show the increasing of national fuel demand at industry sector from 2004-2005 with the average growth about 6.6%. Energy demand in Java Island is the biggest one compare with another region it is about 70% from national demand. The most energy demand is from industry group of cement, lime and gypsum, and goods from cement and lime, Industry Group 264.