COMPARATIVE ANALYSIS OF AGRICULTURAL BIOMASS CHEMICAL EXERGY PREDICTION ON MOIST AND DRY CONDITION WITH ANN AND ANFIS METHOD

• Main Author: MAHARANI (NIM : 20216042), JUWITA
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/28369
• Institution: Institut Teknologi Bandung

In order to be wise in energy utilization of agricultural biomass, we need to know about maximum work in those biomass. The exergy (maximum work) which is contained in biomass is influenced by their element compositions. So this study focused on the chemical exergy of agricultural biomass. It can be influenced by physical character such as moisture content. With various moisture contents, directly determining the chemical exergy is difficult. By using element compositions data (C, H, O, and N), High Heating Value on dry basis (HHVd), and the entropy of biomass fuel (sF) which is known from the others fuel, then HHVd and sF is estimated by multiple regression analysis and ordinary least square method. The results of HHVd and sF estimation can produce 7.67% and 8.84% error relative to reference, respectively. The result of this study show that the chemical exergy is high when biomass is dry (moisture=0%). In this study were used alternative models such as ANN and ANFIS to predict the chemical exergy of agricultural biomass. Input data such as C, H, O, N, and moisture content of agricultural biomass is used to predict the chemical exergy. Based on testing results, ANFIS has higher accuration than ANN. So ANFIS method can be implemented to predicting the chemical exergy values of agricultural biomass with varians of moisture content. The results of this study are expected to be considered in predict the others chemical exergy of agricultural biomass.