OVERBREAK PREDICTION OF UNDERGROUND MINE EXPLOSION USING MULTIPLE REGRESSION ANALYSIS AND ANFIS MODEL

• المؤلف الرئيسي: Akbar, Samsul
• التنسيق: Theses
• اللغة: Indonesia
• الوصول للمادة أونلاين: https://digilib.itb.ac.id/gdl/view/81150

Blasting is still a method commonly used in underground mine excavation because it is efficient in terms of time and cost. However, this method has a series of negative impacts that accompany it, just like overbreaking. If not handled properly, overbreaks can lead to safety and cost issues that hinder the progress of tunnel work. Overbreak is a mismatch in tunnel dimensions before and after blasting which can be avoided with good and optimal blasting design. A series of studies have been carried out to predict overbreak, but the ideal solution to this problem must depend on the rock conditions in each area. The Artificial Neuro Fuzzy Inference System (ANFIS) and Multiple Regression Analysis (MRA) approaches will be tried in this research to build a prediction model by analyzing the relationship between variables that influence the occurrence of overbreak. Data processing will be carried out using Matlab2022b software for ANFIS and SPSS25 for the MRA method. The accuracy or reliability of each method can be seen from the Root Mean Square Error (RMSE) value which is close to 0. The research results show that the RMSE value of MRA and ANFIS can reach 0.14 – 0.10. Overbreak prediction using the MRA and ANFIS methods provides relatively similar accuracy. Research proves that overbreaks at PT Antam Pongkor are influenced by a series of factors such as Perimeter Powder Factor (kg/m3), Maximum Charge Per Delay (kg), especially in poor rock mass conditions.