A Fuzzy Decision in Volcano Warning System Forecast

Abstract The volcanic warning system in Indonesia is divided into four levels namely normal, waspada, siaga and awas. Based on the setting above, a study is needed to design a volcanic activity forecast system using temperature sensors, SO2, CO2 concentration, as well as vibration. The formulation of determining the conditions of volcanic activity is carried out by using fuzzy logic. Volcanic activity measurements are carried out in real time and continuously. Fuzzy logic has been used in the process of determining alert level of volcanic activity based on measured parameters. There are 53 fuzzy rules to determine volcano activity. Best predictions come from the combination of all sensors into a volcano monitoring and forecast system.