OPTIMIZATION OF VOLCANO MONITORING SENSORS FOR MOBILE ROBOT APPLICATIONS USING FUZZY LOGIC CONTROLLER

• Main Author: Apriyana, Nazira
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/71514
• Institution: Institut Teknologi Bandung

Indonesia is one of the countries with the most active volcanoes in the world. Of the 127 active volcanoes in Indonesia, 70 of them are monitored by the Center for Volcanology and Geological Hazard Mitigation (PVMBG). One of them is Mount Tangkuban Parahu which includes active basaltic to andesitic stratovolcano volcanoes. The presence of wireless sensor network technology with data transmission can be an alternative in monitoring existing active volcanoes. Therefore, this research will implement optimizing volcano monitoring sensors on mobile robots that have been developed previously and control these sensors with a fuzzy logic controller. Fuzzy logic with a triangular and trapezoidal membership system used as a control system for this mobile robot has been designed with input in the form of sensor data and output in the form of volcano status based on data from the sensor. In the early stages, the system has been tested on a laboratory scale. But previously, the sensor has been calibrated on the sensor and simulated the fuzzy logic control system. In testing, data from the sensor shows normal conditions as stated by temperatures ranging from 22-27 oC, SO2 concentrations below 0.6 ppm, CO2 concentrations below 0.015 ppm, seismicity with almost zero frequency (in Hz), no soil slope. there is a change during data collection which is expressed by a change in slope that is almost zero. From this sensor input, the system stated that the conditions at that time were normal, which was expressed by a crisp value of 25%, which meant that the status of the volcano at that time was normal. From the results of testing in this laboratory, it is hoped that the system can be used for testing on real volcanoes. After testing on a laboratory scale, testing was then carried out on a real volcano field, namely Mount Tangkuban Parahu. Retrieval of scale test data on Mount Tangkuban Parahu was carried out by collecting data for 5 days in the morning, afternoon and evening. In testing data on Mount Tangkuban Parahu, temperature sensor data shows normal conditions ranging from 29-30 oC (morning data), 24-25 oC (afternoon data), 25-26 oC (afternoon data). While SO2 concentrations ranged from 0-0.000075 ppm (morning data), 0-0.000083 ppm (afternoon data), 0-0.000973 ppm (afternoon data). CO2 concentrations ranging from 0-0.009915 ppm (morning data), 0-0.152869 ppm (afternoon data), 0-0.052869 ppm (morning data), seismicity with almost zero frequency (in Hz) with the x-axis circle, y and z respectively (0.082604; -0.01532; 0.01279), the slope of the land is almost zero change with the x, y and z axes respectively (0.227527; 0.213157; -9.41243) m/s2 . In the development of fuzzy logic to analyze and determine the status of the volcano on the sensor. Based on the temperature detected by the temperature sensor, while other input parameters are supplied internally or externally with the associated sensor. These parameters can determine the appropriate warning (normal, alert alert and alert). The warning results can be entered into the next control system to analyze and determine the status of the volcano that changes on the sensor. From the system input it states that conditions are under normal circumstances which are expressed by a crisp value of 25% or by indicating the status of a volcano which is categorized as normal, so that the condition of the volcano does not show a significant increase in activity. In this case there is no increase in temperature around the crater, there is no decrease in the level of the slope of the land and no earthquake tremors occur