ANALISIS ARTIFICIAL NEURAL NETWORK DALAM PENENTUAN ARRIVAL TIME GELOMBANG P DAN S

• Main Author: Ilyasa Rachmanditya, Bima
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/76745
• Institution: Institut Teknologi Bandung

Earthquake Earthquake Early Warning (EEW) is a system for sending early warning of earthquake events, with the stages of detecting and determining the location of earthquakes. The selection of the arrival time of the P waves and S waves is the parameter used to determine the location of an earthquake. For this reason, the time required to determine the arrival time of P waves and S waves has an important role in EEW. Generally, determining the location of earthquakes is done manually, with artificial intelligence being one of the methods for determining the arrival time of an earthquake automatically. Deep learning is part of artificial intelligence that can process data quickly, has high scalability, and has a high level of accuracy. PhaseNet is a deep neural network (DNN) program, which is a further classification of deep learning with the aim of determining the arrival time of P waves and S waves. In this study, the authors used waveform data from 89 BMKG stations located on the island of Java, Indonesia during the recording time range from April 2020 to September 2020. The duration of the waveform used was five minutes and then PhaseNet was applied to obtain the arrival times of the P waves and S waves. The results of determining the arrival time of the P wave and S wave were compared with the GaMMA program to obtain a consistent distribution pattern of the wave phase selection results. Earthquake locations were determined using the NonLinLoc program. The arrival time results based on PhaseNet obtained 312 out of 333 arrival times of P waves and S waves, these results have a difference of less than one second. The results of determining the location with PhaseNet data have a difference with BMKG with an average longitude of -0.0071o, latitude of 0.39369o and depth of 3.17981 km.