METHOD FOR DETECTING EARTHQUAKES IN MEMS BASED AUXILIARY SEISMIC OBSERVATION NETWORK, RECORDING MEDIUM AND DEVICE FOR PERFORMING THE METHOD

摘要

A method for detecting an earthquake in a MEMS-based auxiliary seismic observation network includes: preprocessing the acceleration data by performing de-trending of removing a moving average from the original acceleration data received from a single sensor of the MEMS-based auxiliary seismic observation network; calculating a short term average/long term average (STA/LTA) by applying a specified filter parameter value based on the preprocessed acceleration data; generating and delivering a message indicating event occurrence or event termination based on the calculated STA/LTA resu when an event occurrence message is generated, calculating a probability of an earthquake through an earthquake detection deep learning model by inputting the pre-processed acceleration data as an input; and analyzing noise by calculating a power spectral density (PSD) from the original acceleration data merged at regular intervals. Accordingly, it provides high-reliability seismic detection by a single sensor event validation process.