長短期記憶神經網路運用於現地型地震預警

Abstract

近震央地區在地震事件中承受著最劇烈的地表振動，也最易受到地震災害所造成的損失。然而，區域型地震預警系統在近震央50 公里的範圍內往往無法及時提供警報，成為區域預警中的盲區。現地型地震預警系統根據初始 P 波前幾秒來估算後續可能的破壞性 S 波，以在振動到來之前發出警告，並且有效的填補區域預警無法提供警報的盲區。前人的研究表示，P 波位移振幅峰值（Pd）是一個穩定可靠的現地預警指標，並且已經持續地應用在台灣的地震預警系統中。然而現地型地震預警在不同測站下各自路徑效應與場址效應之差異所影響，在實作上成為了一個複雜的非線性問題。在這種情形下使用一個固定閾值的單一指標會產生無可避免的誤報與漏報。為了克服上述問題，本研究提出應用長短期記憶（Long Short-Term Memory, LSTM）神經網絡來實現現地型地震預警。利用多層的長短期記憶神經網絡架構，本研究得以訓練一個高度非線性的神經網絡。該模型將強地動地震儀所記錄的加速度以及經過一次積分的速度波形作為特徵輸入，並在每個時間點上輸出後續振動達 80 Gal 的機率值。本篇研究使用了臺灣近期兩個重大的致災性地震來對模型進行測試，結果顯示長短期記憶神經網絡模型的漏報率為0％，誤報率為1.31％。此模型有效地減少現地型地震預警的漏報及誤報，同時提供了適當的預警時間。

On-site Earthquake Early Warning (EEW) systems estimate possible destructive S-waves based on initial P-waves and issue warnings before large shaking arrives. On-site EEW plays a crucial role to fill up the blind zone of regional EEW systems near the epicenter, which often suffers from the most disastrous ground shaking. Previous studies show that peak P-wave displacement amplitude (Pd) may provide a possible indicator of destructive earthquakes. However, the attempt to use a single indicator with fixed thresholds suffers from inevitable misfits, since the diversity in travel paths and site effects for different stations introduce complex nonlinearities. To overcome the above problems, this study presents a deep learning approach using Long-Short Term Memory (LSTM) neural networks. By utilizing the properties of multi-layered LSTM, it is able to train a highly non-linear neural network that takes initial waveform as input and gives an alert probability as the output on every time step. The LSTM neural network is then put into test with 2 major earthquake events that occurred recently in Taiwan, giving the results of a missed alarm rate of 0% and false alarm rate of 1.31%. This study shows promising outcomes in reducing both missed alarms and false alarms while also providing an adequate warning time for hazard mitigation procedures.