DOI: 10.3724/SP.J.1087.2013.03608

Journal of Computer Applications (计算机应用) 2013/33:12 PP.3608-3610

Application of geometric moving average martingale algorithm in anomaly analysis before earthquake based on sliding window

There are various abnormal phenomena before the earthquake, and how to effectively extract exception information before the earthquake is a very important research topic. The geometric moving average martingale algorithm based on sliding window was proposed to extract the anomaly features before earthquake. The seismic data were processed by geometric moving average martingale and sliding window feature extraction, and the anomaly features of earthquake could be effectively extracted before earthquake. Through the analysis of the NOAA (National Oceanic and Atmospheric Administration satellite outgoing long wave radiation information before the Wenchuan earthquake and Lushan earthquake, the experiments show that the algorithm can detect that the earthquake area is more obviously abnormal than the surrounding area. This anomaly can help researchers determine earthquake area before the earthquake.

Key words:sliding window,geometric moving average martingale,earthquake,anomalies analysis,feature exaction

ReleaseDate:2014-07-21 16:59:00