摘要
通过对地震前兆数据的分析与研究,研制了一套地震前兆数据异常自动检测与报警系统。该系统可实现测项分量与检测方法的自由定制,通过利用相关分析、差分检测、固体潮相关检测、方差检测、台阶检测与金日光第四统计力学群子统计理论等多种检测方法,实现了对数据的自动异常检测与报警。试验证明,该系统有效可行,可为地震数据监测人员与地震预报人员提供一定的参考。
We have developed a auto earthquake precursor data outlier detection and warning system through research on earthquake precursor data features. The system can complete the＇ data outlier detection and warning task automatically. The user can customized select the detec- ting objects and detection algorithm, the system realized the auto outlier detection and warn- ing of earthquake precursor data through the use of the correlation analysis, the time domain differential detection, the earth tide correlation analysis, the standard variance detection, the data step detection and the Fourth Statistics-JRG Sub-cluster Statistics Theory and so on. The experiment proves that the system work effectively, and it can provide some reference for the earthquake prediction persons.
出处
《地震地磁观测与研究》
2011年第1期101-108,共8页
Seismological and Geomagnetic Observation and Research
基金
中央级公益性科研院所基本科研专项(J220981301)
关键词
地震前兆
数据异常检测
时域差分检测
标准方差
固体潮
数据异常报警
earthquake precursor, data outlier detection, time domain differential detection, standard variance, solid earth tides, data abnormity alarm
