考虑属性和空间逻辑的地震灾情数据清洗器设计与实现

针对目前地震灾情信息普遍存在的质量问题,根据灾情数据的属性逻辑,采用逻辑分离的清洗框架中基于规则引擎的数据清洗方法进行初步清洗;再从灾情数据的空间逻辑关系出发,通过聚类分析进一步修正数据中存在的错误,弥补基于逻辑分离的清洗框架在孤立点异常检测方面的不足。基于上述规则引擎与空间聚类相结合的清洗技术,设计并实现了一种考虑属性和空间逻辑的地震灾情数据清洗器,并应用于2017年"8·8"九寨沟7.0级地震,分析四川省地震灾情快速上报接收处理系统接收到的客观震感灾情信息。结果表明:地震灾情数据清洗器可快速检测并修正异常的灾情信息,有效提高接收到的灾情数据质量。

Design and Implementation of Seismic Disaster Data Cleaner Considering Attributes and Spatial Logic

Aiming at the general quality problems of the current earthquake disaster information,we use the rules engine-based data cleaning method in the logically separated cleaning framework to perform preliminary cleaning of the disaster data according to the attribute logic of disaster data.Then starting from the spatial logical relationship of disaster data,we further correct the errors by cluster analysis,which can make up for the shortcomings of outlier detection based on logical separation cleaning framework.By the cleaning technology combining the above rule engine and spatial clustering,we designed and implemented an earthquake disaster data cleaner that considers attributes and spatial logic,and applied it to the 2017 "Agu.8" Jiuzhaigou M7.0 strong earthquake to analyze the objective seismic sensation disaster information received by the Sichuan earthquake rapid report receiving and processing system.The results show that the earthquake disaster data cleaner can quickly detect and correct abnormal disaster information,and effectively improve the quality of the received disaster data.