一种基于时空关联的灾害溯源算法

A Disaster Traceability Algorithm Based on Spatiotemporal Correlation

实现灾害溯源,从灾害源头对其进行抑制处理,能够有效杜绝灾害的扩散从而减少损失,对于灾害防治工作有着重要的意义。时空关联是基于大量的时空数据,实现对个体或是现象的连续跟踪或观察。基于时空关联的思想,结合数据挖掘技术,提出一种灾害溯源算法。算法通过时间约束和空间约束建立灾害时空事务表,对表中的潜在关联规则使用Apriori算法进行规则挖掘,并定义了灾害紧急度指标来衡量规则的有效性及关联强度;最终将筛选出的有效规则合并成有向图结构,实现直观的灾害溯源。最后以车公庙地铁站水灾为背景构造仿真模型,针对某一点的水灾事件进行溯源研究,证明通过该算法,能够有效地找出目标点灾害发生的源头。

Achieving disaster traceability and suppressing disasters from the source can effectively prevent the spread of disasters and thus reduce losses,which is of great significance for disaster prevention and control.Spatiotemporal correlation is based on a large amount of spatiotemporal data to achieve continuous tracking or observation of individuals or phenomena.Based on the idea of spatiotemporal correlation and data mining technology,a disaster traceability algorithm is proposed.The algorithm establishes a table of disaster spatiotemporal affairs through temporal and spatial constraints,mines the potential association rules in the table using Apriori algorithm,and defines a disaster urgency index to measure the effectiveness and association strength of the rules;finally,the filtered effective rules are merged into a directed graph structure to achieve intuitive disaster traceability.Finally,a simulation model is constructed with the background of flooding in Chegongmiao subway station,and a traceability study is conducted for a flooding event at a certain point.It is proved that the algorithm can effectively find out the source of disaster occurrence at the target point.