基于Apriori算法的地铁施工致险非线性因素的映射指标研究

RESEARCH ON MAPPING INDICATORS BETWEEN THE NONLINEAR RISK FACTORS OF SUBWAY CONSTRUCTION BASED ON APRIORI ALGORITHM

地铁施工环境极其复杂,诸多风险因素发生并组合后易引发重大灾害事故,但是对所有风险因素及组合同时进行监控很难实现.文章研究目标是在降低风险因素组合维度的基础上,指出它们的映射方式.首先从地铁施工灾害事故中找出风险因素组合,即事故的最小割集集合.然后,利用关联规则Apriori算法进行数据压缩,确定频繁项集.最后通过风险因素出现的频数确定其在传递过程中的作用,分析出地铁施工事故的非线性风险因素之间的映射关系指标.研究成果在实际应用中,降低了风险因素与灾害事故之间的维数,并为提高识别精度和节省控制成本带来一定的贡献.

The environment of subway construction is extremely complicated, many risk factors and their portfolios can cause the explosion of serious disaster accidents. However, it is difficult to monitor all the risk factors and their portfolios simultaneously. The aim of this paper is to reduce the dimensions of risk factors and based on that to propose their mapping ways. We first find out the risk factors in the disaster accidents and their combinations, which are the minimal cut sets of the accident. Second, we employ the association rules Apriori algorithm for data compression and determine frequent item sets. Finally, according to the frequency of risk factors, we determine the effect of risk factors on the risk transition to determine the mapping indicators among nonlinear risk factors in subway construction accidents. The ap- plication of our results can reduce the dimensions between disaster accidents and risk factors, and provide great potential contribution to improve the identification accuracy and reduce relevant costs.