摘要
为更好地控制桥梁施工的安全风险,开发一种基于贝叶斯网络理论的安全性评价方法。首先,构建风险因素、风险事件的贝叶斯网络关系模型;然后,引入连锁故障的事故链理论,以环节控制为核心,由影响传递、重要度、失效影响等指标组成环节关联性指标,通过K-means聚类分析选取关联性最大的下级节点,与上级节点组成环节,构成事故链,形成事故链集;最后,利用层次分析法(AHP)对各事故链评分排序,识别诱发风险事件的关键线路。用武汉某斜拉桥上部结构施工实例验证所提方法,结果表明,该工程项目应从强化安全意识、加强安全监管方面控制风险。
In order to have a better control of safety risk in bridge construction, a new safety assessment method based on Bayesian network was worked out. Firstly, a model was built for relationship between risk factors and risk events based on Bayesian network. Then, fault chain theory in cascading failures was introduced, link control was taken as core and relevance indicators consisting of delivery influence indicators, importance influence indicators, and failure impact indicators were identified. The branch with the highest value of relevance was selected as the next failure node by K-means clustering algorithm, links were composed of next failure node and superior node, and fault chains were composed of links, then fault chain set was formed. Finally, AHP was used to score for every fault chain, the critical line of risk events can be identified. The construction of superstructure of a certain cable-stayed bridge in Wuhan was taken as an example. The results in this case show that safety consciousness and inspection should be strengthened.
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2016年第7期157-162,共6页
China Safety Science Journal
关键词
贝叶斯网络
事故链模型
关键线路识
聚类算法
层次分析法(AHP)
Bayesian network
fault chain model
identify critical line
clustering algorithm
analytical hierarchy process（AHP）
