基于熵算法改进的故障树-云模型风险分析方法  

Visual risk assessment method based on the fault tree analysis coupled cloud models using modified entropy algorithm

作　　者：龙丹冰[1] 解丽萍 杨成[1,2] LONG Dan-bing;XIE Li-ping;YANG Cheng(School of Civil Engineering,Southwest Jiaotong University,Chengdu 610031,China;National Engineering Laboratory for Technology Geological Disaster Prevention in Land Transportation,Southwest Jiaotong University,Chengdu 610031,China)

机构地区：[1]西南交通大学土木工程学院,成都610031 [2]西南交通大学陆地交通地质灾害防治技术国家工程实验室,成都610031

出　　处：《安全与环境学报》2022年第2期633-641,共9页Journal of Safety and Environment

基　　金：国家自然科学基金项目(51808455,51778537);国家重点研发计划项目(2016YFC0802205)。

摘　　要：为了改善建筑运维期失效风险的诊断和预测能力,提升其信息化程度,以一幢有显著劣化特征及病害且损伤分布特征复杂的古旧木结构建筑为例,基于改进的熵算法,提出了故障树-云模型(IEFTC)风险评估方法;考虑不同评价结果的标准差,给出了新的熵和超熵算法,既保留了专家对事件风险的判断,又量化了评价意见的离散度。该方法先利用故障树算法描述事件间的相关性并形成对应的风险地图,再结合云模型的数学分析功能,通过故障树的逻辑门进行模糊信息传递,可以在充分保留风险信息的基础上得到更为合理的分析结果。最后,实现了基于BIM的风险等级判断,并利用显示色度差异展示不同构件的风险程度。研究充分利用了云模型在模糊性和随机性方面进行数学分析的优势,提升了传统故障树方法处理和传递模糊信息的效果,并借助BIM进行可视化,有望在复杂病害结构的分析鉴定中提升评估及维护决策的准确度和效率。The present paper studies a typical old wooden structure building with the characteristic of significant deterioration,multiple diseases,and complex damage distribution to improve the ability to diagnose and predict potential failure risks during the period of operation and maintenance for ancient wood structures,as well as to enhance the level of building information utilization.Based upon the improved method of entropy calculation,the paper proposes a risk assessment method named IEFTC.The IEFTC method consists of fault tree analysis(FTA)and cloud model and a discrete coefficient is used for the calculation of entropy and hyper-entropy.Taking the variance of different evaluating results into account,the method not only retains the experts’evaluations on potential risks but also quantifies the dispersion degree of different evaluations.To implement the proposed algorithm,firstly,the fault tree analysis is constructed to form a risk map of failure mode by describing the correlation between events.Secondly,the characteristic values of the bottom events are generated and aggregated by the cloud model algorithm with the improved method of entropy calculation.When the bottom events get their basic cloud model,the cloud models of upper events will be aggregated by the aggregation algorithms in the cloud model theory,and the aggregation algorithms vary with different logic gates in FTA.Consequently,the fuzzy information is transferred and a more reasonable analysis result is obtained based on the retained risk information.Finally,we realize the visualized presentations of the risk levels of different components by using color legend in building information models.The study fully utilizes the advantages of the cloud model in processing information with fuzziness and randomness.Furthermore,the ability of FTA to process and transmit fuzzy information is improved by using the cloud model with the improved calculation of entropy.The results show that the proposed method can effectively improve the efficiency and accuracy

关 键 词：安全工程 风险诊断 故障树 云模型 BIM 风险可视化 

分 类 号：X93[环境科学与工程—安全科学]