Human Reliability Analysis Of Falling From Height Based On HFACS-BN Models

At present,most of the human factors reliability research in the construction industry is based on the human factors data collection based on the safety accidents that have occurred,and there are many factors that lead to unsafe behaviors.Analyzing directly from the root cause,it is easy to ignore the utility between other factors.The interaction among various causative factors in the construction process is rarely considered.Therefore,based on various safety accidents of falling from heights in construction from 2012 to 2019,this paper,according to accident case reports,starts with indirect and direct causes,and sorts out the human factors that have an important impact on the construction process.Using HFACS（Human Factors Analysis and Classification System）,the hidden and explicit factors were divided into four levels,with a total of 12 elements and 37 index variables.Construct the HFACS-BN model of falling from heights,design a questionnaire for experts,and collect the safety importance evaluation of each factor by safety managers in the construction industry.The Cooder-Leeson reliability test was performed on the questionnaire data.The frequency of each index in the questionnaire was analyzed.Bayesian network.To explore the potential relationship among the internal levels,to analyze the chain of the main causes of unsafe behavior,and the sensitivity of each node,to verify the feasibility of the model with a practical case,and to propose specific corrective measures.Further reduce or avoid the building construction process in the occurrence of high fall accident.It can provide a new way of work for the safety production manager,which is helpful to understand the relationship between different human factors and improve the field safety management level.It has provided the safeguard to the work personnel’s life safety,has brought the positive function to the entire construction profession development.