Scenario Construction And Evolution Analysis Of Construction Crane Accidents Based On Dynamic Bayesian

The rapid development of the construction industry drives the rapid development of China economic level,and provides a large number of employment positions for China.However,there are many uncertain factors such as complex construction environment,numerous construction procedures,and long construction period at the construction site.The crane is one of the machines widely used in the current construction in China,because the structure of the machine is complex,the operation is difficult,and the personnel different management levels make it difficult to detect hidden risks,resulting in frequent crane accidents.In order to improve this phenomenon,based on the theory of accident causes and combining with the characteristics of crane operation,this paper constructs a scenario model of crane accidents,identifies the causes of crane accidents and discusses the evolution process of crane accidents,which can provide reference for the safety management of crane construction.Taking construction crane injury accidents as the research object,this paper collected 168 crane accident investigation reports from 2012 to 2021.Reading and analyzing accident investigation reports,the time and types of accidents,the number of casualties and the number of dead jobs are counted,and the characteristics and rules of crane injury accidents are summarized by analyzing these data.Then the actual accident investigation report content combined with HFACS framework analysis,the cause of the accident is summarized from the four aspects of human-machine-environment-management,and the fault tree model of the construction crane accident scenario is constructed.In order to quantitatively calculate and dynamically analyze the relationship between the causes,the fault tree is mapped to a static Bayesian network structure according to the corresponding transformation rules.The key accident causes are obtained by quantitative analysis of static Bayesian network.According to the actual crane accident scenario,the model is simplified and adjusted,and the dynamic accident nodes are screened out.The model is transformed into a dynamic Bayesian network to analyze the change of each node in the Bayesian network with time and the probability of the whole process.In order to study the failure process of the network,a cascading failure model is designed based on Bayesian network reasoning.Finally,the corresponding countermeasures are put forward according to the analysis results.The research results show that:inexperienced operation,unprofessional operation,crane without maintenance,collision of lifting objects are the most important factors leading to crane injury accidents.In the whole period of construction,the probability of accidents at the beginning stage,the use stage and the end stage is affected by various factors,and the related reasons are analyzed.It provides reference and preventive measures for related management parties to prevent crane accidents and control the risk of crane injury accidents.