Study On Safety Assessment Of Building Construction Based On Genetic Algorithm And Bayesian Network Hybrid Method

The construction industry has always been an important industry in our country.In recent years,with the development of our country's economic the construction industry is also growing,but the construction safety in the construction industry has always been a problem that has plagued domestic practitioners and scholars.The frequent construction safety accidents in our country have caused many irreversible losses to society and individuals.Both the Chinese government and building construction agencies have formulated many policies and measures for responding to accidents,but they still cannot effectively prevent the occurrence of safety accidents.This requires us to fully study and analyze the construction site conditions,monitor the safety status of construction projects in real time,and formulate corresponding preventive measures to stop the occurrence of safety accidents at the source.On the basis of fully researching construction safety accidents,this paper introduces BIM technology into the construction site and monitors the construction project in real time to collect relevant data information.Through the analysis and processing of data,the safety status of the construction project can be obtained and the weak links of the construction safety management can be found out,it can provide a reference for the construction safety management.The main research work is as follows:(1)Research on Evaluation Index System of Safety in Building Construction.Based on the 2017 construction safety accident report and literature study,a comprehensive evaluation index system for building construction safety was obtained,and the connotations of various influencing factors in the indicator system were analyzed.(2)Research on attribute reduction using genetic algorithms in safety evaluation of building construction.Using BIM technology to collect real-time monitoring data of 10 construction projects in Xi'an City,and use genetic algorithms to analyze and process the data,and to screen the influencing factors of construction safety and finally 20 most representative key influence factors were screened out from 29 influencing factors.(3)Research on safety evaluation modeling of building construction based on Bayesian Network.Bayesian network was used for structural learning and parameter learning.Finally,a building construction safety evaluation model was established.The model was applied to a BIM construction project in Zhengzhou for safety evaluation.Finally,we conclude that the probability of this construction project being in different security levels is 0.028,0.203,0.495,0.228,and 0.046,which are in a basically safe state.(4)Sensitivity analysis of a BIM construction project in Zhengzhou City.We found that the project's sensitivity factors are: construction risk(C1),continuous operation time(C6),protective measures(C14),material hazard(C17),maintenance times(C23)and construction method(C24).