Research on risk management of construction safety based on Bayesian network

In order to solve the shortcomings of existing building construction risk management methods in dealing with uncertainty, a construction risk management analysis method based on Bayesian network (BN) theory is proposed. Firstly, based on management experts decision-making methods, the main factors affecting construction risk management are determined, and the security risk Bayesian network topology is constructed. Then use Bayesian network forward reasoning to predict the probability of construction risk in different situations, and analyze the cause of risk by combining backward reasoning. Finally, a sensitivity analysis based on the mutual information index method was used to identify the sensitive risk factors. Combining with the historical data of domestic construction projects, this method is applied to the practice of construction project safety risk management. The results show that the construction risk probability is 3.36%; It is the biggest risk factor that the safety risk existing is not solved timely in construction.