Research Of Large-scale Engineering Projects’Schedule Risk Research Based On Bayesian Network

The goal of schedule management is an important part of large-scale engineering projects. Successful and efficient schedule management of the project is the guarantee of not only the economic benefits, but also the social benefits. Because the large-scale engineering projects contain lots tasks and have long durations, they are always in a complex and long-term dynamic environment, facing the threat of variety uncertainties. So it is necessary to study the schedule risk management of large-scale engineering projects.This article is based on Bayesian network to study the schedule risk of large-scale engineering projects. Firstly, it overviews the researches about schedule risk management and Bayesian network, which have been done domestic and abroad. Then it describes and analyzes the common risk factors of schedule management in large-scale engineering projects, organizing the general steps of risk management. After that, it describes the Bayesian Theorem, also the definition and characteristics of Bayesian networks.Based on the above theory, this paper identifies the schedule risks of large-scale engineering projects through the use of historical literature analysis, brainstorming and questionnaire method. It defines15risk factors and then analysis them based on the risk level matrix, considering the possibility of the delay to the projects and the occurrence of the risks.In this paper, a Bayesian network has been built based on the K2algorithm、Data Learning and correlation analysis, using the obtained data to learn Bayesian networks parameter, and get conditional probability distribution of the variables. Then it uses sensitivity analysis and the largest chain of causation to get the most critical and sensitive risk factors, proposing some advice to improve the efficiency of large-scale engineering project management on schedule.At last, this paper applies the schedule risk Bayesian network model to the Nanguang Railway Yujiang Bridge to analysis its schedule risk, obtained a result of the construction period of Yujiang Bridge. This paper identifies and analyses the risk of large-scale engineering projects’schedule management, building the Bayesian network model and validating the feasibility and applicability for large-scale engineering projects’schedule risk management. It provides an effective and convenient management tools for quantitative prediction schedule risk management of large-scale engineering projects.