Risk Management Of Southwest International Trade City Based On Bayesian Network

With the economic and social progress and people’s quality of life to pursue, large and complex construction projects are flourishing, and these large-scale projects filled with more uncertainty and complexity. For the general contractor, although the contract of a large construction project can bring huge economic benefits to the enterprise, various uncertainties may also bring the risk of loss in the construction process. In the construction of the building process, using Bayesian network project risk management to improve the resilience of the consequences of risk prediction and risk measures has important practical and theoretical significance.The purpose of using Bayesian networks for the project risk management is to grasp the internal laws of the risk factors that affect the cost of the project, schedule, quality and safety problems caused by the risk factors. Combined the Bayesian probability formula and the conditional probability tables obtained from the parameters learning of risk variable, the likelihood of each risk and risk events occurring can be determined. The ability to predict the risk of the consequences was improved, targeted risk response and control measures were proposed, risk management of projects was improved, the degree of risk of loss of the project was reduced, which has an important practical and theoretical significance on the smooth progress of the project.Based on the studies of domestic and international risk management of construction projects, combination my own learning practice in the project-Southwest International Trade City of Guiyang, this paper used the Bayesian network model for risk management in a study of its practical application, mainly in the following research:The theory of project risk management and risk factors were described through relevant information, the inherent law of development between the risk factors and the risk events was analyzed, the main risk factor variables and risk events were sort out.A Bayesian network model of project risk was established by using Bayesian theory and application of a causal relationship between risk factors。The establishment of project risk management model was used for structure learning and parameter learning, risk value of direct and indirect risk variables for safety, quality, schedule and cost were estimated the risk variables, and then the risk of the entire project was evaluated. We could propose measures of control risk factors to reduce the severity of the risk events, which can reduce the loss of the project caused by these risk factors.Southwest International Trade City of Guiyang is a large engineering project under construction, various risk factors and their impact in the proportion change dynamic during the construction. In the implementation process, we can establish a good Bayesian network model based on repairing risk factors, then Bayesian network probabilistic algorithms was used to derive probability of safety, quality, schedule and cost issues, for which managers can improve ability of risk prediction and reduce the risk or avoid risk through effective measures. Finally, more detailed recommendations of risk control measures were put forward for Southwest International Trade City.In this paper, the major contributions are following aspects:(1)A Bayesian network model of project risk was established based on the development of the law of the major project risk factors and risk events. The major risks and risk events of decision variables in the Bayesian network were used for structural learning and parameter learning. Direct and indirect risk could be obtained from structure learning, come to the decision variables, the affecting between each of the risk variables and the decision variables was quantified and the value of the risk can be estimated by the parameter learning. We use the methods of both structural description and quantitative deduction.(2)Southwest International Trade City of Guiyang was applied to the method of the Bayesian network risk management, the impact of risk factors on the risk events analyzed, the risk value of the various risk factors and risk events could be obtained by Bayesian theory. The extent of the impact on risk factors and risk events was calculated by using the methods of both structural description and quantitative deduction. The Bayesian network risk management was applied to the practical project, and the dynamic management of Bayesian networks was achieved.