Research On Correlation Measure And Decision Of Construction Engineering Project Risk

With the development of market economy in China,construction projects gradually tend to be large-scale and modernized,and the risk of construction projects has become more and more uncertain,and the loss is becoming more and more serious.In order to effectively prevent the occurrence of risk,or to minimize the loss caused by the risk,how to determine scientific and effective decision making in risk management is of significant.Moreover,the risk factors of the engineering project interact and influence each other,so there is a certain correlation,which can be measured by the loss caused by different risks and the probability of risks.On the one hand,the portrayal of risk correlation can improve the theory of risk management and make the system of construction project risk management more complete;on the other hand,it also improves and consummates the existing model of objective risk decision.Before the risk correlation is portrayed,it is necessary to identify the risk first.Risk identification is the basis of risk management in construction projects,and any risk management can not be separated from risk identification.This paper first expounds the risk identification of construction project,and establishes the risk list.Then,with the aim of minimizing the risk loss expectation and the risk control cost,a multi-objective risk decision-making model is established.According to the expression of risk loss expectation,this paper quantitatively analyse risk correlation from the loss and probability level respectively,with the correlation of loss level measured by fuzzy measure,and the correlation of probability level measured by Bayesian network.Under the premise of known prior probability and conditional probability,the paper used Bayesian network to obtain posterior probability,and transform from static Bayesian network to dynamic Bayesian network by introducing time series.Considering the correlation based on the original multi-objective risk decision-making model,a multi-objective decision-making model of construction project considering risk at loss level and probability level is established accordingly.The model reflects the multi-objective risk decision-making problem and the minimum risk loss expectation and the minimum risk control cost are conflicting goals.How to balance it can be realized by multi-objective particle swarm optimization.Compared to the traditional algorithm,Multi-objective particle swarm optimization(MOPSO)has faster convergence and simpler operation,and it can optimize multiple target functions at the same time.Through the adoption of multi-objective particle swarm optimization to solve the model,and cited example of its risk control problem for correlation analysis,the results found that the model considering the correlation is more practical than the model without the correlation.Besides,the dynamic Bayesian network for risk probability assessment is more accurate than static Bayesian network,and by solving the decision plan can also provide strong scientific basis for the actual risk control then helps to improve the risk management level of the decision-makers.