为解决军队工程代建模式风险评估过程中的主观性与不确定性,提出一种模糊层次分析法与贝叶斯网络相结合的风险评估方法.以军队代建工程质量安全事故数据为基础,建立模糊层次分析模型对代建工程风险损失权重进行评估.构建贝叶斯网络,利用Netica软件推理风险事件发生概率.结合风险等级隶属函数,得出代建工程整体风险等级.该方法可以有效降低风险评估的主观因素,及时发现高风险节点,更准确地把握风险态势.实例评估结果与工程实际相符,验证了评估方法的合理性和适用性,对工程代建模式风险评估具有一定的参考价值.
In order to solve the subjectivity and uncertainty in the risk assessment of the agent construction in military,a risk assessment method of project in agent construction mode that combines fuzzy analytic hierarchy process and Bayesian network has been proposed. On the basis of the statistics of safety incidents of military agent construction,the loss weight of agent construction is assessed by FAHP method. Bayesian network is built and Netica software is used to deduce the probability of risk incidents. Combined with the membership function of risk level,the risk rank of the whole agent construction is concluded. This method can be used to efficiently lower the subjective elements of risk assessment,find high risk node immediately and more correctly grasp the risk situation. The conformation between the assessment outcome of example and actual project has verified the rationality and feasibility of the assessment method,which has a certain reference value for the risk assessment of agent construction mode.