基于Bayesian网络结构学习的绿色建筑项目风险网络分析

Trait Analysis on Network of Risk Factors of Green Building Project Based on Structure Learning of Bayesian Network

构建客观、准确的绿色建筑项目风险网络有利于对项目风险进行整体管控。针对绿色建筑项目风险网络构建缺乏客观性这一问题,在识别出项目风险因素的基础上,利用Bayesian网络结构学习从项目实地风险因素调查数据中挖掘客观的风险因素网络;其次利用社会网络分析方法分析绿色建筑项目风险因素网络中不同因素的地位和作用并构建了风险的解释结构模型。结果表明,绿色建筑项目风险因素网络中不同因素的地位和作用存在差异;网络中存在作为切割点的风险因素;绿色建筑项目的解释结构模型分为四个层级,风险因素网络中高层级的风险因素属于独立性因素,而底层的风险因素为联系型因素。研究基于数据驱动的风险因素网络发现为绿色建筑项目管理者提供了新的风险管理思路和借鉴。

Building an objective and accurate risk network of green building(GB) projects is conducive to the overall management and control of project risks. In view of the lack of objectivity in the construction of risk network associated with GB projects, based on the identification of project risk factors, Bayesian network structure learning was used to mine the objective risk factor network from the field survey data of project risk factors;secondly, the social network analysis method was used to analyze the status and role of different factors in the risk factor network of GB projects, and the interpretive structural model on network of risk factors was constructed. The results show that the status and roles of different factors in the network of risk factors are different, and there are risk factors in the network that act as cutting points;the interpretive structure model of GB projects has four hierarchies;in the network of risk factors on GB project, the risk factors with high level belong to independent factors, while the risk factors at the bottom level are linked factors. The research on network discovery of risk factors using the data-driven method provides a new risk management idea and reference for green building project managers.