基于直觉语言集和变精度粗糙集的工程项目风险因素排序

Construction Project Risk Factors Ranking Based Intuitionistic Linguistic Set and Variable Precision Rough Set

针对具有信息模糊性和语言描述性的工程项目风险因素排序问题,利用直觉语言集和变精度粗糙集,建立工程项目风险因素排序模型,预测主要风险因素。该模型,首先利用直觉语言数表示评价信息;然后利用变精度粗糙集的属性依赖性构造判断矩阵,利用判断矩阵的相似性计算专家权重,利用判断矩阵的特征向量计算条件属性权重;再结合直觉语言集的运算规则,计算排序值;最后通过实例验证所构建模型的可行性。研究结果表明,所建模型不仅可以利用变精度粗糙集减少专家主观性造成的信息误差,更客观地计算专家和条件属性权重,还可以利用直觉语言数表达模糊性和语言描述性。这有助于更真实地计算工程项目风险因素排序值,帮助项目管理者更准确地了解主要风险因素。

For information fuzziness and linguistic description in construction project risk factors ranking problem, the construction project risk factors ranking model is proposed to forecast main risk factors based on intuitionistic linguistic set and variable precision rough set. Firstly, use intuitionistic linguistic numbers to express assessment information. Secondly, use the attribute dependency of variable precision rough set to construct judgment matrixes. Objective expert weight is calculated based on judgment matrix similarity. Condition attribute weight is calculated based on judgment matrix eigenvector. Thirdly, the risk factors ranking values are calculated based on the operational rules of intuitionistic linguistic set. Finally, an illustrative example shows the feasibility of the proposed method. The research results show that the proposed model can not only use variable precision rough set to eliminate information error caused by experts＇ subjective and calculate expert weight and condition attribute weight more objectively, but also use intuitionistic linguistic number to express fuzziness and linguistic description. It is conducive to calculate risk factors ranking values more veritably and help project managers to know key risk factor more exactly.