BOT项目风险多属性群决策识别方法

本文提出的BOT项目风险识别方法可以客观计算专家的权重和风险属性的权重,反映不同专家、不同属性对决策结果的影响。通过把BOT项目风险决策矩阵转化成多维向量,用两向量的夹角余弦计算专家的相似度和差异度;利用熵值推导得到属性的权重,如果某个属性对决策所起的作用越大,则其熵值就越小、权重越大,并根据综合决策值排序得到风险识别结果。最后,本文通过一个典型的BOT项目验证了该识别方法的可行性和有效性。

Multiple Attribute Group Decision Method of Risk Identification of BOT Project

A new risk identification model for BOT project is established to calculate the weights of experts and attributes more objectively,and to reflect the impacts of different decision-makers and different attributes more reasonably. Firstly, angle cosine of two-space vector transformed from multi-dimensional vector is derived into the similarity and the difference of experts. Secondly, the weights of attributes reflecting the importance degree for decision-making are derived from entropy expressing information content of attributes. While the entropy of a attribute is smaller,the weight is bigger. Risk identification result is based on the value of integrated decision-making. A typical case proves that the method is feasible and reasonable.