基于集对分析的工程评标未知权重多属性决策  

Unknown weight multiple attribute decision for engineering appraisal bidding based on set pair analysis

作　　者：何美丽[1] 刘浪[2,3] 王宏伟[1] 刘霁[1,3] 

机构地区：[1]湖南城市学院土木工程学院,湖南益阳413000 [2]长安大学地质工程与测绘学院,陕西西安710054 [3]中南大学资源与安全工程学院,湖南长沙410083

出　　处：《中南大学学报（自然科学版）》2012年第10期4057-4062,共6页Journal of Central South University:Science and Technology

基　　金：全国优秀博士学位论文专项基金资助项目(200449);湖南省科技计划项目(2011sk312);湖南省教育厅一般项目(11C0245)

摘　　要：基于工程项目管理需要统筹考虑各投标单位的情况,通过对工程项目评标管理进行分析,建立基于集对分析的工程评标未知权重多属性决策模型。该模型利用集对分析理论,结合指标值之间的差异性,全面又突出重点地构建工程评标模型。将集对分析与灰色关联分析相结合,利用关联度对差异度系数赋值,建立评标分级模型。为避免权重赋值的主观性,采用变异系数法确定指标权重,以提高权重的客观性与适应性。构造正、反同一度矩阵,使决策更加全面。研究结果表明:采用该模型所得结果与其他所得结果的平均关联度达到83.3%,从而验证了该模型的合理性与可行性。Based on the fact that project management needs the overall consideration of bidding units, through analyzing engineering appraisal bidding management, a multi-attribute decision making model with weight unknown based on set pair analysis was established. The theory of set pair analysis combined with difference among index values was comprehensively and emphatically used to make construction model of engineering appraisal bidding. Difference degree coefficient assignment was made to establish a grading model of engineering appraisal bidding using grey correlation degree. In order to avoid subjectivity, the variation coefficient method was used to calculate the weights of indexes to improve objectivity and adaptability. In order to obtain more comprehensive decision-making, positive and negative identical degree matrices were constructed. The results show that the average correlation degree of results between this model and other models is 83.3%, which verifies the rationality and feasibility of this method.

关 键 词：工程评标 集对分析 变异系数法 灰色关联分析 多属性决策 

分 类 号：TU72[建筑科学—建筑技术科学]