一种基于联系向量“垂面”距离的改进TOPSIS多属性决策方法

An Improved TOPSIS Method Based on Vertical Projection Distance of Connection Vector

提出一种属性权重未知情境下,基于联系向量“垂面”距离的改进TOPSIS多属性决策方法VSPATOPSIS.该方法将正理想点与负理想点视为不确定系统中相互对立的集合,在测度目标方案与正理想点和负理想点的联系度时,充分考虑了对立集合的存在.通过引入联系向量“垂面”距离的概念.计算相对贴近度.从而在一定程度上克服了SPA-TOPSIS法逆排序的不足。决策中针对传统客观加权方法忽略自然权重的不足,将“动态最大均值逼近标准化方法”与"传统离差最大化方法''进行组合,实现了消除自然权重的离差最大化定权.保证了所提方法的可操作性。最后,从实际应用与定量交叉比较研究两方面验证了该方法的合理性与有效性。

This paper proposes a novel method for decision-making problems, in which the ideal point and negative idealpoint are regarded as an opposite set in a system both having certainly and uncertainty. When inspects the connectiondegree between the objective project and ideal point or negative ideal point is inspected≫ the existence of an opposite set isfully considered. By introducing the concept of the "vertical" distance of the connection vector≫ the relative closenessis calculated, which to some degree improves the SPA-TOPSIS method. Aiming at the deficiency of traditionalobjective weighting method which neglects natural weights, we combine the **dynamic maximum mean approximationstandardization” method with traditional maximizing deviation method in the evaluation process to realize themaximization of the deviation of the natural weight elimination and guarantee the operability of the proposed method.Furthermore, the rationality and validity of the method are verified from two aspects: practical application and quantitativecross-comparison.