一种基于理想区间数和熵的多指标评价法

The Multi-object Weights Evaluation Method Based on Ideal Interval Number and Entropy

利用评价人员对待评方案给出的区间数评价值与理想区间数之间的差异和指标权系数的随机性, 给出一种新的多指标评价方法。利用优化方法建立数学模型,以实际评价值与理想评价值之间的差异和最小化与权系数熵的最大化为优化目标,用拉格朗日乘子法给出模型的最优解,得到指标权系数。该方法能够结合评价人员的主观意志和客观事实,精确确定各指标的权系数。

A new assigning method of index weights coefficients in multiple index evaluating is presented. Applying the ideal interval number of evaluated projects and the randomicity of index weight coefficients, a new mathematical programming model is established. The objective is to minimize the difference of real evaluating value and ideal evaluating value and to maximize the entropy of weights of each coefficient. The optimal solution is the best index weight coefficients, which is solved by the Lagrange multiple method, and the exact and reliable evaluation results combining the subjectivity and the facts are obtained. Finally, an example demonstrates the validity of this method.