Abstract: When using the entropy method to assign weight，as the discrete degree of an index is too large, the weight of the index will be very large. In order to solve this problem, using the ideas of Analytic Hierarchy Process method（AHP） to improve the entropy method, the index weights are solved through judgment matrix，which is obtained by the pair comparisons of indexes’ difference coefficients. The normalization method of maintaining the same index information entropy of the index initial data, the steps of calculating the mapping values from the 1 to 9 scale of AHP method to the 1 to 9 scale of improved entropy method, and the methods of constructing judgment matrix based on the information entropy and solving the index weights are given. The six index weights of fighter air-to-air combat capability are assigned using the entropy method and the improved entropy method respectively, and the fighter air-to-air combat capability is evaluated. The results show that the improved entropy method makes up the shortcoming of entropy method, the assignment weight result is more reasonable, and air-to-air combat capability evaluation results are more accurate.

Key words: effectiveness evaluation, entropy method, Analytic Hierarchy Process（AHP） method, improved entropy method, judgment matrix

摘要： 针对熵值法赋权时当某个指标的离散程度太大时，该指标的权重会过大这一问题，借鉴层次分析法（AHP）的赋权思路，对熵值法进行了改进，通过对指标差异性系数进行两两比较获得的判断矩阵来求解指标权重。详细给出了保持指标信息熵不变的指标初始数据的规范化方法，计算从AHP赋权法的1~9标度到改进熵值法的1~9标度映射值的步骤，以及构造基于信息熵的判断矩阵并求解指标权重的方法。分别利用熵值法和改进的熵值法对战斗机空空作战能力的六项指标赋权，对战斗机空空作战能力进行评估。结果表明改进熵值法弥补了熵值法赋权时单一指标权重可能过大的不足，赋权结果更加合理，空空作战能力评估结果也更为准确。

关键词: 效能评估, 熵值法, 层次分析（AHP）法, 改进熵值法, 判断矩阵