时序趋优偏好下的TOWA算子及其应用  

The Time Ordered Weighted Averaging Operator and Its Application with Time Series Optimal Preference

作　　者：张发明[1] 何意雄 于婷[1] 杨红暾 ZHANG Fa-ming;HE Yi-xiong;YU Ting;YANG Hong-tun(School of Economics and Management, Nanchang University, Nanchang 330031 ,China)

机构地区：[1]南昌大学经济管理学院,江西南昌330031

出　　处：《系统工程》2018年第2期146-150,共5页Systems Engineering

基　　金：国家自然科学基金资助项目(41661116;71361021);江西省教育厅科技重点资助项目(GJJ150027);江西省社学科学"十二五规划"重点资助项目(15ZQZD01);江西省赣鄱英才555工程资助项目(第三批);江西省青年科学家(井冈之星)资助项目(第九批)

摘　　要：由于传统的时序加权平均(TOWA)算子在对动态评价信息进行集结时忽略了不同时刻系统发展程度不同所带来的影响,因此对其进行了拓展与完善。首先运用改进后的灰色关联分析法求得被评价对象在各个时刻的静态综合评价值差分序列与最优时序数据差分序列的几何相似度,以测算系统发展的时序趋优程度;然后将偏好时序趋优数据的管理思想融合在时间权向量的确定方法中,根据时序趋优度的分布密度得到"趋优度"时间权向量和融合"时间度"和"趋优度"时间权向量的确定方法;最后分别运用两种偏好下的TOWA算子进行动态综合评价并与原有方法进行结果的对比分析,验证其有效性和可行性。Traditional time ordered weighted averaging （TOWA） operator ignore the influence produced by the differences of system development at different times when the operator aggregates the dynamic evaluation data. In this paper, the problem is solved and traditional TOWA operator is expanded. Firstly, The improved gray incidence analysis method is used to measure the similarity of each time＇s evaluated objects＇ static evaluation value＇s difference to the best time series data＇s difference in order to compute the evaluated system＇s each time＇s optimal trending status. Next, this paper tries to integrate the decision makers＇ management mind of preferring time series＇ optimal data into time weights solving method, and according to the distribution density of time series＇ optimal degree, figure out ＂Optimal degree＂ time weights solving method and the time weights solving method which integrates ＂time degree＂ and ＂optimal degree＂. Finally, it uses these two TOWA operators in different preferences to dynamic comprehensive evaluation and compare it with the results acquired from the original method and verify the efficiency and feasibility.

关 键 词：综合评价 TOWA算子 时序趋优度 信息集结 

分 类 号：C931[经济管理—管理学]