面向多任务的非线性群决策系统多模型匹配算法  

Multi-task Oriented Non-linear GDSS Multi-models Matching Algorithm

作　　者：曹玉辉[1] 王卫红[1] 刘超飞[1] 张森[1] 

机构地区：[1]西安交通大学电子与信息工程学院

出　　处：《计算机应用研究》2007年第1期37-39,共3页Application Research of Computers

基　　金：国家"863"计划资助项目(2003AA412020);陕西省科学技术发展计划"十五"攻关资助项目(2000K08-G12)

摘　　要：针对多任务的非线性群决策系统提出了多模型匹配算法。为了对各种不同性质的任务进行理性决策,首先根据不同任务的特性选取针对不同任务的满意决策群体,构成动态满意决策群体集;在此基础上为每个决策任务构造多个决策方案模型集,然后根据方案模型的匹配程度和决策者的决策权重以及决策的目标权重来控制决策过程的动态变化,从而得到最优匹配决策方案模型;最后给出该算法的应用实例,实践表明该算法是有效的,能够灵活选取模型库中的不同模型,提高决策的速度和精度。Aiming at multi-tasks non-linear GDSS, a multi-models matching algorithm was provided. In order to rationally make decisions to tasks with all kinds of character, the paper selects a satisfactory decision-making group fitting different tasks based on the character of different tasks. Then, build a multi-schemes model set for each decision-making task, and control the change of decision-making process based on the matching degree of scheme model and decision-making weight of decisionmaker. So, an optimized decision-making model was obtained. At last, an application instance was given about the algorithm. The results indicate the algorithm is effectually, it can choose flexible different models from model database and can improve the speed and precision of decision-making.

关 键 词：多模型 群决策支持系统 模型匹配 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]