群决策专家权重自适应算法研究.pdfVIP

第 28 卷第 2 期 计算机应用研究 Vol. 28 No.2 2011 年 2 月 Application Research of Computers Feb.2011 群决策专家权重自适应算法研究* 王俊英 l 气李德华 1 (1.华中科技大学图像识别与人工智能研究所，武汉 43∞74; 2. 三峡大学计算机与信息学院，湖北宜昌 443则) 摘 要:为获得科学合理的群体决策结果，并在综合集成研讨厅中维护一个动态的专家权重信息库，提出了丰 种基于 TOPSIS 法和灰色关联度的专家权重自适应调整算法，算法综合考虑了专家个体与群体在方案重要属性 以及方案序的判定上的相似度，并根据决策问题的不同，对算法的参数进行相应选取，算例表明算法是有效的， 收敛速度也比较快。 关键词:多属性群决策:专家权重;逼近理想解排序法:决策关联度;自适应调整 中圄分类号: 1白91 文献标志码 :A 文章编号: 1 ∞ 1-3695(2011)02-0532-04 doi: 10. 3969/j. issn. 1 ∞ 1-3695.2011.02.035 Research on self-adjusting algorithm of group decision-making experts weight WANG Jun_yingl.l , U De-hua1 ( 1. Institute for p幽， mReω， 阴阳 A耐性也1 /，脚M帆风 H1UUhong Univer. 仰 0/ 协脚 Technology. Wuhan 43∞74 ， Chína; 2. College ?fCo.呻阳 lnfonootion Technology , China Three 臼ges Univer哟 ， Yicha略 Hubeí 443棚 . China) Abstract: ln order to obtain scientific ratíonal 伊Oup decision-making results and maintain a dynamic experts weight infor? mation base in HWME ( hall for workshop of meta synthesis engineering) , this paper proposoo an experts weíght 回 lf-adaptive adjustment algoríthm bωed on TOPSIS meth叫 and giey cQlTelation degree. 咀 is algorithm took into account the símilarity be? tween the individuals and groups of experts in the judge on the important attríbutes and order of the altematives. Meanwhile , ac∞rding to different decision-making problems , this algorithm could select corresponding parameter. Numerical example shows that the algorithm is effective and its convergence rate is relatively fa