主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2014, Vol. 22 ›› Issue (11): 79-86.

• 论文 • 上一篇    下一篇

基于最小偏差组合权重的突发事件应急能力评价模型

刘德海1,2, 于倩1,2, 马晓南1,2, 尹丹1,2, 王维国1,2   

  1. 1. 东北财经大学经济计量分析与预测研究中心, 辽宁 大连 116025;
    2. 东北财经大学数学与数量经济学院, 辽宁 大连 116025
  • 收稿日期:2013-07-12 修回日期:2014-03-08 出版日期:2014-11-20 发布日期:2014-11-21
  • 作者简介:刘德海(1974-),男(汉族),辽宁辽阳人,东北财经大学数学与数量经济学院,教授,研究方向:演化博弈理论、社会突发事件问题.
  • 基金资助:

    国家自然科学基金资助项目(70901016,71171035,71271045)

Combination Weight Model of Emergency Ability Evaluation with Minimum Deviation

LIU De-hai1,2, YU Qian1,2, MA Xiao-nan1,2, YIN Dan1,2, WANG Wei-guo1,2   

  1. 1. Center of Econometric Analysis and Forecasting, Dongbei University of Finance & Economics, Dalian 116025, China;
    2. School of Mathematics and Quantitative Economics, Dongbei University of Finance & Economics, Dalian 116025, China
  • Received:2013-07-12 Revised:2014-03-08 Online:2014-11-20 Published:2014-11-21

摘要: 在突发事件应急能力评价问题中存在着不同评价方法的冲突性,即群决策评价结果的方法依赖性问题。本文以高速公路突发事件应急能力评价为研究对象,提出了基于最小偏差组合权重的评价模型。首先,系统地分析了高速公路突发事件应急能力的评价问题,建立了基于过程管理的高速公路突发事件应急能力评价的指标体系。根据语言评价指标分别运用二元语义层次分析法、G1主观赋权法和基于熵值法的二元语义赋权法等三种方法确定各评价指标的权重,然后综合各种评价方法构建了基于最小偏差的组合权重线性规划模型,确定准则层的最优复合权重,得出高速公路应急能力的评价值。最后,以上海"9·14"重大道路交通事故作为应急能力评价的分析案例,比较了不同评价模型的结果,并对评价指标进行敏感性分析。研究结果表明,基于最小偏差组合权重的评价模型能够在多个评价模型中达到最大的一致性,从而有效地减少了群决策中不同主客观评价方法的"极化"效应。

关键词: 应急能力评价, 突发事件, 最小偏差, 组合权重, 多属性群决策, 极化效应

Abstract: Emergency ability evaluation has a problem with the result conflict among the different evaluation methods, that is to say, the evaluation result of group decision depends on the evaluation method adopted. Emergency ability evaluation of accident occurred on expressway is studied in this paper. There have some problems that deserve some special consideration. Because the emergency rescue time gets tight, and road needs to be cleared as soon as possible, the index system of emergency ability evaluation in accident occurred on expressway should consider those factors. The evaluation method should deal with the linguistic information of expert evaluation. Furthermore, considering result differences among various evaluation methods, there has an optimization problem among various evaluation methods. A combination weight model of emergency ability evaluation is built, which is constituted of the linear combination of series of evaluation methods including 2-tuple linguistic AHP, G1 subjective weighting method, and 2-tuple linguistic entropy evaluation method, according to the minimum deviation principle. Along with the research idea of "the index system of emergency ability evaluation-various evaluation methods and corresponding combination weight model-case analysis", the paper mainly contains the following three parts. Firstly, the emergency ability in accident occurred on expressway is systematically analyzed and the index system of emergency ability evaluation in accident occurred on expressway is built based the idea of process management, Which includes the criterion layer with four first-level indexes and the index layer with 22 second-level indexes. Secondly, based on the linguistic assessment indexes, 2-tuple linguistic AHP, G1 subjective weighting method, and 2-tuple linguistic entropy evaluation method are used to compute the index weights, and then the linear programming model of combination weights with minimum deviation is built, which integrates above evaluating methods, to compute the optimal composite weight of the criterion layer. So the emergency ability evaluation value is obtained in accident occurred on expressway. Shanghai expressway emergency ability of "9·14" major road traffic accident is evaluated by the above model, and the different results derived from four evaluation methods are compared. In addition, the sensitivity analysis of evaluation indexes is discussed. Results show that the combination weight model with the minimum deviation principle achieves the maximum degree of consistency among different evaluation methods, and reduces "the polarization effect" of various evaluation methods in group decision, which can be applied into the field of emergency ability evaluation. The combination weight model with the minimum deviation principle can be not only applied to optimization problem among other various evaluation methods, but also other types of multiple attribute group decision making problems.

Key words: emergency ability evaluation, accident, minimum deviation, composite weight, multiple attribute group decision making problem, polarization effect

中图分类号: