论文部分内容阅读
针对传统语言群决策方法专家权重难以合理求取且决策属性值为不确定语言变量的问题,提出一种基于蒙特卡洛经验模态分解(Mentor Carlo-Empirical Mode Decomposition,MC-EMD)提取专家语言评价信息的多属性大群体决策方法。考虑专家期望偏差越小为宜,建立偏差最小单目标优化模型求解属性权重;运用EMD方法分解各专家的综合语言评价值,得到客观趋势成分和主观随机成分,以客观趋势成分的均值作为评价结果;鉴于不同专家顺序可能有不同的分解结果,从而导致评价结果的不确定性,基于蒙特卡罗思想随机抽取专家排序,通过计算模拟获取专家评价的总体客观趋势,并借以进行方案优选排序。案例分析验证了该方法的有效性和可行性。
Aiming at the problem that traditional expert group decision-making method is difficult to reasonably weigh and the decision attribute value is uncertain linguistic variables, this paper proposes an MC-EMD-based expert language extraction method based on Monte Carlo empirical mode decomposition (MC-EMD) Multi-attribute large group decision-making method for evaluating information. Considering that the expert expectation deviation is smaller, it is reasonable to set up the minimum deviation single objective optimization model to solve the attribute weight. Using the EMD method to decompose each expert’s comprehensive linguistic assessment value, the objective trend component and the subjective random component are obtained, and the average of the objective trend component is taken as the evaluation result In view of the fact that the order of different experts may have different decomposition results, leading to the uncertainty of the evaluation results, the experts are randomly selected based on Monte Carlo’s thought, and the general objective trend of expert evaluation is obtained through calculation and simulation, and the priority of the programs is sorted. Case studies verify the effectiveness and feasibility of this method.