论文部分内容阅读
动态多属性决策问题包含有多时间阶段的属性值,反映了各时间节点对于决策对象的状态评估.基于模型预测控制思想,提出了一种基于相似关系的动态赋权重方法,通过灰色预测模型 GM(1,1)对属性值进行预测,对集成权重实施滚动优化,从而有效地解决动态多属性决策问题,实例分析表明了所提方法的有效性,并通过预测模型精度验证方法的可靠性.“,”Dynamic multi-attribute decision making problems usually involve multi-time attribute values which represent the state assessments of the decision-making objectives over the time horizon. Inspired by the model predictive control philosophy, a dynamic weighting method based on similarities is established in this paper. Combining the GM(1,1) grey model based predictions for attribute values along with rolling optimization strategies, the dynamic multi-attribute decision-making problems are effectively resolved. Case studies illustrate the benefits of the contribution.