Abstract: The current individual decision-making scheme in shared resource matching does not consider the interaction,which makes the decision unstable and results in decision competition and decision failure.To address the problem,based on the fact that the interactive individual decisions in the decision-making time window are group decision-making,this paper uses the group decision-making theory was used to construct a group decision-making method for the selection of decision-making time point in resource matching by considering each individual decision-making competitor as a group decision-making member of effective decision-making.The combination of individual decisions was regarded as the alternative of group decision-making,and the psychological preference coefficient was established to describe the degree of attribute preference of decision-makers at different time points by considering the influence of time factors in the decision-making process on the psychological behavior of decision-makers,so as to modify the decision information and optimize the decision-making.Based on Probabilistic Linguistic Term Sets (PLTSs),language evaluation information and numerical evaluation information were fused in the form of probability,and the probability calculation formula of evaluation value was given in combination with psychological preference coefficient.The evaluation value probability sets were combined with the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) to sort the decision schemes,and the optimal decision scheme under group consensus was obtained,which proved to be the best decision-making time point set of all decision makers.The effectiveness of the model was verified by taking the vehicle search in the logistics information platform as an example.The proposed method could provide direction for the platform to solve the matching competition and implement decision control and resource scheduling.

Key words: shared resource matching, matching decision, decision time window, group decision-making, decision organization and optimization, probabilistic linguistic term sets, technique for order preference by similarity to ideal solution

摘要： 针对共享资源匹配中个体决策方案未考虑个体决策相互影响且不稳定,导致决策竞争和产生决策失效的问题,基于决策时间窗内相互影响的个体决策实质为群决策的特征,引入群决策理论,将各个体决策竞争者视为有效决策的群决策成员,构建了资源匹配中关于决策时间点选择的群决策模型。以各个体决策的组合为群决策备选方案,考虑到决策过程时间因素对决策者心理行为的影响,建立了心理偏好系数描述不同时间点的决策者对属性偏好的程度,以此修正决策信息、优化决策。基于概率语言术语集(PLTSs)以概率形式融合语言评价信息和数值评价信息,并结合心理偏好系数给出了评价值的概率计算公式。最后将评价值概率集合与逼近理想解排序法(TOPSIS)结合,对各决策方案排序,求出群体共识下的最优决策方案,该决策方案为所有决策者的最佳决策时间点集合。以物流信息平台中的车找货为例验证了模型的有效性。所提出的方法可为平台解决匹配竞争而实施决策控制以及资源调度提供方向。

关键词: 共享资源匹配, 匹配决策, 决策时间窗, 群决策, 决策组织优化, 概率语言术语, 逼近理想解排序法