Evaluation of urban taxi-carpooling matching schemes based on entropy weight fuzzy matter-element
Evaluation of urban taxi-carpooling matching schemes based on entropy weight fuzzy matter-element
复制标题
基于熵权模糊物元的城市出租车拼车匹配方案评价
DOI：
10.1016/j.asoc.2019.105493
复制
发表时间：
2019-08-01
影响因子：
8.7
通讯作者：
Zhang Wei
中科院分区：
文献类型：
--
作者：
Xiao Qiang;He Ruichun;Zhang Wei
城市出租车拼车匹配度评价是出租车拼车成功与否的关键指标。本文引入物元理论。建立了物元评价模型，提出了基于模糊熵权的物元评价算法。该算法根据匹配数据的特点，将拼车匹配方案中到拼车目的地的距离、到拼车起点的距离、拼车距离和到拼车起点的时间作为评价类别的特征指标。通过指标模糊值法建立复合模糊物元矩阵，构造优选隶属度物元矩阵和标准差模糊物元矩阵。通过熵权法确定指标权重，并计算物元矩阵的贴近度。该方法避免了传统物元分析方法中人工确定特征值指标和计算关联度的缺点。实验结果表明，该算法的评价结果与实际评价结果较为接近。此外，该算法的评价效果优于传统的物元法。(C)2019 Elsevier B.V.版权所有。
The evaluation of urban taxi-carpooling matching is the key index for the success of taxi carpooling. In this research, matter-element theory is introduced. Moreover, a matter-element evaluation model is established, and a matter-element evaluation algorithm based on fuzzy entropy weight is proposed. In accordance with the characteristics of the matching data, the algorithm takes the distance to the carpool destination, the distance to the carpool starting point, the carpool distance and the time to the carpool starting point in the passenger-matching scheme as the characteristic indexes of the evaluation category. A compound fuzzy matter-element matrix is established through an index fuzzy value method, and a matter-element matrix of a preferred membership degree and a fuzzy matter-element matrix of standard deviation are constructed. An index weight is determined through entropy weight method, and the closeness degree of the matter-element matrix is calculated. It avoids the shortcoming of traditional matter-element method which uses artificial determination of an eigenvalue index and correlation degree calculation. Experimental results show that the evaluation result of this algorithm is close to the actual evaluation result. In addition, the evaluation effect is better in this algorithm than in the traditional matter-element algorithm. (C) 2019 Elsevier B.V. All rights reserved.