Abstract: In order to improve the loading rate of whole vehicle transportation and the satisfaction of cargo owners, a vehicle-cargo matching method based on view similarity and considered the evaluation of cargo owners was proposed, which was based on case retrieval thinking and hesitancy fuzzy comprehensive term set method in case based reasoning (CBR) technology. According to the case retrieval thinking, the cargo condition attribute set was regarded as the problem to be solved (quasi case), the vehicle condition attribute set was regarded as the condition attribute set of existing case, and the trust attribute index of the vehicle owner was retained. Based on the adaptive K-means method, the nearest cluster was determined, and only the vehicle data stored near the cluster center within the nearest cluster was retained, and the matching space was double compressed. The entropy weight method was used to determine the weight of conditional attributes, and the view similarity between the goods data set to be matched and the vehicle data set in the matching space was calculated based on the European distance. The calculation results were arranged in descending order and the vehicle owner information corresponding to the TOP 3 vehicle data sets was retained. Based on the hesitant fuzzy linguistic term sets (HFLTS) method, the fuzzy evaluation information of the cargo owner's hesitation to the vehicle owner was transformed into a specific value through the score function, and the view similarity and evaluation results were comprehensively taken into account to recommend the optimal matching vehicle owner for the cargo owner. Through a case study, it is verified that the matching efficiency of the proposed method is increased by 88.79% compared to conventional methods, and the ratio between the goods to be matched and the recommended vehicle condition attributes is the maximum, which effectively improves the vehicle loading rate.

摘要： 为提高整车运输装载率及货主满意度，基于案例推理(CBR)技术中的案例检索思维和犹豫模糊综合术语集(HFLTS)，提出了一种基于视图相似度并考虑货主评价的车货匹配方法。依据案例检索思维，将货物条件属性集视为待解问题(准案例)，将车辆条件属性集视为既有案例条件属性集，保留车主的信赖属性指标；基于自适应K均值法确定最近聚类，只保留最近聚类内堆积在聚类中心附近的车辆数据，双重压缩匹配空间；通过熵权法确定条件属性权重；基于欧式距离计算待匹配货物数据集与匹配空间中的车辆数据集之间的视图相似度，降序排列计算结果并保留TOP 3车辆数据集所对应的车主信息；基于HFLTS方法将货主对车主犹豫模糊的评价信息通过得分函数转化为具体的数值，并综合考虑视图相似度及评价结果为货主推荐最优匹配车主。通过某案例分析，验证了该方法的匹配效率比常规方法提升了88.79%；待匹配货物与推荐车辆条件属性比值最大，能有效提升车辆装载率。