【作者】 李东民；

【导师】 钟佩思；

【作者基本信息】 山东科技大学 ， 机械电子工程， 2009， 博士

【摘要】 物流业在国民生产总值中占有重要的比重,我国已经将现代物流服务业列入国家“十一五”计划的重点发展产业。传统物流运作方式效率低下、灵活性差,已不适应物流行业发展的客观要求,物流信息化成为我国现代物流发展的必由之路。物流运作由于跨地域运作频繁、运作流程多变,导致一般信息系统很难满足其分布性和灵活性等需求。而Web服务特别适用于可变、灵活的分布式应用环境,能够满足现代物流灵活可变的多样化跨地域运作需求。将物流服务封装为Web服务并进行发布,通过服务匹配与选择,能够有效解决物流服务有效供给与有效需求之间的矛盾,进而提升物流服务水平。本文深入分析了现有物流服务匹配与选择及其关键支撑技术的研究和应用现状,结果显示物流服务匹配与选择存在许多问题,匹配与选择结果不理想。针对以上问题,通过深入研究物流业务运作和本体理论,建立了物流领域本体和物流服务质量本体实例;提出了物流服务语义标注方法;提出了基于本体概念属性距离的三级物流服务匹配算法;提出了基于联运服务质量综合评价体系的物流服务模糊选择算法;基于物流服务管理平台框架验证了算法在汽车制造企业集群的联运应用场景下的正确性。通过研究Web服务全生命周期的活动流程,建立了物流服务全生命周期框架;在此基础上根据物流业务运作过程,提出了物流服务匹配与选择框架;整个框架以物流领域本体、物流服务匹配算法及物流服务选择算法为核心和主线。分析和比较了分别采用语义标注Web服务描述语言SAWSDL和服务本体描述语言OWL-S进行物流服务语义标注的优劣,并选定SAWSDL进行语义标注;根据物流业务过程和物流服务分类体系,建立了物流领域本体和物流服务质量本体实例;提出并实现了基于SAWSDL采用物流领域本体实例进行物流服务语义标注的方法。物流服务匹配即物流服务需求模板与候选物流服务匹配的过程。根据匹配过程,建立了匹配模型,提出了基于本体概念属性距离的物流服务类别匹配、概貌匹配和接口匹配的三级物流服务匹配算法,将三级匹配过程转换为相应映射本体概念语义相关度问题。类别匹配转换为物流服务类别本体概念语义相关度,概貌匹配转换为WordNet本体词汇库中词汇概念语义相关度,接口匹配转换为物流领域本体概念属性语义距离计算问题;对匹配过程进行了形式化描述,给出了三级匹配的总匹配度计算方法。分析了物流服务质量中最重要的联运服务质量指标,包括联运服务业务质量和Web服务质量指标,建立了联运服务质量模型;针对联运服务业务和质量指标特征及对物流服务的评价方式和影响程度,提出了基于主观不确定性的联运服务质量综合评价体系。基于物流服务质量评价的不确定性,提出了采用梯形模糊数描述服务质量评价初始值的方法;根据物流服务匹配与选择框架,将物流服务选择转换为模糊多属性决策问题,并提出了基于联运服务质量综合评价体系的物流服务模糊选择算法。在物流服务管理平台框架下,设计并开发了物流服务匹配与选择原型系统并在汽车制造企业集群的联运场景下对物流服务匹配算法和选择算法进行了应用验证。验证得到了各级和总匹配度并选择出了最优物流服务。更多还原

【Abstract】 Logistics is a central part of modern economy. Modern logistics service industry has been especially emphasized in key development of The National Eleventh Five-year Plan. Traditional logistics operations are not flexible nor available for the logistic industry growth’s requirements. While informationization has been admitted as one of the primary means to normalize the operational process. Therefore, logistics informationization is necessary for development of modern logistics in our country. Due to the cross regions, changed processes, general information systems cannot satisfy the requirements of distribution and flexibility. Web service is available for alterable distributed circumstance and diversity of logistic operations across region. Firstly, logistics services are encapsulated in the form of Web services and published. Then services are matched and selected. Finally, logistics operations are managed dynamically and the comprehensive logistics service system is constituted.Current research and application in the filed of logistics services matchmaking and selection and the key technologies are analyzed in this dissertation and many problems are found. To solve above problems, logistics process operation and ontology are studied, then instances for logistics domain ontology and logistics services quality ontology are built. Besides, semantic annotations are proposed and accomplished. Furthermore, Distance of Ontology concept property based three-level Matchmaking algorithm for logistics services—DisOnMatch and SUQMulT based Fuzzy algorithm for Logistics services Selection—SUQMulT-FuzLS are proposed. Finally, DisOnMatch and SUQMulT-FuzLS are validated under multimodal transport among automobile manufacture enterpreise centralization scene based on service management platform framework.The framework of all life cycle for logistics services is abstracted from activities of all life cycle for Web services; The framework of logistics services matchmaking and selection is proposed according to logistics process operations, which pays more attention to logistics domain ontology, algorithms for both services matchmaking and service selection.Differences between SAWSDL and OWL-S for semantic annotation are studied and analyzed, and SAWSDL is adopted to annotate services. Instances of logistics domain ontology and logistics service quality ontology available are set up according to logistics operation process and logistics services category; The method for logistics services annotation with SAWSDL based on logistics domain ontology is proposed and accomplished.Logistics services are matched between services template and candidate logistics services. DisOnMatch is proposed according to matchmaking process. Three-level matchmaking are converted to semantic similarity on mapping ontology concepts by modeling for services matchmaking process; Category matchmaking is converted to semantic similarity on mapping concepts in logistics service category ontology. Profile matchmaking is converted to concepts similarity on vocabularies in WordNet ontology. Interface matchmaking is converted semantic distances of logistics domain ontology concept properties. After formal description for matchmaking process is finished, total matchmaking score is obtained.This dissertation studies the indexes of service quality for multimodal transport, which include qualities of multimodal transport process and Web services; A model on logistics service quality is built; Subject Uncertainty based comprehensive evaluation system on Quality for Multimodal Transport services—SUQMulT is proposed according to quality indexes, evaluation type and influence degree for logistics services.Logistics service qualities are evaluated by trapezoidal fuzzy number according the subject uncertainty; Logistics service selection is converted to fuzzy multi-attribute decision making according to the framework of logistics services matchmaking and selection; Also SUQMulT-FuzLS based on SUQMulT is proposed.DisOnMatch and SUQMulT-FuzLS are validated in a multimodal transport among automobile manufacture enterpreise centralization scene based on service management platform framework; Each level and total matchmaking scores are got by DisOnMatch, and then optimal logistics service is selected among the logistics services, whose matchmaking degrees are higher than threshold value of matchmaking score.更多还原