基于RFID异常事件检测和转移信用模型的热带农产品可视化智能防伪新方法及案例研究

项目名称： 基于RFID异常事件检测和转移信用模型的热带农产品可视化智能防伪新方法及案例研究

项目编号： No.61462023

项目类型： 地区科学基金项目

立项/批准年度： 2015

项目学科： 自动化技术、计算机技术

项目作者： 姚孝明

作者单位： 海南大学

项目金额： 45万元

中文摘要： 假冒伪劣产品猖獗需求新的智能防伪方法。基于RFID的热带农产品供应链通过RFID自动识别、追踪、追溯及定位等特性构建虚拟信息世界中农产品的可见性来获取对其物理运动的有效监控及防伪能力。目前对这种可见的真实性的信任评估方面等深入研究仍然缺乏。本项目拟根据常见农产品供应链运营管理场景，突出读入错误与人工篡改关联逻辑差异，采用异常事件检测技术和转移信用模型，对RFID数据模型、事件模型、复杂事件检测模型以及供应链各节点的可见性信用评估模型展开研究，由此提出基于时序推理的农产品RFID供应链智能防伪新方法(包括推理规则与核心算法)以及适于海南案例的解决方案，最后通过搭建简易原型系统实地模拟等方法来验证其模型与算法的可行性和有效性。所得研究成果不仅为农产品RFID供应链各环节可能出现虚假的真实提供可靠的信用依据，也为海南农产品供应链信息服务平台满足用户随时随地进行可视化智能防伪检测奠定应用基础。

中文关键词： 数据安全；数据流；RFID；异常事件检测；防伪

英文摘要： That fraud transactions and counterfeits are rampant in current open and competitive market calls for novel intelligent anti-counterfeiting technologies.The RFID-based supply chain system for tropical agricultural products attempts to build up the visibility of the agricultural products in the virtual information world by exploiting the RFID features such as contactless identification, tracking, tracing and locating to acquire the power of anti-counterfeiting and effectively monitoring, managing their physical movements. However, the in-depth researches regarding the aspects such as how to decide if this intuitional visibility is reliable are rare up to now, few work about the model of measuring or evaluating such visibility was ever reported in public literatures available. Therefore, in this project, it is planned that we start our research on RFID data model, event model, and complex event detection model, and consequently the belief index of the visibility and its evaluation model by using abnormal event detection technology and transferable belief model in applications of typical scenarios for supply chain systems of agricultural products, focusing on discriminating the logic difference with respect to the relations between the intentionally tampering and the regular cases when RFID data readings might be incorrect or uncertain, and subsequently on this basis, propose the novel temporal inference-based intelligent anti-counterfeit methods (including inference rules and core algorithms) and solutions of case studies specific for Hainan agricultural markets. The effectiveness and feasibility of the methods, models and algorithms are validated by simultation experiments in a simplified prototype system specially developed for this research in several real supply chain nodes carefully chosen. The results derived from this research will not only provide with a reliable evaluation model regarding the belief index of the visibility of the RFID supply chain for the agricultural products where the faked real can be most likely found, but lay a technological foundation for Hainan local applications of information service platform for the supply chain of tropical agricultural products to satisfy the users who want to visually self-validate the product with the intelligent anti-counterfeit detection anywhere and anytime.

英文关键词： data security;data streams;RFID;abnormal event detection;anti-counterfeiting