基于模糊理论和D-S推理的开箱报警模型与算法  

Alarm model and algorithm for container-opening based on fuzzy theory and D-S reasoning

作　　者：周鹏飞[1] 王开放[1] 方金灿[1] 

机构地区：[1]大连理工大学建设工程学部,辽宁大连116024

出　　处：《沈阳工业大学学报》2014年第2期210-217,共8页Journal of Shenyang University of Technology

基　　金：国家自然科学基金资助项目(71101014);大连市科技计划资助项目(20111061)

摘　　要：为提高集装箱运输过程中非法开箱自动报警的可靠性,在对集装箱非法开箱报警机制分析的基础上,建立了基于箱门传感器、箱内传感器等多源信息的非法开箱报警决策模型,提出了基于证据模糊相似度和证据源先验可靠度修正证据权的改进D-S证据融合算法来减少证据冲突,形成了集装箱非法开箱多源信息融合报警系统.实验结果表明,提出的证据融合算法能够有效降低证据冲突系数和合成悖论次数,报警模型具有较好的容错性和适应性.To improve the reliability of automatic alarm for illegal container-opening during the container transportation, an alarm decision-making model for illegal container-opening based on such multi-source information as the container door sensors and container inside sensors was established with the analysis on the alarm mechanism of illegal container-opening. An improved D-S evidence fusion algorithm based on the revised evidence weight with both evidence fuzzy similarity and evidence source prior reliability was proposed to reduce the evidence conflicts. An illegal container-opening alarm system with the multi-source information fusion was realized. The experimental results show that the proposed evidence fusion algorithm can effectively reduce the evidence conflict coefficients and fusion paradox times. Furthermore, the proposed alarm model has good fault tolerance and adaptability.

关 键 词：集装箱运输 非法开箱 自动报警 多源信息 D—S证据理论 模糊理论 证据权修正 信息融合算法 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]