基于神经网络的船舶冷藏集装箱故障远程监控、预警及诊断研究

【作者】 舒象海；

【导师】 郑学林；

【作者基本信息】 上海海事大学 ， 制冷与低温工程， 2007， 硕士

【摘要】 随着世界经济全球化速度步伐的加快,信息技术的快速发展,以信息技术为代表的高新技术的变化日新月异,广泛并且深远地影响了世界各国的经济与社会,经济全球化和信息网络化相互依存、相互促进的关系愈加明显。随着经济全球化步伐的加快,世界各国之间的贸易往来愈加频繁,由于冷藏集装箱技术先进,保鲜期长,能将鱼肉、果蔬和鲜花等货物始终保持在各自最佳的低温状态,可灵活、可靠、方便地将其运送到世界任何地方。因此,它迅速发展成为国际贸易中的一种重要运输手段。但是,由于冷藏集装箱是一个流动的冷库,其营运是一个系统工程,稍有一个环节出现差错就会造成巨大的经济损失。在冷藏集装箱运输量增大的同时,由于机组机械故障、货物储藏不当、未有效控制好冷藏温度、货物质量差等原因给航运公司带来了巨大的经济损失和信誉损失。为了减少货差货损、明确权责及降低航运公司运输成本,所以基于信息化建立一个冷藏集装箱的远程故障诊断系统是非常必要的,本文将依次为背景,利用涉及到计算机网络、信息集成、专家系统等多门学科的综合技术——远程故障监控、预警及诊断技术,基于神经网络开发了船舶冷藏集装箱故障远程监控、预警及诊断系统。本文分为五个部分,第一部分,提出了该课题的研究背景、研究意义及研究内容。第二部分,指出了远程故障诊断技术、人工神经网络应用于故障诊断技术和冷藏集装箱的监测系统的国内外研究现状。第三部分,介绍了人工神经网络,并基于BP神经网络建立了BP神经网络故障诊断模型;介绍了远程故障诊断技术及结构,并建立了船舶冷藏集装箱故障远程监控、预警及诊断系统的体系结构。第四部分,介绍了船舶冷藏集装箱的制冷系统,并利用故障树分析法对其故障进行了调查与分析,利用上海海事大学冷藏集装箱试验平台进行样本采集,通过资料查阅,结合调查与分析的结果建立了船舶冷藏集装箱故障知识库。第五部分,利用组态王6.5.1开发了基于BP神经网络的冷藏集装箱故障监控、预警及诊断系统,本软件可以对冷藏集装箱参数进行监控、并对故障进行预警,当故障出现时,可以通过查询故障数据知识库找出解决办法。更多还原

【Abstract】 With the quickening step of the globalization of world economy and the great development of information technology, especially the representative of IT of the advanced technology develops changing with each passing day. It influences the world economy and social life widely and profoundly. The globalization of world economy and information network are interrelated and mutually dependent obviously, with the development of world economy globalization, the trade of different nations becomes more and more frequently. Because of the advanced refrigeration container, it can keep fish , meat ,fruit .vegetable ,and fresh flowers in the optimum temperature so to be shipped to anywhere in the world frequently and convinently. So the great development of refrigeration container has become an important transportation means in international commercial.But because refrigeration container is a mobile cold storage if the mistake of one part will bring great loss, so its transportation is a systematical project. With the increasing transportation of refrigeration container, if mechanism problem, improper goods storage, not controlling the temperature effectively and the bad quality goods happens it will bring great financial loss and bad reputation to the shipping company. In order to reduce the damage of goods, make sure the responsibility, and reduce the transportation cost, it is necessary to establish a long-distance failure diagnosis system of refrigeration container. The article uses the multi-technology of computer network, integrated information, experts system-long-distance failure controlling, forewarning and diagnosis technology the long-distance failure controlling of refrigeration container shipping, forewarning and diagnosis system.The article is divided into 5 parts. The first part puts forward the background, significance and content of this problem. The second part shows the internal and external current research situation of long-distance failure diagnosis technology, the failure diagnosis technology by man-made nerve network and the controlling system of refrigeration container The third part introduces the man-made nerve network and establishes BP nerve network failure diagnosis model on the basis of BP nerve network and also introduces the technology and structure of long-distance failure diagnosis and establishes the long-distance controlling of refrigeration container shipping, forewarning and diagnosis system. The fourth part introduces the refrigeration container shipping’s refrigeration system and investigates and analyzes the failure by using the analyzing way of failure tree. It establishes the refrigeration container shipping failure information data base by using the refrigeration container experimental flat in Shanghai Maritime University. The fifth part develops the refrigeration container failure controlling, forewarning and diagnosis system on the basis of BP nerve network and using 6.5.1. this software control the refrigeration container parameter, forewarn the failure and solve the problem by inquiring the failure information data base.更多还原