船舶智能信息管理系统的构造及其方法

【作者】 张劲；

【导师】 李海量； 汤天浩；

【作者基本信息】 上海海运学院 ， 电力电子与电力传动， 2002， 硕士

【摘要】 在科学技术空前辉煌和科学理性充分发展的今天，信息论所引发的信息技术革命浪潮方兴未艾，它正深刻地变革着我们的生活方式与生产方式。本文试图在这样一个大背景下，就航运管理系统的信息处理问题展开探讨。 本文在强调必要性的基础上阐述了构造航运智能信息管理系统的总体思路，并就其各个子系统作了具体的讨论，总结了将众多原本相对独立的子系统逐级地组成航运智能信息管理系统的过程所依赖的三个关键技术：计算机集成技术、信息集成技术和智能集成技术，并以机舱智能信息管理子系统为例，进一步阐述了三个集成技术的细节，包括现场总线控制技术、数据库技术、信息综合显示技术、趋势预测与故障诊断技术，等等。 作为重点，本文辟用了较大的篇幅讨论面向应用(主要是趋势预测与故障诊断)的集成智能信息处理技术，在介绍相关领域的发展情况和总体格局之后，重点阐述了几种基于神经网络的智能算法，包括基于对角递归神经网络(DRNN)的直接多步非线性自回归滑动平均(NARMA)预测模型，以及基于广义概率和(GPS)与广义概率积(GPP)两种算子的模糊神经网络模型，给出了它们的详细算法及面向应用的运算步骤。 文章最后，介绍了在理论研究的基础上研制开发的基于集成智能的船舶机舱信息管理系统软件，并且对该软件的集成环境、结构框架和功能实现等作了详尽的剖析。本文将基于模糊技术、专家系统和人工神经网络的智能集成方法应用于复杂系统的故障诊断和预测分析，进而在自动集中监测系统的基础上，建立了以集成智能为核心，以诊断知识为根本，以辅助决策为手段的具有故障自动发现、辅助诊断、趋势预报等功能的船舶机舱信息管理系统。其研究成果对于整个航运智能信息管理系统来说，具有普遍意义。更多还原

【Abstract】 In the past century, a breakthrough has been made in science and technology around the globe, and the revolution of information technology is still deeply affecting our manufacturing and life style. Under such a background, this paper attempts to discuss those problems related to the information processing of Marine Management System.This paper states the idea on the structuring of the Marine Information Intelligent System after emphasizing its necessary, discusses specifically every sub-systems, and concludes the three key technologies that support thecombination of sub-systems——computer integrated tech., information integratedtech. and intelligence integrated tech. By the example of Engine Room Information Intelligent System, this paper states fatherly the details of three integrated technologies which include field-bus control tech., database tech., information integrated display tech., trend prediction and fault diagnosis tech., etc.The information intelligent processing technology facing the application is presented as an emphasis. After introducing the development situation and the whole pattern on related fields, this paper describes several algorithm applied in the simulation experiment, including direct multi-steps nonlinear autoregressive-moving average(NARMA) prediction model based on diagonal recurrent neural networks and fuzzy neural networks model based on Generalized Probability Sum(GPS) and Generalized Probability Product(GPP), and lists the algorithm steps facing the application.At last the paper describes the software of Engine Room Information Intelligent System, which is developed on the base of theory research. We expounded upon the external supporting, the structure and the function of the software. This simulation experiment applies the integrated intelligence bases on fuzzy logical,expert system and artificial neural networks to solve the problems of fault diagnosis and prediction analysis of the complicated system. Then on the basis of the Integrated Monitoring System, we build the Engine Room Information Intelligent System that is based on the diagnosis knowledge, centered with integrated intelligence, and by the way of aided decision, the system provides the functions of fault auto-detection, aided diagnosis, trend prediction and etc. The fruit of research has general meaning to the whole Marine Information Intelligent System.Zhang Jin (Power Electric and Electronic Drive) Directed by Prof. Li Hailianq and Prof. Tang Tianhao更多还原