【作者】 李婷婷；

【导师】 吴中如；

【作者基本信息】 河海大学 ， 水工结构工程， 2006， 博士

【摘要】 本文集成多学科知识，将现代计算机、通信、可视化、人工智能、管理学等多类知识综合应用于大坝安全监控领域的研究中，深入探讨了研制混凝土坝健康诊断及其预警系统的实现方法和技术。主要研究内容如下： (1)探讨了大坝健康诊断及其预警系统的功能和总体结构体系，同时，以先进的UML建模语言为基础，并考虑大坝健康诊断及其预警系统的特点和实际情况，研究了系统的多维软件体系结构。 (2)构建了混凝土坝结构健康诊断的结构体系。据此，以混凝土坝结构健康状况为诊断目标，依据诊断理论的思想，利用粗糙集理论、模糊逻辑、神经网络等多领域知识，进行有机智能融合，提出了基于粗糙集的模糊神经网络健康综合诊断方法。 (3)建立了基于数据仓库的大坝安全监测信息的智能集成结构，从多个角度对不同类型的大坝监测信息进行智能集成，使得大坝的各种信息的潜在价值能够得到充分地发挥。 (4)提出了融合小波分析对大坝数据仓库数据预处理的方法，进而更加有效地提高了大坝数据仓库中数据预处理的效率和数据信息的质量。同时，研究了利用数据挖掘技术的关联分析模式，挖掘出蕴藏于大坝数据仓库中的潜在有用知识。 (5)探讨了利用现代可视化技术提高大坝安全管理水平的方法。将可视化数据挖掘技术应用到大坝健康诊断及其预警系统的可视化中，通过可视化技术将大坝监测数据进行三维立体可视化，并采用OLAP技术对其进行多维分析，以求更加有效地剖析数据。 (6)综合上述研究成果，利用数据挖掘技术和可视化技术，以古田溪一级大坝为例，开发了福建省大坝健康诊断及其预警系统。更多还原

【Abstract】 In order to realize health diagnosis and early-warning system for concrete dam, many discipline knowledge are integrated for the research of dam safe monitoring, such as computer technique, communication, visualization, artificial intelligent, management, and so on. The main contents are as follows.(1) The general function and structure of Dam Health Diagnosis Early-Warning System (DHDEWS) is studied. According to the characteristic of dam safety monitoring, multi-dimensional software structure of DHDEWS based on UML is obtained.(2) The structure system of comprehensive health diagnosis for concrete dam is built. The method of comprehensive health diagnosis for concrete dam based on rough sets and fuzzy neural network is purposed according to diagnosis, rough sets, fuzzy logic and neural network.(3) Intelligent integrated structure of dam safety monitoring information based on data warehouse is offered. In order to make full use of the dam safety monitoring information, it is intelligent integrated in various views.(4) The data pre-treatment combined wavelet with data warehouse is purposed. The data pre-treatment of data warehouse is more efficient and the quality of information is improved. In the same time, the association analysis of data mining is used to mine the hidden available knowledge in dam data warehouse.(5) The method of improving dam safe management based on visualization is researched. Visual data mining is studied for visualization of dam health diagnosis and early-warning System. Three dimensional visualization of dam monitoring information is realized. In the same time, on line analytical processing technique is used for the more efficient analysis of data.(6) Data mining technique and visual technique is used to develop dam health diagnosis and early-warning System for Fujian province.更多还原