【作者】 高占凤；

【导师】 杜彦良；

【作者基本信息】 北京交通大学 ， 车辆工程， 2010， 博士

【摘要】 结构健康监测的目标是对结构的整体行为实现实时在线监测与预测预报,要达到这一目标监测系统必须具有快速大容量的信息采集、传输和处理能力,并能够实现数据的网络共享。其中高速数据获取与传输,海量数据的存储与管理,监测数据的预处理、解读、分析和利用等是建立长期健康监测系统的关键技术问题。本论文针对上述关键技术问题开展结构健康监测中信息获取与处理的智能化研究,主要研究结构健康监测系统中的网络化数据采集与传输技术、智能信息处理技术、动态数据管理与查询技术。(1)首先提炼出建立结构健康监测系统的指导思想,构建具有普遍意义的大型结构长期健康监测系统的体系结构；自主开发了光纤传感测试系统、基于网络数据采集的电致传感测试系统和数字温度传感测试系统,构成“小集中”分布式数据采集与传输系统,以此为基础构建了大型结构健康监测系统通用化硬件平台；开发了基于虚拟仪器技术的以数据库系统为核心的软件平台,完成对硬件系统的集成。(2)研究了小波分析在实际结构监测信息处理中的应用。提出了一种改进的小波阈值降噪算法,该阈值函数表达式简单,能够更好地改进滤波效果,提高降噪质量；对比分析了分别以小波包系数节点能量值和傅里叶变换后的固有频率值作为损伤特征值的敏感度,分析结果表明以小波包系数节点能量值作为特征指标时敏感度更高,证明了利用小波包进行结构健康监测中的特征值提取是可行的。(3)研究了在实际结构监测信息处理中利用BP神经网络进行损伤识别的技术。首先建立BP网络模型,对模型进行性能优化,利用性能优化后的神经网络模型对混凝土斜拉桥模型的损伤状况进行了诊断,损伤识别结果表明通过结构振动模态频率的改变很容易地识别出结构的损伤情况。(4)自主开发了数据库管理与查询系统,采用标准关系型数据库技术对监测数据进行规范化统一管理,解决海量数据的存储、分析处理、查询、数据备份、数据安全等技术问题。(5)上述研究成果已成功地应用于芜湖长江大桥和郑州黄河大桥两座具有标志性意义的重要桥梁的监测系统中。两座大桥监测系统运行情况表明：自主开发的基于光纤应变测试系统、电致传感测试系统和数字温度传感测试系统的“小集中”分散数据采集系统实现了信息获取和传输的智能化；所采用的小波分析技术成功的完成对监测数据的降噪处理和特征值提取,所构建的BP网模型有效地对结构进行了损伤识别,自主开发的数据管理与查询系统能实时存储监测数据、实时显示查询结果、查询方便快捷,以上技术实现了对监测数据的实时分析、处理、存储及查询,提高了信息处理的智能化程度。研究成果将进一步丰富结构健康监测系统的工程应用理论和试验基础,具有较大的理论意义和工程实用价值,将产生巨大的社会经济效益。更多还原

【Abstract】 The goal of structural health monitoring is to achieve real-time monitoring and prediction for the overall behavior of the structure. To achieve this goal, the monitoring system should have a fast large-capacity information collection, transmission and processing capabilities, and data sharing capabilities through network. High-speed data acquisition and transmission, mass data storage and management, monitoring data pre-processing, interpretation, analysis and utilization are the key technical issues to establish the long-term health monitoring system. For the above key technical problems, research on intelligent information acquiring and processing for structural health monitoring are carried out in this paper. The research focuses on the network-based data acquisition and transmission technology, intelligent information processing technology, and dynamic data management and query technology.(1) The guiding ideology of building a structural health monitoring system is refined firstly. Based on the guiding ideology, a general-purpose platform of long-term health monitoring system for large-scale structure is built. The strain test system based on the optic sensor, the electro-sensor test system based on network-based data acquisition and digital temperature sensor testing system are developed independently. The above three test systems, which build up the "small concentration" distributed data acquisition system, are the core of the hardware platform and integrated by the software platform based on the virtual instrument technology and the database system.(2) The usage of wavelet analysis in monitoring information processing for the actual structure is studied. An improved wavelet thresholding algorithm is proposed. The new algorithm, which has a simple expression, performs wave filtering well and improves the quality of noise reduction. The sensitivity of the injury characteristic value expressed by the wavelet packet node energy coefficient and the inherent frequency value after Fourier transform is compared and analyzed. The results show that the injury characteristic value expressed by the wavelet packet node energy coefficient is more sensitive, which proves the usage of wavelet packet in extraction of the injury characteristic value is feasible.(3) The usage of BP neural network in the damage identification for the actual structure monitoring information processing is discussed. A BP network model is established first, and then it is optimized, at last it is used to diagnose the damage of concrete cable-stayed bridge model. The structural damage identification results show that the BP neural network model is easy to identify the damage by the changing of vibration modal frequency.(4) A dynamic data management and query system based on standard relational database is developed independently to implement the unified and standardized management of mass monitoring data. The system provides a good solution to the problems such as mass data storage, data analysis and processing, data query, data backup, data security etc.(5) The above study results have been successfully applied to the Wuhu Yangtze River Bridge and the Zhengzhou Yellow River Bridge, which are two important bridges of symbolic significance. The operation of the two bridge monitoring system shows that the strain test system based on the optic sensor, the electro-sensor test system based on network-based data collection instrument and digital temperature sensor testing system can obtain the corresponding monitoring data well provide intelligent information acquiring and transmitting. The operation of the two bridge monitoring system also shows that wavelet analysis technique selected can complete noise reduction and feature extraction of monitoring data successfully and the BP network model constructed can identify the damage of the structure effectively and the data management and query system developed is able to store monitoring data timely, display query results in real-time and query conveniently. The above technologies achieve analysis, processing, storage and query of monitoring data in real-time and improve the intelligence degree of information processing.The results will further enrich engineering application theories and experimental basis of the structure health monitoring system. And it is of great theoretical significance and practical engineering value. In addition, it will generate significant social and economic benefits.更多还原