北方寒冷地区特大钢箱梁斜拉桥基于健康监测模型修正及损伤识别

发布时间：2018-01-13 15:10

本文关键词：北方寒冷地区特大钢箱梁斜拉桥基于健康监测模型修正及损伤识别　出处：《沈阳大学》2014年硕士论文　论文类型：学位论文

【摘要】：近二十几年，我国大跨径斜拉桥建设数量居于世界前沿。实际成桥后桥梁的受力状况及运营状态下桥梁的结构状态，是目前大型斜拉桥研究的热点。本文针对大型斜拉桥结构健康监测系统中涉及到的几项关键科学与技术问题进行了研究分析，研究内容包括：健康监测系统数据提取、健康监测系统数据滤波、结构的损伤预警、结构的模态参数识别、有限元模型建立及模型修正。主要研究内容如下：第一，总结了桥梁健康监测系统测点布置及数据采集的特点，，结合辽河特大桥健康监测系统，制定出一套针对是用于各类大桥海量数据处理的方法。第二，提出了对于不同类传感器有针对性的建立数据滤波方法，该方法为首先对采集的数据进行加权递推平均，然后对处理后的数据进行小波分析过滤掉数据高频部分得到低频部分数据，最后对低频数据进行单支重构，得到稳态的结构状态数据。该方法能够有效的过滤掉结构因为白噪声及偶然因素产生的噪声数据。第三，利用大型通用有限元软件ANSYS、同济大学的桥梁博士软件分别建立了辽河特大桥模型，并对建立的模型进行了分析，静力学分析中通过对弯矩、剪力、轴力及位移的对比，桥梁博士及ANSYS分析结果相近。动力学分析结果符合桥梁实际的规律，通过有限元模型的建立为辽河特大桥养护提供了指导，也为桥梁的损伤识别、安全评定及预警打下了基础。第四，通过运用健康监测系统对辽河特大桥扁平钢箱梁跨中横截面的温度和纵向应力进行监测，深入研究了辽河特大桥钢箱梁在温度作用下所产生的温度效应、钢箱梁温度、纵向应力日变化趋势、总体变化规律及温度对纵向应力的影响情况。分析结果表明：根据规范JTG D60-2004算出的钢箱梁顶板温度最大值小于实际监测值；钢箱梁连续24h温度变化服从正弦曲线分布，纵向应力每天前6小时变化服从线性分布，后18小时服从高斯曲线分布。根据有两大软件分析结果提取出成桥阶段钢箱梁应力与实际健康监测数据进行累加，得到桥梁真实的应力变化情况。第五，将应变计采集的应变数据通过雨流计法得到一个一个应力幅，将应力幅进行统计，并结合英国疲劳规范对钢箱梁顶板进行了疲劳损伤分析。最后提出了一种基于温度和支座位移相关性的桥梁健康监测系统采集的海量数据处理方法，该方法为运用小波分析法对数据进行了滤波与重构，将重构后的支座位移与环境温度数据建立相关性模型。并构建BP神经网络，建立人工神经网络评估模型。
[Abstract]:In recent 20 years, the number of long-span cable-stayed bridges in China is in the forefront of the world. It is a hot topic in the research of large cable-stayed bridges at present. In this paper, several key scientific and technical problems involved in the structural health monitoring system of large cable-stayed bridges are studied and analyzed. The research contents include: data extraction of health monitoring system, data filtering of health monitoring system, structure damage warning, modal parameter identification of structure, finite element model establishment and model modification. Firstly, this paper summarizes the characteristics of point layout and data collection of bridge health monitoring system, and develops a set of methods for mass data processing of all kinds of bridges combined with Liaohe super bridge health monitoring system. Secondly, a new method of data filtering for different kinds of sensors is proposed. The method is first weighted recursive average of the collected data. Then wavelet analysis is used to filter out the high-frequency part of the data to get the low-frequency part of the data. Finally, the low-frequency data is reconstructed by a single branch. The steady-state structural state data can be obtained. This method can effectively filter out the noise data of the structure due to white noise and accidental factors. Third, using the large-scale universal finite element software ANSYS, Tongji University bridge doctor software respectively to establish the model of the Liaohe super large bridge, and the established model is analyzed, the static analysis through the bending moment. Compared with the shear force, axial force and displacement, the results of the bridge doctor and ANSYS analysis are similar. The dynamic analysis results accord with the actual rules of the bridge. The finite element model provides guidance for the maintenance of the Liaohe super bridge. It also lays a foundation for bridge damage identification, safety assessment and early warning. In 4th, the temperature and longitudinal stress of the middle cross section of flat steel box girder of Liaohe Bridge were monitored by using health monitoring system. The temperature effect of the steel box girder of Liaohe super bridge under the action of temperature, the diurnal variation trend of temperature and longitudinal stress of the steel box girder are studied in depth. The analysis results show that the maximum temperature of steel box girder roof is less than the actual monitoring value according to the code JTG D60-2004. The temperature change of steel box girder for 24 hours was distributed according to the sine curve, and the longitudinal stress changed from linear to linear in the first 6 hours of each day. According to the analysis results of two softwares, the stress of the steel box girder in the bridge stage and the actual health monitoring data were extracted and accumulated, and the true stress changes of the bridge were obtained. In 5th, the strain data collected by the strain gauge were used to obtain a stress amplitude through the rain flow meter method, and the stress amplitude was counted. The fatigue damage analysis of the steel box girder roof was carried out in combination with the British fatigue code. Finally, a massive data processing method of bridge health monitoring system based on the correlation between temperature and support displacement is proposed. The method uses wavelet analysis to filter and reconstruct the data. The correlation model between the reconstructed support displacement and the ambient temperature data is established, and the BP neural network is constructed, and the evaluation model of the artificial neural network is established.
【学位授予单位】：沈阳大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U446;U448.27

【参考文献】