基于EEMD的桥梁挠度分离研究

基于EEMD的桥梁挠度分离研究(任务书,开题报告,中期检查,外文翻译,论文13000字)
摘要
桥梁挠度是桥梁安全评估的关键因素，反应了桥梁的整体特性，影响着桥梁的稳定性能、美观程度及安全性能。影响桥梁挠度的因素有很多，除了桥梁自重、混凝土收缩徐变等恒载以外，还包括车辆冲击荷载以及外界环境（如风，地震，环境温度）产生的其他荷载。准确分离复杂因素的影响，求出各因素单项贡献值的作用，是评估桥梁安全性能的重要前提。
首先，根据不同影响因素具有时间多尺度的特性，可以将不同成分引起的挠度进行分离，从而得到单项因素的量值贡献。其中，活载挠度是动态的，温度挠度的变化频率较小，混凝土收缩徐变引起的挠度变化接近于0。了解了不同因素的时间特性差异之后，便可从信号处理的角度对不同挠度成分进行分离，得到不同时间尺度的作用效应。
其次，对于车辆荷载等高频信号，可采用滤波算法进行分离。对于温度挠度、混凝土收缩徐变挠度等低频信号，传统的信号处理方式难以分离，本文提出了集合经验模态分解与独立成分分析相结合的算法，更精确地分离出不同成分的挠度响应。其中，集合经验模态分解(EEMD)是传统经验模态分解(EMD)的改进算法，分离效果更好，精度更高。独立成分分析（ICA）则解决了EEMD端点效应太过明显的缺点。
最后，本文通过对总挠度信号的仿真模拟，验证了该方法的可行性。对桥梁施加单项荷载，获得相应的模拟信号，将单项模拟信号进行混合叠加，再经过集合经验模态分解和独立成分分析对单通道混合信号进行分离，分离结果与源信号进行比对发现，两者的相关系数达到0.9以上，说明了分离方法的可行性和准确性。
关键词：集合经验模态分解；独立成分分析；桥梁挠度；改进的PCA

ABSTRACT
Bridge deflection is a key factor in bridge safety assessment, which reflects the overall characteristics of the bridge and affects the stability, aesthetics and safety of the bridge. There are many factors affecting the deflection of the bridge, in addition to the dead weight of the bridge, concrete shrinkage and creep and other constant load, including vehicle impact load and the external environment (such as wind, earthquake, environmental temperature) generated by other load. It is an important prerequisite to accurately separate the influence of complex factors and calculate the contribution value of each factor.
First of all, according to the time-scale characteristics of different influencing factors, the deflection caused by different components can be separated, so as to obtain the magnitude contribution of single factor. Among them, the live load deflection is dynamic, the change frequency of temperature deflection is small, and the deflection change caused by concrete shrinkage and creep is close to 0. After understanding the difference of time characteristics of different factors, the components of different deflection can be separated from the perspective of signal processing to obtain the effect of different time scales.
The second, high frequency signals such as vehicle load can be separated by filtering algorithm. Traditional signal processing methods are difficult to separate low-frequency signals such as temperature deflection and concrete shrinkage and creep. EEMD is an improved traditional EMD algorithm with better separation effect and higher precision. Independent component analysis (ICA) solves the disadvantage that EEMD endpoint effect is too obvious.
The last, the feasibility of this method is verified by the simulation of the total deflection signal. Bridge applying single load to the background, to get the corresponding analog signal, mix single analog signal superposition, through the collection of empirical mode decomposition and independent component analysis was carried out on the single channel of mixed signal separation, compares with the source signal separation results found that the correlation coefficient reaches 0.9 above, the separation method is feasible and accurate.
Keywords: collective empirical modal decomposition; Independent component analysis; Bridge deflection; The improved PCA
 

目录
第一章  绪论    1
1.1 研究背景及意义    1
1.1.1大跨斜拉桥特点与发展    1
1.1.2大跨斜拉桥挠度问题研究    1
1.2 桥梁挠度信号的分离研究    2
1.2.1 桥梁挠度信号分离研究现状    2
1.2.2 挠度组分分析    3
1.2.3 温度挠度分离方法研究    3
1.基于统计回归的方法    3
2.基于EMD和ICA的方法    3
1.3 本文主要研究内容    3
第二章  基于信号分离的理论分析    5
2.1 EMD基本理论及其改进算法    5
2.1.1 EMD算法的基本理论    5
1.EMD的应用范围    5
2.EMD算法的主要问题    5
2.1.2 EEMD算法及其步骤    5
1.EEMD基本理论    5
2.EEMD分解步骤    6
2.2 改进的主成分分析PCA    6
2.2.1 改进的PCA基本原理    6
2.2.2 改进的PCA算法流程    6
1.改进的PCA算法流程    7
2.二次降维    7
2.3 独立分量分析ICA    7
2.3.1 独立分量分析的定义    7
2.3.2 FastICA算法    8
2.4 本章小结    9
第三章  桥梁挠度分析计算    11
3.1 Midas/civil模型    11
3.1.1 Midas/civil软件介绍    11
3.1.2 斜拉桥模型模态分析    11
3.1.3 温度荷载挠度及建模    13
1.温度荷载挠度的成因    13
2.温度挠度建模    13
3.2 仿真总挠度的分离    15
3.2.1 基于ICA算法的低频挠度信号分离    15
3.2.2 分离结果评价    18
3.3 本章小结    19
第四章  桥梁实测挠度信号分离    20
4.1 活载挠度测取    20
4.1.1 工程概况    20
4.1.2 项目概述及监测方法    20
4.2 实测数据处理    21
4.2.1 二月份实测数据处理    21
1.桥梁主跨跨中挠度数据处理    21
2.桥梁边跨跨中挠度数据处理    26
4.2.2 七月份、十月份实测数据处理    27
4.2.3 一年内实测数据处理    28
4.3 数据结果分析及评价    29
4.4 本章小结    30
第五章  结论与展望    31
5.1 结论    31
5.2 展望    31
参考文献    32
致谢    34