考虑复杂环境影响的大跨度斜拉桥异常状态统计识别

In the process of long-term health monitoring, structural condition features were not only affected by structural conditions, but also influenced by some random factors, such as complex environmental factors, measurement noise and analysis errors. The traditional structural abnormal condition assessment methods, which were based on deterministic theories, cannot apply in this condition. So a statistical identification method of abnormal conditions in long-span cable-stayed bridges considering complex environmental effects was proposed. At first, these structural condition features, were extracted from the monitoring data of long-span cable-stayed bridges. The major environmental factors, which significantly affected structural condition features, were identified and the influcing mechanism was analyzed. The analysis model of complex environmental factors effects was established and environmental effects were decoupled. After that, structural abnormal condition can be detected based on the non-parameters sequential probability ratio testing method. Meanwhile according to the monitoring data， such as extreme environmental factors and boundary condition, the type of structural abnormal condition can be correctly determined. Structural abnormal position can be identified based on the rejection probability estimation methods. Combined with the finite element model of structural benchmark condition, structural abnormal extent was estimated by Bayesian statistical methods. In this study, in order to consider the effects of various uncertainty factors，a statistical analysis method of abnormal conditions in cable-stayed bridges was proposed and the abnormal conditions of cable-stayed bridges can be correctly assessed.

基于长期健康监测数据提取的结构状态特征不仅取决于结构自身状态，还会受结构所处的复杂环境因素、测量噪声、分析误差等不确定性因素的影响。传统的基于确定论的异常状态识别方法难以取得满意的应用效果，因此本研究注重考虑复杂环境影响的大跨度斜拉桥异常状态统计识别方法。主要研究内容有：基于斜拉桥长期监测数据提取状态特征；结构状态特征的环境因素识别、影响机理分析及建模，解耦环境因素的影响效应；基于非参数序贯概率比检验对斜拉桥异常状态统计判别，并判别异常状态类型；基于拒绝概率估计方法对结构异常位置统计识别；基于贝叶斯统计和基准状态有限元模型的结构异常程度统计估计。本研究考虑各种不确定性因素的影响，提出有效的斜拉桥异常状态统计分析方法，以获得对斜拉桥异常状态准确的评估.

基于长期健康监测数据提取的结构状态特征不仅取决于结构自身状态，还会受结构所处的复杂环境因素、测量噪声、分析误差等不确定性因素的影响。传统的基于确定论的异常状态识别方法难以取得满意的应用效果，因此本研究注重考虑复杂环境影响的大跨度斜拉桥异常状态统计识别方法。主要研究内容有：基于斜拉桥长期监测数据提取状态特征；结构状态特征的环境因素识别、影响机理分析及建模，解耦环境因素的影响效应；基于非参数序贯概率比检验对斜拉桥异常状态统计判别，并判别异常状态类型；基于拒绝概率估计方法对结构异常位置统计识别。本研究考虑各种不确定性因素的影响，提出有效的斜拉桥异常状态统计分析方法，以获得对斜拉桥异常状态准确的评估.