The safety monitoring and condition assessment system (SMCAS) constructed for the Nanjing Yangtze River Bridge (NYRB) and its functions are introduced in this paper. The SMCAS, which allows the bridge to be monitored and evaluated under normal operating conditions, is composed of sensor system, signal sampling and processing system, and structural monitoring and condition assessment system. The objective is to evaluate objectively and accurately the integrity and safety of the bridge and operating conditions in real time for the NYRB. According to requirements of safety monitoring for the NYRB, representational and dominative monitoring items and layout of measuring points are determined by combining theoretical analysis with measured data analysis. Key techniques to build and realize the SMCAS for the NYRB, such as automation of safety monitoring, reliability, special software designing, massive monitoring data analysis and processing, data communication etc., are also discussed. The SMCAS for the NYRB can be used for reference for a similar monitoring system.

Large cable vibrations of a cable-stayed bridge have been observed frequently around the world in the past decade. They are very harmful to the bridge. For this reason, a new and smart magneto-rheological (MR) damper first used to mitigate vibrations of cables for the Dongting Lake Cable-stayed Bridge (DLB) is investigated in this paper. MR damper has some obvious properties such as insensitivity to temperature variation and the low power required for operation. Its damping coefficient can be easily adjusted by changing the voltage level applied on the MR damper. In order to demonstrate the effectiveness of cable vibration mitigation, experimental devices including supporting system and excitation system are designed, and tests for cables with MR dampers under different voltage levels and without MR dampers on the DLB are completed. For determining a modal damping ratio which is one of the important parameters of stay cables, a new approach to identification of equivalent modal damping ratios of cables incorporated with the MR dampers is also presented in this paper. Results of field tests show that the MR damper is very effective to vibration mitigation of cables. The approach proposed in this paper is practical and feasible.

本文提出一种基于信号能量分析计算结构阻尼比的新方法。推导出自由振动响应信号能量与阻尼比关系的理论表达式后，求出阻尼比。与传统的对数衰减率法相比，此法精度、抗噪声干扰能力强、稳定性好。振动系统信号的能量之所以会衰减，是因为阻尼的存在。本文提出的这种识别阻尼比的方法物理概念清晰，易于理解。仿真计算和洞庭桥实测试验结果表明所提出的方法有效、可行量分析，参数识别，阻尼比。

回顾桥梁健康监测的发展历程。介绍桥梁健康监测系统的组成、监测内容功能及特点。对信号分析与处理的传统谱分析方法与小波变换、希尔波特黄变换（Hilbert. Huang）进行比较。介绍最新的信号时频分析方法希尔波特黄变换的具体算法，及其在结构参数识别、损伤识别、消除实测南京长江大桥应变信号中的对讲机干扰等方面的具体应用。对动力指纹分析法、模型修正与系统识别法、神经网络法、遗传算法5种损伤检测方法的基本原理、特点及发展动态进行概述和总结，指出神经网络技术结合遗传算法是结构损伤检测的发展方向之一。

高阶谱较之功率谱有三个显著特点：（1）可用来检查过程高斯性；（2）可估计非高斯过程的相位；（3）可检醒系统的非线性。高阶谱的特点使得它可用于故障诊断。本文分析了频响函数三阶谱对刚度变化的敏感，并效它用于故障诊断。数字仿真实物试验结果表明，本方提出的方法是可行和有效的。

本文提出一种利用自由振动响应通过年顿。拉夫逊迭代同时识别出桥梁全部8个颤振气动导数的方法。此法具有对迭代初值要求不高、识别结果稳定的特点，并具有一定的抗噪声干扰能力。数字仿真与实物试验结果表明本文方法有效、可行。

斜拉桥拉索模态参数（固有频率、阻尼比）在索力测试、拉索减振、实时控制等方面起着重要作用。文中利用随机减量技术从岳阳洞庭桥环境激励下拾取的拉索的加速度响应中分离出自由衰减振动加速度响应信号，将获取的加速度响应信号表达为一理论形式，综合运用参数识别、最优估计理论，识别出拉索的模态参数，与理论值吻合良好。本文方法简单，试验容易实施，具有工程实用价值。数字仿真与工程测试结果表明了方法的有效性和可行性。

提出一种新的由结构自由振动响应识别结构阻尼比的方法。对于单自由度振动系统，推导出响应历程与时轴所围各面积之间的确定性关系后，利用各面积之间的关系来确定自由衰减振动的阻尼比系数。与传统的对数衰减率法比较，它具有抗噪声干扰能力强、精度高、稳定性好及简便实用等特点。对于多自由度振动系统，先将结构上某测点的自由振动响应表达为一理论解析式x（t），以eat相乘x（t），预给a初值范围，通过牛顿二分法（或黄金分割法）搜索a值，直至eatx（t）做等幅振荡，则该阶模态的各模态参数得以确定。从总响应中扣除该阶模态对总响应的贡献后，重复这一过程，则可识别出响应信号中各阶模态的模态阻尼比。仿真计算与岳阳洞庭湖斜拉桥拉索实测试验结果表明了本文方法的有效性和实用性。

运用系统识别理论识别桥梁颤振气动导数的过程，可分为复模态识别与模态坐标转换为物理坐标两个阶段。文中着重研究坐标转换算法，该算法建立一个带约束的二次泛函，利用惩罚函数法以一定的搜索方向使二次泛函取得极小值，并同时求得8个气动导数。数字仿真与实物试验结果表明文中方法有效、可行。

主要研究基于神经网络技术的南京长江大桥桥域船撞力识别。在建立了南京桥4#桥域的空间有限元分析模型后，对4#桥域进行了动力特性分析，基fMATllAB平自建立了“船撞力响应”神经网镕，自沃辛试验船撞力曲线计算得到了船对桥域的撞击响应并作为神经网络训练的样本。仿真计算结果表明，自桥域域顶的响应可有效识别船舶撞击力的大小、方向和撞击位置，该神经网镕可用于有、无噪声干扰的船撞力识别，并且具有良好的窖错性和鲁棒性。