【作者】 潘洪良；

【导师】 赵树山；

【作者基本信息】 哈尔滨工业大学 ， 固体力学， 2006， 硕士

【摘要】 模态分析是研究结构动力特性的一种近代方法,是系统识别方法在工程振动领域中的应用。模态是结构的固有振动特性,每一阶模态都具有特定的固有频率、阻尼比、振型、模态刚度和模态质量。这些模态可以由计算或试验分析取得,这样一个理论和实验的分析过程就称为模态分析。本文依照实际工程的背景,对一个钢框架结构模型进行模态参数识别。首先,对钢框架结构建立有限元模型,并分析其动力特性,得到钢框架前10阶的固有频率和振型。然后,对该钢框架模型进行模态试验,试验通过力锤分别对钢框架进行随机敲击和单次敲击,通过传感器和数据采集系统得到结构的试验数据,再对试验数据进行后处理。本文采用的参数识别方法是不测力法,即基于环境激励的模态参数识别方法。不测力法主要通过自己编写的MATLAB程序(响应谱分析法,随机子空间法)对试验数据进行处理。响应谱分析法是传统的识别方法,通过自谱、互谱图等来确定固有频率、阻尼比和振型。随机子空间(SSI)是一种比较先进的模态参数识别方法,但是传统的子空间识别方法是建立在有输入的基础之上。本文对该方法进行了改进,适用于无输入条件下的识别。本文结论部分通过实验和理论的分析探讨了传统的参数识别方法和随机子空间参数识方法的一些问题,分析了不测力法在具体的结构的参数识别上的问题和误差大小。更多还原

【Abstract】 Model analysis is the procedure of determining a structure’s dynamic characteristics recently , is a application of system identification method in vibration domain . Mode is structure’s inherence characteristics , each mode has specifically resonant frequencies , damping values , mode shapes, mode stiffness and mode mass. To obtain these modes , we can calculating and experimental analysis , such the course of theory and experimental analysis called Model analysis .Based on the background of practical engineering , modal parameters identification of a steel frame structure in this paper . Fist , the dynamic character of the structure are analyzed by the Finite Element Model (FEM) obtain the low ten model resonant frequencies and model shapes . Then model experimental on the steel frame structure , experiment uses random knock and single knock separately on the steel frame structure through a force hammer . We can gained the data of experiment using transducers and the data acquisition system and post treatment of the data .Modal parameters identification include ODS ,ODS mainly use MATLAB procedure(Response spectrum method , SSI) for dealing with the experiment data .Response spectrum method is using APS , CPS to identify modal parameters including resonant frequencies , damping and model shapes . Stochastic subspace identification(SSI) is a advanced modal parameters identification method , but traditional subspace parameters identification method based on having input data .An improved method of stochastic subspace identification is given in this paper . The problems existing for structural model parameter identification , ODS and SSI in the conclusion of this paper .更多还原