Parameters Identification Of Building Structures Using Time Series Analysis Methods

Time domain identification method of structures'modal parameters means a method which can be used to identify structures'modal parameters in time domain. The research and application of time domain identification method are developed later than those of frequency domain method. And time domain identification method is a new technique developed with application and development of computers in last decades. Time domain method can over the defects of frequency domain method. Especially for huge and complex structures, such as dams, bridges, high buildings and aircrafts, ships etc., which are usually suffered from wind, wave, and pulse of the earth. The loads are hard to measure when the structures are working, but the response signals of them are easy to measure, so it's definitely important for us to identify parameters just using response signals in time domain. At present time domain identification methods of modal parameters mainly include least square method in time domain, prony method, and ARMA model time series analysis method. Besides, time domain methods can also be used to do online identification for working equipments, such as generators, large chemical units etc.. These parameters identified in working conditions can actually reflect the dynamic characteristics of structures. Because time domain methods of parameter identification techniques just need response signals in time domain, they can greatly save excitation equipments, test time and costs, which are all advantages that frequency domain methods don't have.The contents and methods contained in this thesis are mainly as follows:(1) First, basic theory of parameter identification, the development situation of parameter identification methods and basic parameter identification methods in both time and frequency domains are introduced. Then the research and application of time series method of parameter identification at home and abroad currently are simply elaborated.(2) The principles and basic theory of time series models are elaborated. The concept, expressions, important time domain characteristics of time series models are significantly introduced. (3) The modeling process of time series parameter identification method is elaborated which includes parameter estimation, modeling steps, order determination and true and false modes distinguishing methods of time series models.(4) The process and related test results of shaking table test of high-rise structure are introduced. Then the important functions and advantages of MATLAB software used for modeling in this thesis are simply introduced. At last, the process of parameter identification of test model structure based on ARMA model using MATLAB software is elaborated. Then identification results and test results are compared, and some conclusions are obtained. And some problems which need to be researched and improved in future are proposed.