基于自回归模型和主成分分析的结构损伤识别方法分析-analysis of structural damage identification method based on autoregressive model and principal component analysis.docxVIP

AbstractDuring the life period of civil engineering structures, invasions by natural environmental causes and man-made factors are unavoidable to these structures, which lead to degradation and damage in the structures. Failure to detect and repair these functional deficits may lead to catastrophe events. Therefore, the structural damage identification is developing into a research area which are of great significance both in theoretical and application aspects. In this paper, time series autoregressive model and principal component analysis are employed for structural damage detection, and concentration is made on the removal of environmental effects on the structural damage identification results. T he main contents are as follows:Studies on topics of environmental effects on the structural dynamic response and the damage identification theory based on structural vibration are reviewed. The basic theory of the autoregressive model is introduced. Model order determination criteria and parameter estimation methods are compared. The recursive least squares estimation method combine d with the BIC criterion can ensure the accuracy and fast calculation for parameter estimation, and it is suitable for online monitoring.Two damage identification methods based on the autoregressive model are introduced, namely, the Euclidean distance discrimination method and the Itakura distance discrimination method based on the model residuals. Through experimental studies on a truss model, the effectiveness of the two methods is verified, but they are susceptible to environmental factors.The damage detection method based autoregressive model is improved by introducing the principal component analysis (PCA). Autoregressive model is constructed using acceleration of structures. Principal component analysis is applied to extract and eliminate the environmental component s from autoregressive model coefficients. Reconstructed coefficient vectors of the autoregressive model are used to