结构参数识别中的数字滤波

时间序列分析法已成功地用于结构的参数识别。本文所应用的数字滤波技术可使响应采样序列中高阶的信息消除，以便降低ＡＲＭＡＶ模型的阶数而克服拟合其参数时出现的困难。 文中指出了如何选择合适的滤波器，由模拟的向离散的滤波器的转换方法，以及消除其相位失真的逆置方法． 最后给出了应用滤波技术的时间序列法参数识别的仿真实例和实验实例。

Digital Filtering in Parameter Identification of Structures

Time series analysis techniques have been successfully used in the parameter identification of structures. In this paper, a digital filtering technique is developed to reduce the adequate order of the Autoregressive Moving Average Vector(ARMAV) model so as to overcome the difficulties associated with the higher order modeling. Problems in digital filtering with the ARMAV model are analyzed, including the selection of an appropriate filter, the methods of transformation from an analog filter to discrete filter, and the time reversal method for eliminating the phase distortion. Finally, a simulation example and experimental results are illustrated using the digital filtering technique presented herein.