Variance and bias confidence criteria for ERA modal parameter identification. [Eigensystem Realization Algorithm]
Variance and bias confidence criteria for ERA modal parameter identification. [Eigensystem Realization Algorithm]
复制标题
ERA 模态参数识别的方差和偏差置信标准。
DOI：
--
复制
发表时间：
1988
期刊：
影响因子：
--
通讯作者：
J. Juang
中科院分区：
文献类型：
--
作者：
R. Longman;M. Bergmann;J. Juang
对于ERA系统的识别算法，摄动方法被用来开发表达式的方差和偏差的模态参数。基于测量噪声的统计数据，方差结果通过指示真实参数位于其识别值的任何选定区间内的可能性来充当置信度标准。这取代了使用昂贵和耗时的蒙特卡罗计算机运行来获得类似的信息。偏差估计有助于指导ERA用户选择使用哪些数据点以及使用多少数据以获得最佳结果，从而在偏差和分散之间进行权衡。此外，当偏差的不确定性足够小时，偏差信息可用于校正ERA结果。此外，奇异值的方差和偏差的表达式作为工具，以帮助ERA用户决定适当的模态顺序。
For the ERA system identification algorithm, perturbation methods are used to develop expressions for variance and bias of the identified modal parameters. Based on the statistics of the measurement noise, the variance results serve as confidence criteria by indicating how likely the true parameters are to lie within any chosen interval about their identified values. This replaces the use of expensive and time-consuming Monte Carlo computer runs to obtain similar information. The bias estimates help guide the ERA user in his choice of which data points to use and how much data to use in order to obtain the best results, performing the trade-off between the bias and scatter. Also, when the uncertainty in the bias is sufficiently small, the bias information can be used to correct the ERA results. In addition, expressions for the variance and bias of the singular values serve as tools to help the ERA user decide the proper modal order.