On the hysteretic Bouc-Wen model - Part II: Robust parametric identification
On the hysteretic Bouc-Wen model - Part II: Robust parametric identification
复制标题
DOI：
10.1007/s11071-005-0070-x
复制
发表时间：
2005-10-01
影响因子：
5.6
通讯作者：
Rodellar, J
中科院分区：
文献类型：
--
作者：
Ikhouane, F;Rodellar, J
本文讨论滞回Bouc-Wen模型参数的识别问题。在现有的文献中，致力于这个问题的方法主要依赖于数值模拟，并没有，在很大程度上，一个严格的数学证明。我们的方法包括在激励滞后系统与周期性输入，并获得所需的参数，由此产生的极限环。我们提出的识别方法具有严格的数学基础，因为它基于极限环的解析描述，并且，与现有的Bouc-Wen模型识别方法不同，给出了（未知的）精确模型参数及其相应估计之间的保证相对误差。我们还证明了这种方法是强大的常数和T-周期性干扰通常存在于任何实验室实验。数值模拟的例子说明了我们的识别方法的使用。
This paper deals with the problem of identifying the parameters of the hysteretic Bouc-Wen model. In the existing literature, the methods devoted to this problem rely mainly on numerical simulations and do not have, to a very large extent, a rigorous mathematical justification. Our method consists in exciting the hysteretic system with a periodic input and obtain the desired parameters from the resulting limit cycle. The identification method that we propose has a rigorous mathematical basis as it based on the analytic description of the limit cycle, and, unlike existing identification methods for the Bouc-Wen model, gives guaranteed relative errors between the (unknown) exact model parameters and their corresponding estimates. We also prove that this method is robust with respect to constant and T-periodic disturbances commonly present in any laboratory experiment. A numerical simulation example illustrates the use of our identification method.