密集模态挠性结构模型多变量频域辨识和控制

Multivariable Frequency Domain Maximum Likelihood Identification and Control of Flexible Structures with Close Modes

密集模态挠性结构的模态不仅阻尼小而且耦合程度高,给模型辨识带来了很大困难。频域辨识是获取空间挠性结构模型的一个有效方法,但是目前频域辨识要么模型结构定义有缺陷,要么是集中在单变量情形,不适合作大型空间挠性结构多变量模型的辨识。本文提出了多变量极大似然频域辨识,给出了其模型结构定义、算法推导及实现,并将其应用于H型密集模态挠性板的模型辨识,根据辨识结果,设计主动控制律,实现了对密集模态挠性板振动的有效抑制,表明了辨识算法及主动控制的可行性。

It is difficult to id modal damping and high mo entify the model of flexible structures with close modes for its lightly dal coupling. Frequency domain identification is an effective method to obtain the model parameters of flexible structures, but it is mainly focused on Single Input Single Output system and can not be used to identify the model of flexible large space structures, which usually requiring multivariable description. The multivariable frequency domain maximum likelihood identification is proposed in this paper for the first time. The algorithm derivations and realization are given in details. An H-like flexible plate with close modes is given as the model structure. Active controller was designed based on the model identified by the proposed method. Good control effect validates the effectiveness of the active controller and the feasibility of the identification method.