基于特制信号频域模型集辨识.pdfVIP

一种基于特制激励信号的线性对象频域模型集辨识方法 李飞 ，熊智华 清华大学自动化系, 北京100084 E-mail: f-li09@mails.tsinghua.edu.cn 摘 要: 本文针对一类存在不确定性的对象，采用特制激励信号来辨识对象的频域响应，并建立对象的频域模型集. 在 采用特制激励信号获取对象的时域辨识数据之后，应用chirp-z变换来计算对象输入输出信号的离散频谱，进而得到对 象的离散频率响应. 在对象的先验知识较少的情况下，利用Laguerre正交基函数拟合得到对象的频域标称模型，并利 用非线性规划算法计算对象的频域模型集，给出了描述对象真实模型的可能范围，从而为系统的控制器设计打下了基 础. 最后，在一个二维机翼机构模型上进行了仿真分析，验证了方法的有效性. 关键词: 特制激励信号,离散频率响应, 正交基函数拟合, 模型集 Identification of Frequency Domain Model Set for Linear System Based on Tailored Excitation Signals Li Fei, Xiong Zhihua Department of Automation, Tsinghua University, Beijing, China, 100084 E-mail: f-li09@mails.tsinghua.edu.cn Abstract: In this paper, tailored signals are used as excitation signals to obtain frequency domain model set of a class of plants with uncertainty. After collecting the data from identification test, chirp-z transform is adopted to calculate the discrete frequency spectra of the input and the output signals. Then the discrete frequency response is obtained. With little priori knowledge of the plant and the type of uncertainty, Laguerre orthogonal basis functions are chosen to construct the nominal frequency domain model. Furthermore, the model set of the plant can be determined by using nonlinear programming algorithm, and it can depict model errors and some unknown uncertainties. The result is helpful for system analysis and controller design etc. The proposed method is illustrated on a case of a two-dimensional wing section model. Key Words: Tailored Signals, Chirp-z Transform, Discrete Frequency Response, Model Set 定的情况，或者受到外部环境变化的影响而引起系 1 引言 统变化，例如高超声速飞行器在飞行测试过程中飞 在传统的线性对象频域辨识