全文下载:
201906001.pdf
文章编号: 1672-6987(2019)06-0001-18; DOI: 10.16351/j.1672-6987.2019.06.001
丁锋1,2, 徐玲1, 刘喜梅2
(1.江南大学 物联网工程学院, 江苏 无锡 214122; 2.青岛科技大学 自动化与电子工程学院, 山东 青岛 266061)
摘要: 工程中,频率特性又称频率响应。针对不同极点惯性环节并联而成的系统,利用正弦激励信号作为输入,通过测量系统的频率特性观测数据,基于二次优化和非线性优化技术,推导了估计传递函数参数的最小均方算法、随机梯度算法、多新息随机梯度算法、递推梯度算法、多新息递推梯度算法、牛顿递推算法,以及结合实频特性和虚频特性观测数据的联合递推辨识算法和耦合递推辨识算法。文中的方法可以推广用于其他传递函数描述的动态系统参数辨识,如具有共轭极点、重极点传递函数参数的辨识以及任意非线性函数的参数估计。
关键词: 传递函数; 参数估计; 递推辨识; 迭代辨识; 频率特性; 梯度搜索; 牛顿搜索
中图分类号: TP 273文献标志码: A
引用格式: 丁锋, 徐玲, 刘喜梅. 传递函数辨识(11):频率响应递推参数估计(并联情形)\[J\]. 青岛科技大学学报(自然科学版), 2019, 40(6): 1-18.
DING Feng, XU Ling, LIU Ximei. Transfer function identification. Part K: Recursive parameter estimation based on the frequency characteristics\[J\]. Journal of Qingdao University of Science and Technology(Natural Science Edition), 2019, 40(6): 1-18.
Transfer Function Identification. Part K: Recursive Parameter Estimation
Based on the Frequency CharacteristicsDING Feng1,2, XU Ling1, LIU Ximei2
(1.School of Internet of Things Engineering, Jiangnan University, Wuxi 214122, China;
2.College of Automation and Electronic Engineering, Qingdao University of Science and Technology, Qingdao 266061, China)
Abstract: Frequency characteristics is also called the frequency response in engineering. For different pole of inertial link parallel systems, using the sine excitation signal as input and measuring the system frequency characteristics of the observation data, this paper studies the least mean square algorithms, the stochastic gradient algorithms, the multi-innovation stochastic gradient algorithms, the recursive gradient algorithms, the multi-innovation recursive gradient algorithms and the Newton recursive algorithms to estimate the parameters of the transfer functions based on the quadratic optimization and the nonlinear optimization techniques, as well as the joint recursive identification methods and the coupled recursive identification methods.Although the methods in this paper are proposed for transfer functions with distinct poles, they can be extended to the parameter estimation of transfer functions with conjugate poles and duplicated poles and nonlinear systems.
Key words: transfer function; parameter estimation; recursive identification;iterative identification; frequency response; gradient search; Newton search
收稿日期: 2019-10-28
基金项目: 国家自然科学基金项目(61873111).
作者简介: 丁锋(1963—),男,博士,“泰山学者”特聘教授,博士生导师.