| 一类时变参数的递推辨识方法 |
| |
| 引用本文： | 韩兵,陈新海. 一类时变参数的递推辨识方法[J]. 西北工业大学学报, 1988, 0(4) |
| |
| 作者姓名： | 韩兵 陈新海 |
| |
| 作者单位： | 西北工业大学，西北工业大学 硕士 郑州工学院讲师，教授 |
| |
| 摘 要： | 本文对线性时变系统的参数辨识方法进行了分析和研究,提出了一类新的时变参数递推辨识方法。算法假定时变参数由自回归模型表示,并保证辨识算法结构传递矩阵正实或严格正实。本文分别给出了白噪声和有色噪声情况的两种递推辨识算法,证明了它们的一致收敛性。结构分析表明,最小二乘估计和朗道自适应串联辨识算法是这类辨识方法的特例。在给定的结构条件下,辨识算法对周期和单调变化的时变参数有一定的跟踪能力。
|
| 关 键 词： | 递推辨识算法 快速时变参数 结构传递矩阵 |
On New Recursive Identification Algorithms for Rapidly Time-Varying Parameters |
| |
| Han Bing Chen Xinhai. On New Recursive Identification Algorithms for Rapidly Time-Varying Parameters[J]. Journal of Northwestern Polytechnical University, 1988, 0(4) |
| |
| Authors: | Han Bing Chen Xinhai |
| |
| Affiliation: | Department of Astronautical Engineering Northwestern Polytechnical University |
| |
| Abstract: | This paper presents recursive identification algorithms for parametersthat vary rapidly.The algorithm assumes that time-varying parameters are fromARMA model and guarantees that the structure transfer matrix is positive realor strictly positive real.In this paper,two kinds of recursive identificationalgorithms which are proved to rave consistent convergency are given in thecase of white noise and color noise.The strueture analysis indicates that theleast square estimation and Landau adaptive estimation are special forms ofthe authors' identification algorithms. |
| |
| Keywords: | recursive identification algorithm rapidly time-varying parameter structure transfer matrix. |
| 本文献已被 CNKI 等数据库收录！ |
