| 方程误差系统的多新息辨识方法 |
| |
| 引用本文： | 丁锋.方程误差系统的多新息辨识方法[J].南京气象学院学报,2015,7(5):385-407. |
| |
| 作者姓名： | 丁锋 |
| |
| 作者单位： | 江南大学 物联网工程学院, 无锡, 21412;江南大学 控制科学与工程研究中心, 无锡, 214122;江南大学 教育部轻工过程先进控制重点实验室, 无锡, 214122 |
| |
| 基金项目： | 国家自然科学基金(61273194);江苏省自然科学基金(BK2012549);高等学校学科创新引智"111计划"(B12018) |
| |
| 摘 要： | 多新息方法可以用于线性系统和非线性系统的自适应滤波、参数估计、自校正控制、自适应故障检测与诊断等.线性系统包括两种基本类型:方程误差类系统和输出误差类系统.本文将多新息辨识应用到方程误差滑动平均(EEMA)系统(即CARMA系统),研究多新息增广随机梯度算法和多新息增广最小二乘算法,应用到方程误差自回归滑动平均(EEARMA)系统(即CARARMA系统),提出基于分解的多新息广义增广随机梯度算法和基于分解的多新息广义增广最小二乘算法,以及基于滤波的多新息广义增广随机梯度算法和基于滤波的多新息广义增广最小二乘算法.
|
| 关 键 词： | 参数估计 递推辨识 梯度搜索 最小二乘 滤波 分解 辅助模型辨识思想 多新息辨识理论 递阶辨识原理 方程误差系统 线性系统 |
| 收稿时间： | 2015/9/20 0:00:00 |
Multi-innovation identification methods for equation-error systems |
| |
| DING Feng.Multi-innovation identification methods for equation-error systems[J].Journal of Nanjing Institute of Meteorology,2015,7(5):385-407. |
| |
| Authors: | DING Feng |
| |
| Institution: | School of Internet of Things Engineering, Jiangnan University, Wuxi 21412;Control Science and Engineering Research Center, Jiangnan University, Wuxi 214122;Key Laboratory of Advanced Process Control for Light Industry (Ministry of Education), Jiangnan University, Wuxi 214122 |
| |
| Abstract: | The multi-innovation methods can be applied to adaptive filtering,parameter estimation,self-tuning control,adaptive fault detection and diagnosis,etc.for linear systems and nonlinear systems.Linear systems include two basic types:equation-error type systems and output-error type systems.This paper studies the multi-innovation (MI) extended stochastic gradient algorithm and the MI extended least squares identification algorithm for equation-error moving average (EEMA) systems (namely,CARMA systems),and presents the decomposition based MI generalized extended stochastic gradient (GESG) algorithm and the MI generalized extended least squares (GELS) algorithm,and the filtering based MI-GESG algorithm and the filtering based MI-GELS algorithm for equation-error autoregressive moving average (EEARMA) systems (namely,CARARMA systems). |
| |
| Keywords: | parameter estimation recursive identification gradient search least squares filtering decomposition auxiliary model identification idea multi-innovation identification theory hierarchical identification principle equation-error system linear system |
|
| 点击此处可从《南京气象学院学报》浏览原始摘要信息 |
| 点击此处可从《南京气象学院学报》下载免费的PDF全文 |
|
