• 首页
  • 期刊简介
  • 编委会
  • 投稿指南
  • 收录情况
  • 杂志订阅
  • 联系我们
引用本文:安秋羽,景绍学.基于依赖性度量的Hammerstein系统的时滞估计[J].软件工程,2024,(2):36-39.【点击复制】
【打印本页】   【下载PDF全文】   【查看/发表评论】  【下载PDF阅读器】  
←前一篇|后一篇→ 过刊浏览
分享到: 微信 更多
基于依赖性度量的Hammerstein系统的时滞估计
安秋羽1, 景绍学2
(1.南京邮电大学计算机学院, 江苏 南京 210023;
2.淮阴师范学院, 江苏 淮安 223300)
b21030307@njupt.edu.cn; jingsx3@126.com
摘 要: 时滞在工业系统中广泛存在,但是现有的Hammerstein辨识算法大多没有考虑时滞,使得这些算法的应用受到局限。针对具有整数时滞的Hammerstein非线性系统,提出了一种新的时滞估计方法。该方法基于输入、输出数据的依赖性工作,首先计算输入数据和输出数据的边缘分布;其次基于2-copula函数构建它们的联合分布,利用该分布定义一个依赖性度量方法;最后基于这个度量推断时滞。该方法独立于参数估计时滞,并且避免了基于优化的时滞估计方法中存在的取整运算问题。通过数值实验和实际案例,验证了该方法可以准确估计Hammerstein系统的时滞。
关键词: Hammerstein系统;时滞估计;依赖性度量;2-copula函数
中图分类号: TP391.9    文献标识码: A
Time Delay Estimation of Hammerstein Systems Based on Dependence Measure
AN Qiuyu1, JING Shaoxue2
(1.School of Computer Science, Nanjing University of Posts and Telecommunications, Nanjing 210023, China;
2.Huaiyin Normal University, Huaian 223300, China)
b21030307@njupt.edu.cn; jingsx3@126.com
Abstract: Time delay is widely present in industrial systems, but most existing Hammerstein identification algorithms do not consider time delay, which limits the application of these algorithms. This paper proposes a new Time Delay Estimation(TDE) method for Hammerstein nonlinear systems with integer time delays. This method is based on the dependence of input and output data. Firstly, the margin distributions of the input and output data are calculated. Then their joint distribution based on the 2-copula function is constructed and a dependence measure method using joint distribution is defined. Finally, the time delay is inferred based on this measure. The proposed method is independent of parameter estimation and avoids rounding operations in optimization-based delay estimation methods.Through numerical experiments and real cases, it has been verified that this method can accurately estimate the time delay of Hammerstein systems.
Keywords: Hammerstein系统;时滞估计;依赖性度量;2-copula函数


版权所有:软件工程杂志社
地址:辽宁省沈阳市浑南区新秀街2号 邮政编码:110179
电话:0411-84767887 传真:0411-84835089 Email:semagazine@neusoft.edu.cn
备案号:辽ICP备17007376号-1
技术支持:北京勤云科技发展有限公司

用微信扫一扫

用微信扫一扫