Author Login Editor-in-Chief Peer Review Editor Work Office Work

Computer Engineering ›› 2011, Vol. 37 ›› Issue (22): 162-163. doi: 10.3969/j.issn.1000-3428.2011.22.053

• Networks and Communications • Previous Articles     Next Articles

Maximum-likelihood Identification Method for State-missing Multivariate System

ZHONG Lu-sheng 1,2, FAN Xiao-ping 1, YANG Hui 2, QU Zhi-hua 1, YAN Zheng 2, QI Ye-peng 2   

  1. (1. College of Information Science and Engineering, Central South University, Changsha 410083, China; 2. School of Electrical and Electronic Engineering, East China Jiaotong University, Nanchang 330013, China)
  • Received:2011-04-19 Online:2011-11-18 Published:2011-11-20

状态缺失多变量系统的极大似然辨识方法

衷路生 1,2,樊晓平 1,杨 辉 2,瞿志华 1,颜 争 2,齐叶鹏 2   

  1. (1. 中南大学信息科学与工程学院,长沙 410083;2. 华东交通大学电气与电子工程学院,南昌 330013)
  • 作者简介:衷路生(1979-),男,副教授,主研方向:系统辨识,信号处理;樊晓平,教授、博士生导师;杨 辉,教授;瞿志华,教授、博士生导师;颜 争、齐叶鹏,硕士研究生
  • 基金资助:

    国家自然科学基金资助项目(60870010, 60864004, 6090 4049);国家“863”计划基金资助项目(2008AA04Z129)

Abstract: Maximum likelihood identification is proposed for parameter estimation of multivariable state-space models subject to missing states. The likelihood function conditioned on input, output and missing series is constructed. The influence of parameter estimation by missing mode is analyzed. And the modified Kalman filter suitable for state estimation with missing state is presented. And parameter estimation algorithm for maximization of likelihood function is given. Numerical simulation results show the effectiveness of the proposed method.

Key words: system identification, maximum-likelihood identification, multivariate system, data-missing, Kalman filtering

摘要: 提出一种极大似然辨识方法,用于解决状态缺失多变量系统的参数估计问题。通过构造以输入-输出序列为条件概率的似然函数 表达式,以及分析数据缺失程度对参数估计的影响,设计适用于状态缺失情况的卡尔曼状态估计器,在此基础上提出极大化似然函数的参数计算方法。数值仿真结果证明了该方法的有效性。

关键词: 系统辨识, 极大似然辨识, 多变量系统, 数据缺失, 卡尔曼滤波

CLC Number: