本文已被:浏览 2442次 下载 1次
投稿时间:2011-10-13 修订日期:2012-03-26
投稿时间:2011-10-13 修订日期:2012-03-26
中文摘要: 针对多传感器数据融合的误差配准问题,提出了一种自适应的误差配准方法。该方法使用无迹卡尔曼滤波方法训练神经网络,在偏差先验模型未知的条件下,通过学习待配准目标量测与配准目标之间的误差变化,实时估计配准误差并同时将其应用于目标状态估计。仿真实验表明,该方法能够实时有效地估计目标配准误差和目标状态。
Abstract:An adaptive target error registration method based on neural networks(NN) and unscented Kalman filter(UKF) was proposed.The method applied UKF training neural Networks. In the case of prior model of bias unknown, the method learned the difference between unregistered target measurements and registered track and estimated the registered error and target state real time. The simulation results showed the effectiveness of this method.
文章编号:201100903 中图分类号: 文献标志码:
基金项目:国家自然科学基金资助项目(60736046)
作者 | 单位 |
刘宇 | 四川大学 计算机学院 |
陈昕 | 海军空管办 |
王运锋 | 四川大学 计算机学院 |
刘洪 | 四川大学 计算机学院 |
Author Name | Affiliation |
Liu Yu | School of Computer Sci.,Sichuan Univ. |
Chen Xin | ATC System Office of Navy |
Wang Yunfeng | School of Computer Sci.,Sichuan Univ. |
Liu Hong | School of Computer Sci.,Sichuan Univ. |
作者简介:
引用文本:
刘宇,陈昕,王运锋,刘洪.一种基于神经网络和UKF的自适应目标误差配准方法[J].工程科学与技术,2012,44(3):101-105.
Liu Yu,Chen Xin,Wang Yunfeng,Liu Hong.An Adaptive Target Error Registration Based on Neural Networks and UKF[J].Advanced Engineering Sciences,2012,44(3):101-105.
引用文本:
刘宇,陈昕,王运锋,刘洪.一种基于神经网络和UKF的自适应目标误差配准方法[J].工程科学与技术,2012,44(3):101-105.
Liu Yu,Chen Xin,Wang Yunfeng,Liu Hong.An Adaptive Target Error Registration Based on Neural Networks and UKF[J].Advanced Engineering Sciences,2012,44(3):101-105.