首页 | 本学科首页   官方微博 | 高级检索  
     检索      

模糊神经网络预测在卫星姿态测量中的应用
引用本文:王俊璞,田蔚风,金志华.模糊神经网络预测在卫星姿态测量中的应用[J].中国惯性技术学报,2004,12(1):43-48.
作者姓名:王俊璞  田蔚风  金志华
作者单位:上海交通大学仪器工程系,上海,200030
基金项目:中国航天科技集团公司“航天科技创新基金项目”资助
摘    要:以满足对地观测卫星测姿精度为目标,将由惯性基准、红外地平仪和太阳敏感器测姿过程视为典型的建模问题,讨论了基于自适应神经网络的模糊推理系统(ANFIS)的卫星姿态预测。仿真结果表明,ANFIS预测能够满足卫星姿态测量精度的要求,具有较强的容错性,同时该方法可将俯仰、横滚和偏航三个姿态分离建模,有利于提高卫星姿态测量的可靠性,为卫星姿态测量信息处理提供了一种新的方法。

关 键 词:对地观测卫星  姿态测量  ANFIS  卡尔曼滤波
文章编号:1005-6734(2004)01-0043-06
修稿时间:2003年12月18

Attitude Measurement of Remote Sensing Satellites Using Neural Network-Fuzzy Reasoning System
WANG Jun-pu,TIAN Wei-feng,JIN Zhi-hua.Attitude Measurement of Remote Sensing Satellites Using Neural Network-Fuzzy Reasoning System[J].Journal of Chinese Inertial Technology,2004,12(1):43-48.
Authors:WANG Jun-pu  TIAN Wei-feng  JIN Zhi-hua
Abstract:To ensure the high accuracy of attitude and orbit control in remote sensing satellite applications, the measurement of attitude should employ several kinds of sensors, such as gyros, earth sensors, solar sensors, etc.. This method depends on the system dynamic model and often fails to ensure error convergence when small changes happen to the system. To overcome this weakness, an adaptive network-based fuzzy inference system (ANFIS) is employed to process the information collected from the sensors. A simulation study has been made, which demonstrates the effectiveness of this approach. Meanwhile, the pitch, roll and yaw can be predicted by their respective ANFIS models, which improves the reliability of the measurement of attitude.
Keywords:ANFIS  Kalman filter  attitude measurement  remote sensing satellite
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号