航空学报 > 1985, Vol. 6 Issue (4): 371-378

由后验概率密度递推导出的地形轮廓匹配定位算法

陈必红   

  1. 长沙工学院
  • 收稿日期:1984-06-23 修回日期:1900-01-01 出版日期:1985-08-25 发布日期:1985-08-25

TERCOM ALGORITHMS DERIVED FROM RECURRENCE OF POSTERIOR PROBABILITY DENSITY FUNCTION

Chen Bihong   

  1. Changsha Institute of Technology
  • Received:1984-06-23 Revised:1900-01-01 Online:1985-08-25 Published:1985-08-25

摘要: 飞行器一边飞行,一边不断测出下方地面的海拔高度或飞行器与地面间的高程差,和已知的基准地形图比较,从而确定飞行器的位置,这就是地形轮廓匹配定位问题。本文给出了根据测高数据进行后验概率密度递推的基本公式,由此简化得到的几个具体算法,经计算机在实际地形图上仿真表明,具有递推执行,截获概率和定位精度较高,可用于任意飞行航迹和不等间隔测高,可对付较严重的随机扰动、航速误差和航向偏差等特点。

Abstract: A general form of the optimal algorithm of the Terrain Contour Ma-tching(TERCOM) is derived. By simplifying the general form, several realizable algorithms for different models are obtained. Simulation of them has been done and the results show the following Peculiarities;(1) The measurements are processed recursively. The algorithms begin to be executed as soon as the vehicle receives the first measurement. So they can save more computation time than the algorithms with batch processing.(2) There is no limitation on the flight trace of the vehicle. The vehicle can maneuvre while the algorithms are being carried on.(3) The algorithms can work at lower signal-to-noise ratio.(4) The convolution-multiplication algorithm can work under serious random disturbance and has good robustness to the existing course and speed deviations.These algorithms may also be applied to pattern recognition, image recognition and some filter problems.