Application of Statistical Techniques to Landmark Navigation

Abstract

THEACCURACY OF LANDMARK navigation can be improved by redundant measurements in conjunction with the use of a statistical data reduction method. This paper presents several statistical methods for this purpose, including sample mean, least square regression, and Kalman filtering methods. The measured quantities are bearings of landmarks whose positions are known. Each method is described in detail. An example is given to demonstrate and compare all methods. The Fortran programs used in the computation are included in an appendix.

High-Reliability Strapdown Platforms Using Two-Degree-of-Freedom Gyros

“…The technique used will be "least-square data reduction" [5]. Since the system reconfiguration is accomplished by means of software control, it can readily be incorporated into the data reduction program.…”