NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Rejecting outliers: was: Kurtosis.
From: Gary LaPook
Date: 2011 Jan 1, 11:15 -0800
From: Gary LaPook
Date: 2011 Jan 1, 11:15 -0800
Another way to improve the accuracy an an LOP derived from a series of sights on one body is to take the mid point of the times and the median of the altitudes. This is the principle used in many "averagers"incorporated in bubble octants for use in flight such as the A-7, A-10, A-12 and others, These all use disks or drums attached to the altitude adjustment knobs onto which you impress pencil marks my pressing a plunger when the body is collimated with the bubble. The A-10A makes a mark every second by use of a solenoid activated plunger triggered by a clockwork mechanism. At the end of the series you visually find the median between the marks and use that for your Hs and the mid time of the series as the time of the observation. Since the times of the observations are not marked down and, except for the A-10A, the marks are not evenly spaced in time, using
the midpoint of the time doesn't really represent the time of the median sight but is close enough. Even when using a bubble octant in flight that doesn't have this type of averager the navigator records only the start and end time of the series and then the series of altitude readings. He then finds the average of the altitudes and used the mid point of the time for his observation. Obviously, these methods do not produce LOPs with the precision of a least squares fit or other methods discussed on this thread but good enough for flight navigation. See: https://sites.google.com/site/fredienoonan/topics/pionneer-octant gl --- On Sat, 1/1/11, Geoffrey Kolbe <geoffreykolbe@compuserve.com> wrote:
|