S matrix for Sequential Processor

greenspun.com : LUSENET : CU Statistical Orbit Determination : One Thread

The results for my batch and sequential processors are coming out a little differently. I'm not sure what to use for the initial S_(k-1) in the sequential processor. I've been using

S_(k-1) = -Pbar*Hx'*(R^-1)*Hc

for an initial guess. Does that sound right? Are there any reasons that the batch and sequential processors wouldn't give the same result? Thanks...

-- Rodney Anderson (rla@colorado.edu), March 03, 2000


I'm using the same thing as Rodney, in my sequential, for my first Sensitivity matrix given by:

S_(k-1) = -Pbar*Hx'*(R^-1)*Hc

...and my results from my batch and sequential seem to be pretty consistent with each other

-- MJ (Moriba.Jah@Colorado.EDU), March 03, 2000.

Moderation questions? read the FAQ