|
From: | jinbos.chan |
Subject: | [Paparazzi-devel] Kalman filter |
Date: | Wed, 29 Jul 2009 12:23:54 +0800 |
Hi,
I don't understand this equation:
P_dot = A*P + P*A_transpose + Q ,why not P=APA_transpose +Q, what is
the different between them? and where does that equation come
from?
Ths,
/*
* Typically the covariance update equation is: * * P_dot = A*P + P*A_transpose + Q * * The first of these zeros P_dot, computes part of (A*P + P*A_tranpose) * and adds in the parts of Q that are non-zero. Note that we know that * A has three zero rows at the bottom, so we do not include those in our * math. * * The second part of the covariance update computes the inner portion * of Pdot, the 4x4 region that corresponds to the quaternion. It also * updates the covariance matrix P. */ 2009-07-29
jinbos.chan
|
[Prev in Thread] | Current Thread | [Next in Thread] |