EKF: Fix scaling error in covariance prediction derivation

The covariance prediction derivation was using the noise 1-std value instead of the noise variance. This resulted in the values of IMU process noise required to tune not matching the observed covariance growth.
This commit is contained in:
Paul Riseborough
2016-04-05 22:23:37 +02:00
committed by Roman
parent 8678a939e2
commit ac71ec5d37
2 changed files with 18 additions and 16 deletions
@@ -193,7 +193,7 @@ G = subs(G, {'rotErrX', 'rotErrY', 'rotErrZ'}, {0,0,0});
[G,SG]=OptimiseAlgebra(G,'SG');
% derive the state error matrix
distMatrix = diag(distVector);
distMatrix = diag(distVector.^2);
Q = G*distMatrix*transpose(G);
[Q,SQ]=OptimiseAlgebra(Q,'SQ');