KFUpdator Class Reference

#include <KFUpdator.h>

Detailed Description

Update trajectory state by combining predicted state and measurement as prescribed in the Kalman Filter algorithm (see R. Fruhwirth, NIM A262 (1987) 444).

x_filtered = x_predicted + K * (measurement - H * x_predicted)

x_filtered, x_predicted filtered and predicted state vectors
measurement measurement vector
H "measurement matrix" projects state vector onto measurement space
K Kalman gain matrix
(formulae for K and error matrix of filtered state not shown)

This implementation works for measurements of all dimensions. It relies on CLHEP double precision vectors and matrices for matrix calculations.

Arguments: TrajectoryState & predicted state
RecHit & reconstructed hit

Initial author: P.Vanlaer 25.02.1999 Ported from ORCA.

Author

vanlaer, cerati