The order-recursive Chandrasekhar equations for fast square-root Kalman filtering

The author presents a combination of the Chandrasekhar and square-root philosophies, leading to the order-recursive Chandrasekhar equations. They constitute a fast alternative to the square-root Kalman filtering algorithms. Specialized to a particular parameter estimation problem, they reduce to the fast lattice/fast QR (FLA/FQR) recursive least squares (RLS) algorithms. A number of other applications can also be considered, leading to some novel parameter estimation algorithms.<>