Input-adaptive Kalman-Bucy filtering
Karl G. Brammer · IEEE Transactions on Automatic Control · 1970
The theory of nonlinear filtering (Stratonovitch, Kushner, and Bucy) is applied 1) to real-time identification of the covariance matrix of the input noise in the process model used by Kalman and Bucy [1] and 2) to adaptive mechanization of the matrix Riccati equation and the gain matrix in the Kalman-Bucy filter.