The method of designing the filter for a class of linear discrete generalized stochastic systems

摘要

In this paper, using the singular value decomposition of matrix and generalized inverse of matrix, the filter for linear discrete generalized stochastic system - Ex(k+1)=Ax(k)+Bu(k)+D/spl omega/(k), y(k)=Cx(k)+/spl upsi/(k)(E and A is not a square matrix), has been studied. Based on this, a new method of designing the Kalman filter for this system is given.