Disturbance rejection and tracking using output feedforward control

The disturbance rejection and tracking problems that have particularly nice solutions when full state feedback is used are discussed. Although the output feedback problem, in general, has a very difficult solution, it is seen that for these special cases, the solution involves solving only a linear two-point boundary-value problem, rather than the nonlinear problem associated with the general case. The Kronecker algebra is of considerable usefulness in obtaining the solution, which involves a matrix linear two-point boundary-value problem. The solution has applications in the area of large-scale systems where one may consider the disturbance to be the interconnection vector which reflects how the other states of a large system affect the local dynamical model.<
>