Abstract:
Disturbance decoupling and disturbance decoupled estimation problems in linear, time invariant, dynamical system are studied In a common framework using geometric approach. The basic solvability question is investigated for disturbance docoupling problem by static and dynamic state and measurement feedback. Estimation of the state vector or a function of the state vector of a system in the presence of disturbances is considered. The concepts of observable, unobservable subspaces and observability of a system are generalized for unknown input systems. In the second part of the thesis, solvability of the above problems in a special kind of system which consists of separated dynamic and algebraic parts is considered. In general, the solvability range of the problems is improved by system decomposition. AII results are fully constructive and an appendix is included for the numerical computation of the developed design methods.
