Output feedback of linear systems by error minimization

A gradient search method on the basis of an integral square-error criterion is presented for computing feedback gains in linear time-invariant systems. The system's closed-loop behavior is matched with some desired behavior, specified in terms of assigned poles and static gains.