Minimizing the maximum value of the regulated output to a fixed input

The problem of minimizing the maximum amplitude of a regulated output in response to a prespecified reference signal, which may be unbounded is discussed. The continuous-time case is considered. This optimization problem is converted into a simpler and more tractable one. Methods of solving it are discussed. For every epsilon >0, it is possible to compute the magnitude of the optimal error within epsilon . The authors are restricted to the case of a single regulated output, but there is no constraint on the dimension of the measured outputs or on the control input.<
>