3.4. Direct optimal feedback compensation
This method is proposed by the authors for the mitigation
of multidimensional tonal output disturbances
(Orivuori and Zenger, 2011a). The method shares
some similarities with the FSCF and IDC methods, as
it is essentially based on finding an optimal solution for
the LQ problem related to a plant model containing the
dynamics of the disturbance. However, there are also
some fundamental differences, such as the full state
feedback from the plant states and readily tunable
closed-loop robustness properties. The need for the
method arose from the problem of frequency-varying
disturbances, which was tackled by a nonlinear optimal
control law consisting of a set of continuously gainscheduled
linear control laws (Orivuori et al., 2010,
2012). The tedious task of finding a suitable and generic
design parameterization is addressed in this design. The
robustness and convergence properties of the proposed
method are easy to tune, while it still provides good
control performance even in the presence of significant
static offsets of the controlled quantities.