Achieving perfect causal feedforward control in presence of nonminimum-phase behavior - exploiting additional actuators and squaring down

Control performance is limited by nonminimum-phase zeros, for example through the Poisson integral in feedback control and 'unstable' poles in inverse model feedforward control. The aim of this paper is to exploit the additional freedom in overactuated systems to overcome these limitations. In particular, an approach for causal and exact inversion of nonminimum-phase systems is presented for application in inverse model feedforward control. The proposed method is based on the observation that non-square systems often have no invariant zeros. A squaring-down approach is employed to create a square system without nonminimum-phase zeros to enable direct inversion. The proposed approach is successfully demonstrated on a benchmark system. The method enables exact inversion for non-square systems without requiring preview.