Robust Design of a Fractional-Delay Repetitive Controller for Uncertain Feedback Control Systems and Its Validation on a Magnetic Bearing Test Rig

This article presents a robust design of a fractional-delay repetitive controller in the frequency domain. The control goal is to accurately suppress the control error caused by a disturbance whose period may be a non-integer multiple of the sampling period. Coefficients of the repetitive controller finite impulse response (FIR) filters are calculated by solving a convex optimization problem that includes constraints limiting both the sensitivity function and the control sensitivity function. The proposed new design algorithm is verified by experiments on a test rig to efficiently damp shaft vibrations of an active magnetic bearing (AMB).