This paper investigates the design of composite nonlinear-feedback (CNF) control law for a hard-disk-drive (HDD) servo system. First, a scaled nonlinear function is introduced for the CNF control law, in which a parameter is scaled by the error between the amplitude of the target reference and the initial value of the system controlled output. The closed-loop system under the scaled function has robust transient performance to the variation of the amplitude of the target reference. Then, the parameters of the selected nonlinear function are tuned by optimal tuning methods. More specifically, the parameter-tuning problem is formulated as an optimization problem, which can be solved efficiently via numerical methods. The simulation and experimental results show that the control law designed using the new approach yields excellent performance for both track seeking and track following in the HDD servo system.
Read full abstract