Identification of the secondary actuator dynamics at frequencies beyond the Nyquist Rate in Dual-Stage Actuator Hard Disk Drive systems

Abstract

In Hard Disk Drive (HDD) head positioning control systems, the system output can be observed only at fixed finite time intervals. Identification of the plant dynamics embedded in such an inherently sampled-data system requires special considerations to avoid the aliasing effect caused by the sampling operation. In this paper, non-parametric algorithms to identify the secondary actuator dynamics in a Dual-Stage Actuator HDD system are investigated. Identification methods with continuous-time and discrete-time excitation input are proposed. In the discrete-time method, a multirate digital—analogue converter and multirate filters are employed to reduce the aliasing effect. Numerical simulation and experimental results are presented to show the effectiveness of the proposed methods.