Adaptive Mode-Suppression and Disturbance-Rejection Scheme With Application to Disk Drives

Abstract

The goal of a track-following controller for a disk drive is to maintain as close to perfect tracking as possible. Better tracking allows for more data to be stored on a single disk. There are several factors which make this control problem difficult, such as high-frequency resonant modes, unknown disturbances, and a preset sampling rate. The resonant modes of a disk drive are often uncertain and vary between units and can also lie near or beyond the Nyquist frequency. Suppressing these modes can be difficult. However, a multirate adaptive notch filter and adaptive bandwidth controller is proposed to suppress these resonant modes. A robust online estimator that estimates the mode frequencies is used to allow the multirate adaptive notch filter to track the disk drive's modes. By adding a multirate scheme to the adaptive notch filter, it can suppress modes at higher frequencies, which are close to or beyond the Nyquist frequency of the preset sampling rate. As the multirate adaptive notch filter tracks the plant mode frequencies, the adaptive bandwidth controller ensures that stability and performance requirements are satisfied by using the online estimate of the mode frequencies. An adaptive disturbance rejector is added to the adaptive mode-suppression scheme and broken into two parts: a part to suppress the repeatable runout and another to attenuate the residual disturbance by using a neural model. Both parts are adapted online and combined with the mode-suppression scheme to ensure good disturbance rejection even in the presence of unknown or changing resonant modes. Simulation results with a hard disk drive as the application are included and show that the adaptive mode-suppression scheme is able to maintain stability and better performance than a nonadaptive counterpart for an uncertain resonant-mode frequency. When combined with the adaptive disturbance-rejection scheme, stability is maintained and an improvement of as much as 14.3% in 3sigma is seen. This improvement in tracking can allow for data to be stored more densely on a single disk, providing larger storage media for the same size of disk drive.