High precision motion control for hard disk drive systems
One challenge for hard disk drives (HDDs) servo control is better precision for future drives with ultrahigh densities. The servo control design has to be robust towards uncertainty and the track misregistration (TMR) during track following has to be minimized. A new mixed H_2/H_infinity control des...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Theses and Dissertations |
Language: | English |
Published: |
2008
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/13316 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | One challenge for hard disk drives (HDDs) servo control is better precision for future drives with ultrahigh densities. The servo control design has to be robust towards uncertainty and the track misregistration (TMR) during track following has to be minimized. A new mixed H_2/H_infinity control design via a LMI based approach is proposed, with additional slack variables being introduced to characterize both the H2 and H_infinity performances. This offers additional degree of freedom and resulting in equal or better performance than existing methods. A technique that combines a disturbance observer with a phase stabilized controller is also proposed for rejection of high frequency narrow-band disturbances at two different frequencies. A nonlinear least squares optimization (NLSO) method is applied in sensitivity function shaping to design a feedback controller to attenuate disturbances at specific frequencies. Due to the limited improvement of the NLSO method, an approach based on the generalized Kalman-Yakubovic-Popov (KYP) Lemma with the Youla parameterization is proposed. An H_infinity feedforward decoupler is presented to cancel the interactions amongst the microactuators in a hybrid type servo track writer (STW). In addition, we investigate the self-servo track writing (SSTW) technology, which potentially reduces the production cost as no cleanroom environment is required. The SSTW process is modeled as a 2-D system framework, where the radial error propagation containment is formulated as a 2-D H_2 control problem. |
---|