Discrete Robust Anti-Windup to Improve a Novel Dual-Stage Large-Span Track-Seek/Following Method

• Published in: IEEE transactions on control systems technology 2008-11, Vol.16 (6), p.1342-1351
• Main Authors: Herrmann, G., Hredzak, B., Turner, M.C., Postlethwaite, I., Guoxiao Guo
• Format: Article
• Language: English
• Subjects: Actuators; Anti-windup; Anti-windup (AW) compensation; Applied sciences; Bandwidth; Circle criterion; Compensation; Computer science; control theory; systems; Control system analysis; Control system synthesis; Control systems; Control theory. Systems; Design engineering; Disk drives; dual-stage actuator; Electronics; Exact sciences and technology; Frequency; hard disk drive (HDD) servo-control; Hard disks; Instruments; Memory; Microactuators; Miscellaneous; Robustness; Saturation; Servosystems; Settling; Storage and reproduction of information; Studies; track-seeking/following
• ISSN: 1063-6536; 1558-0865
• DOI: 10.1109/TCST.2008.917876

An application of an advanced anti-windup compensation scheme to a recently developed large-span track-seeking and track-following method for dual-stage actuator systems in hard disk drives (HDDs) is presented. In the original dual-stage servo control approach, micro-actuator saturation was addressed using the Circle Criterion to guide the controller design. However, such an approach does not allow the fine-tuning of seeking and settling speeds and also limits the class of linear controllers for the secondary loop due to the Circle Criterion. Anti-windup compensation removes many of these limitations and improves settling duration, disturbance rejection, recovery from saturation in the secondary actuator loop, and design flexibility. A general anti-windup (AW) compensator is presented in discrete time bringing together a number of AW concepts and results previously given in continuous time. Robustness is taken into account and performance requirements are expressed via frequency weights. The robust performance design is based on an l 2 -optimization approach using linear matrix inequalities (LMIs).