Adaptive time optimal control of a disk drive actuator

Abstract

This paper presents the design of an adaptive time optimal controller for a disk drive actuator. Controller adaptation is used to achieve time optimal seek performance on individual drives with different plant parameters. Implicit model identification is done through the use of overshoots of pre-determined seeks. Adaptation is facilitated through a small, off-line trained, neural network. Adaptive controller maintains time optimal seek performance without compromising tracking accuracy. The algorithm can be easily incorporated into the existing servo system of a disk drive. Comparisons with a fixed parameter PTOS controller indicate the faster access of the proposed control scheme.