Repetitive learning control system using disturbance observer for head positioning control system of magnetic disk drives

Abstract

Magnetic disk drives are widely used as data storage devices, and the demand for larger capacity and smaller size is increasing. In order to meet this demand, a head positioning control technique is expected to improve the head positioning accuracy by reducing the tracking error caused by disturbances such as bias-force and repeatable run-out. We have developed a disturbance compensation system that is a learning system which can reduce the repeatable run-out tracking error by updating the compensation table using a disturbance observer. This method has reduced the tracking error to the background noise level of a position error signal.