Control System Design for Track-Seek Control with HDD Benchmark Problem

To advance research on magnetic-head positioning control in HDDs, we have developed a new open-source HDD benchmark problem focusing on track-seek control for HDDs used in AI data centers. This benchmark incorporates real measurement data from controlled objects and disturbance sources, ensuring that validation results closely reflect actual experimental outcomes. This paper introduces the benchmark problem and proposes a feedforward control strategy for track-seek control utilizing the HDD benchmark framework. The proposed approach employs a sampled-data polynomial that incorporates zero-order hold (ZOH) characteristics, enabling us to mitigate the adverse effects of discretization errors from ZOH and the residual compliance of mechanical components.