Magnetic Disk Learning Control for Transient-Free Head Positioning

The patent applies preliminary action by extending the evaluation section to include post-learning control period tracking errors. This allows the learning control input to be updated in advance to account for transient responses that will occur when learning control is terminated. By evaluating tracking errors beyond the conventional output section boundary, the system proactively compensates for future accuracy deterioration before it happens.

The patent applies dimensionality change by extending the evaluation section in the time dimension beyond the conventional output section boundary. Instead of only evaluating tracking errors during the output section, the system evaluates errors in the extended evaluation section that includes the post-learning control period. This temporal dimension extension allows the system to capture and compensate for transient responses without adding complex spatial or structural elements.

The patent applies continuity of useful action by making the evaluation section continuously overlap with and extend beyond the output section. This continuous evaluation approach ensures that tracking errors are monitored without interruption from the start of learning control through the post-learning control period. The continuous evaluation enables smooth updating of the learning control input, preventing gaps that would allow transient responses to develop unchecked while maintaining efficient processing through uninterrupted data flow.