学習制御の開始時刻ずれの影響を補正する手法とHDDシークへの応用

抄録

Iterative learning control is effective to improve responses of systems that operate repeatedly. Iterative learning control generates feedforward control to reduce errors in the operation, revising the feedforward control using the errors in the previous operation. Iterative learning control may start in the middle of the operation to reduce memory usage. The time to iterative learning control starts are different when iterative learning control starts in the middle of the operation, and those time differences make iterative learning control less effective. In this paper we propose a new method that revise output of iterative learning control for the time differences to start iterative learning control. The proposed method shifts the output of iterative learning control in time because the time differences to start iterative learning control cause time deviations between the output of systems and iterative learning control. Simulation results show that the proposed method is effective and reduces the influence of the time differences to start iterative learning control.