Abstract
This paper is concerned with a digital learing control method in which the present input is modified by the error information at the previous trial to produce a given desired output. First, it is clarified that the input saturation and disturbances cause serious problems concerning to the error convergency and achievable accuracy in the digital learning control system, especially when the sampling period becomes small. Second, a digital learning control method is proposed which overcomes the above problems. Finally, the validity of the proposed method is illustrated by simulation and experiments.
