Study of a class of sampled-data ILC from the point of performance improvement and memory Capacity

This paper presents the design and application of a sampled-data iterative learning control combing feedback and feedforward learning mechanism for nonlinear systems with initial resetting error, input disturbance and output measurement noise. Firstly, a new technical result is rigorously studied to show that the tracking error will converge to a residual set if a sufficient condition is satisfied and the sampling period is small enough. The sufficient condition is guaranteed by a forgetting factor and the feedforward learning gain. Secondly, three types of sampled-data iterative learning controllers are considered from both performance improvement and memory capacity point of view based on the derived technical results. In order to demonstrate the correctness and feasibility, the proposed sampled-data iterative learning controllers are realized by a digital circuit in an FPGA chip and applied to a repetitive position tracking control of DC motor. The experimental results show the effectiveness of the three types sampled-data iterative learning controllers. Comparisons of learning performance and memory capacity are also studied extensively.