Adaptive iterative learning neural control: An error-tracking approach

In this paper, the problem of adaptive iterative learning control using neural networks is addressed by an error tracking approach for systems with arbitrary initial states. The desired error trajectory is pre-specified at the design stage. It is shown that the tracking error is ensured to converge to an adjustable neighborhood of a pre-specified one. The performance improvement is made possible in case of non-zero approximation error, due to the use of an appropriate Lyapunov functional adopted in the design.