ABSTRACT
Iterative learning control (ILC) is a performance-enhancing feedforward control scheme for systems that repeat the same trajectory or task. Before the start of each iteration of the trajectory, the designed ILC algorithm uses the error signal from the previous iteration(s) to generate an updated feedforward control signal. The learning process converges after anywhere from a few to tens of iterations, depending on the algorithm. In the literature it is commonly reported that ILC improves the performance of physical systems by several orders of magnitude, measured by root mean square (RMS) or maximum error, as compared to those systems’ feedback controllers.
