Reducing conservativeness in robust iterative learning control (ILC) design using parameter-dependent Lyapunov functions

Iterative learning control was developed for systems that repeat the same task over a finite duration with resetting to the starting point once each execution is complete. The distinguishing feature of this form of control is the use of information from previous executions of the task to update the control signal to be applied on the next execution and thereby sequentially improve performance. Once an execution is complete, all information generated is available for use in control design and how to best use such information is a dominant issue. In applications, robust control is also a critical feature and the new results in this paper use parameter dependent Lyapunov functions to enlarge the uncertainty range allowed for successful design. These results are developed by treating ILC in a repetitive process setting.