A Robust Iterative Learning Control Algorithm for Uncertain Power Systems

Abstract

Learning control is an iterative approach to the problem of output tracking for processes that are repetitive in nature and operate over fixed time intervals. The adaptability of the control law to changes of model parameters, to unmodelled dynamics and to nonlinearities of real systems has been proven valuable for applications in management and control of components of production and distribution networks. The ILC problem is converted to the design problem of the linear approximation parameters of the input function in a subspace spanned by a set of basis functions. Results on positive invariant sets of linear discrete-time systems allow for the development of a robust ILC algorithm that deals with the uncertainties in the systems model and the restriction of the input and output spaces. The method is presented in view of applications in the management and control of components of production and distribution networks such as Active Power Filters, Inverters for interfacing with renewable energies microgrids and wind turbines.