An Improved Adaptive Control Algorithm Based Upon Unfalsified Control

In this paper, an improved algorithm for data-driven adaptive control that includes ideas from Unfalsified Control is presented. This work extends our previous work on suitable cost functions in Unfalsified Control and the use of optimization to improve the set of controllers by an additional sufficient condition for robust stability that is derived only from the measured data. This condition is also used to falsify controllers in order to avoid that a non-robust controller is put into the loop.