Stability of supervised adaptive control

A stability proof of a recently introduced supervised adaptive control algorithm is presented. The novel premise of the algorithm is its dual model architecture. The approach addresses the well known parameter burst and infinite drift phenomena. The stability result demonstrates that the dual model supervised adaptive control algorithm is robust with respect to small model/plant mismatch and bounded disturbances. Ydstie's Switching Lemma and Lyapunov stability results constitute the foundations of the theoretical results. Monte Carlo simulations illustrate the algorithm's potential.