Robust pole-assignment adaptive controllers with online tuning parameters

Introduces a robust stabilization theorem for pole-assignment adaptive controller, which is based on a robust control criteria in frequency domain and a slow time variation assumption. The robust control criteria indicates that the robust stability of adaptive system can be improved if a low gain adaptive controller is used with respect to a point wise stable modelling error. The authors suggest an `α-adaptive´ closed-loop performance which allows the desired closed-loop polynomial to follow the polynomial of the identified model so that the gain of the adaptive controller becomes controllable in the presence of model-plant mismatch. They also discuss the choice of parameter α online in order to obtain a stable adaptive control system and achieve satisfactory closed-loop performance