This paper introduces a fast learning mechanism to address constrained input in the higher derivatives Newton-based extremum seeking. The proposed algorithm has a two-time-scale structure consisting of: a compensation mechanism (i.e. an anti-windup compensator) with fast dynamics that compensates for the effect of the constrained input, and a slow subsystem to maximize the map’s higher derivatives by regulating the output. The practical asymptotic stability of the new ES algorithm is proved using a modified version of the singular perturbation method. The effectiveness of the proposed algorithm is demonstrated using simulations.
|2022 IEEE International Conference on Systems, Man, and Cybernetics (SMC)
|09/10/2022 → 12/10/2022
|I E E E International Conference on Systems, Man, and Cybernetics. Conference Proceedings