Adaptive switching supervisory control of nonlinear systems with no prior knowledge of noise bounds. (Q1428135)

The authors develop switching supervisory control algorithms for nonlinear plants with uncertain parameters and subject to inaccessible persistent disturbances of unknown maximum size. In the analysis, the authors utilize Lyapunov-based controller falsification tests, which are combined with a suitable switching logic and guarantee the stability of the resulting control system. The analysis is restricted to constant uncertain parameters, and simulation examples show that the algorithm is capable of coping effectively with slowly time-varying or piecewise constant parameters.