Abstract
This paper considers the problem of simultaneous estimation of the system states and the strategy of commutation for a larger class of nonlinear switched systems. First, a hybrid high gain observer is considered to get the exact estimation of the continuous states where the strategy of switching is previously known. Then, an extension to a larger class of nonlinear hybrid systems with arbitrary switching is made. Stability analysis is widely discussed for the two cases to provide a finite-time convergence of the estimation errors. The effectiveness of the proposed hybrid high gain observer has been proved by applying it to a quadruple tank process.