Partial state feedback tracking with ISpS disturbance attenuation via direct adaptive design

Abstract

This paper presents a systematic design procedure for constructing partial state feedback controllers that achieve semiglobally input-to-state practically stable disturbance attenuation for a class of uncertain time-varying nonlinear systems whose unmeasured dynamics do not possess a constant disturbance attenuation level (CDAL). After the identification of a necessary condition for the existence of a CDAL, adaptive neural network control design is presented, where the universal approximation property of neural networks and the domination design technique are combined to overcome the difficulties due to the lack of state information, unknown system nonlinearities, and unknown state dependent disturbance attenuation gain. The proposed method is coherent in the sense that it is applicable to the case in which a CDAL exists.