Robust and adaptive control of uncertain nonlinear and interconnected systems with its applications

Robust and adaptive control of uncertain nonlinear systems is explored in this dissertation. The proposed design procedures are systematic and capable of maintaining robustness to both parametric and dynamic uncertainties and unmeasurable external disturbances as well.; We address the robust adaptive control issue in the presence of input unmodeled dynamics. The proposed controller is robust to nonlinear input unmodeled dynamics and other types of uncertainties and bounded unmeasurable disturbances as well. Nonlinear damping terms are introduced via backstepping to counteract the uncertainties and observer design is also employed to cancel the affects of the nonlinear unmodeled dynamics. To this end, nonlinear systems transformable to the Extended Strict Feedback Form with Input Uncertainties are considered. A robust adaptive controller is designed to achieve semiglobal stability with respect to an invariant set utilizing partial state feedback.; In the context of large-scale nonlinear systems, decentralized robust and adaptive controller has been designed. The design is an extension of the results developed in the centralized case. A novel robust adaptive controller is designed to achieve robust stability in the presence of uncertainties and disturbances via partial state-feedback. To handle the input uncertainties in the interconnected systems, large-scale nonlinear systems transformable to the Form of Large-Scale Nonlinear Systems with Input Uncertainties (FLNS-IU) are considered and each subsystem contains interconnected nonlinear input unmodeled dynamics.; Applications of nonlinear robust adaptive control to stepper motors, manipulators with joint elasticities, friction compensation in servo-drives, and satellite formation-flying are considered next. A controller is designed for a detailed model of the permanent magnet (PM) stepper motor that includes both mechanical and electrical dynamics. The proposed control design removes current measurements and does not require the knowledge of stator resistances. Furthermore, the controller is robust to dynamic uncertainties and bounded disturbances present in the mechanical system such as load variations, cogging effects, and dynamic friction etc. (Abstract shortened by UMI.)...