| Summary: | 碩士 === 輔仁大學 === 電子工程學系 === 91 === In this thesis, we develop two novel approaches about the sliding mode control for uncertain nonlinear systems. For the first controller, we propose a novel sliding mode controller, which guarantees the global reaching condition of the sliding mode in an uncertain nonlinear system with multiple inputs containing both sector nonlinearities and deadzones. They can work effectively for systems either with or without sector nonlinearities and deadzones in the inputs. Moreover, the controllers ensure that the system trajectories exponentially converge to the sliding mode. For the second one, a novel fuzzy-neural sliding mode controller trained by vector-evaluated genetic algorithms (VEGA) is developed for robot manipulators with uncertainties and external disturbance to guarantee the robust stability and tracking performance. Using the VEGA, a fuzzy-neural network (FNN) is established to approximate the regressor dynamics of robot manipulators. A sequential searched crossover point (SSCP) method and a vector-evaluated (VE) mechanism are proposed to determine a suitable crossover point for the VEGA. Simulation results are demonstrated to verify the effectiveness of the proposed sliding mode controller for a class of uncertain nonlinear systems and show that the proposed VEGA-based fuzzy-neural sliding mode controller is effective and yields superior tracking performance for robot manipulators.
|
|---|
