On-line intelligent adaptive control for uncertain nonlinear systems using TS-type fuzzy models with maximum allowable computational time for controller

A new Takagi-Sugeno (TS)-type FNN learning architecture is proposed for the on-line identification of the TS-type fuzzy model of the uncertain system. The dynamical optimal learning rule is adopted to update the linearized TS-type fuzzy model to guarantee the convergence of the on-line training process. To improve the convergence speed of the on-line training process, the least-squared identification is applied to identify the initial parameters of the TS-type fuzzy model. Once the linearized TS-type fuzzy model of the uncertain linear system is obtained in real-time environment, the on-line adaptive controller can be easily designed to accomplish the design specifications. A simplified tracking controller is also proposed to perform the tracking of a reference signal for unknown system. Critical constraint criteria are applied to find the computational time for generating the controller signal. Based on this sampling time, suitable equipments are used in actual hardware implementation. Inverted pendulum system is illustrated to track sinusoidal signal.