A stable neuro-fuzzy controller for output tracking in composite nonlinear systems

A learnable fuzzy controller is proposed for online implementation of a decoupling control action of uncertain composite affine nonlinear plants to track a prescribed trajectory. The controller is mainly composed of decentralized fuzzy systems with embedded two-stage rule credit assignments cascaded with an interconnection-compensating associative memory network. The controller can be parametrized by a set of linear parameters, which represent a combination of the credits of rules, locations and shape factors of membership functions. The parameters are tuned by a dead-zone adaptation algorithm. It is shown that the adaptive fuzzy controller guarantees a given level of attenuation for the tracking error in the presence of unknown but bounded interconnections and disturbances. A simulation of an inverted pendulum is given to demonstrate the effectiveness and robustness of the controller.