Adaptive fuzzy logic controller with changeable universe of discourse using neural fuzzy networks

The accurate input-output universe of discourse (UOD) on which membership functions are defined is hard to acquire in many situations and control accuracy will reduce greatly in the steady state due to limited fuzzy control rules. This paper presents a design scheme for the adaptive model-independent PI- and PD-type fuzzy logic controllers with on-line changeable input-output UOD by using neural fuzzy networks, which can solve two problems mentioned above to a great extent. Fuzzy inference rules defined by the error and change of error were applied to describe the variation law of the input-output UOD according to the current trend of the controlled process and the neural fuzzy networks was used especially to optimize the change of the output UOD considering its importance to the control performance and its complexity to be accurately set by experience. Simulation results for the typical second linear system with large dead time in two cases show the effectiveness of the proposed fuzzy logic controller.