On-Line Inference for Fuzzy Controllers in Continuous Domains

Abstract

Fuzzy logic controllers are successfully applied in many complex industrial process control tasks due to their special features that other controllers lack or do not have. However, in order to cope with uncertainty in process dynamics or control environments, choosing and optimising a fuzzy controller's parameters that significantly affect the properties of the control system are indeed a time-consuming. empirical exercise. In this paper, a simplified reasoning method is employed to tune the scaling factors, which is aimed at increasing the speed of fuzzy inference, and simplifying the computing of the complexity and high dimension matrices, so that it can be executed in real time within continuous domains, as well as easily and simply for on-line tuning.