Nearly-Optimal Operating Parameters for Online Self-Tuning of Model Free Adaptive Fuzzy Controller
Muhammad Umair Arif, Muhammad Bilal Kadri · 2013
This paper implements online self-tuning model free fuzzy controller. Assuming no priori knowledge, the controller is able to intelligently learn, adapt and tune It-self such that it can track the desired set point with minimal error. Real time self-tuning is performed in the fuzzy controller by adapting its rule consequents using plant's qualitative information only. Simulation results show that the error arising at the plant output can be greatly minimized if the correct combination of parameters i.e. Scale factor (C_0) and forgetting factor (β) of the adaptive fuzzy controller are used. This paper aims to identify the ＂optimal operating parameters＂ of this Adaptive Fuzzy controller. This is done by analyzing the calculated RMSE values for different linear and non-linear plants and for various set-points. This study highlights the need for finding the nominal parameter values for such an Adaptive fuzzy controller for satisfactory control action.