Intelligent Fuzzy Controller Design: Disturbance Rejection Cases

Saeed Seyedtabaii | M. Moradi

---

URL :   http://research.shahed.ac.ir/WSR/WebPages/Report/PaperView.aspx?PaperID=159542
Date :  2022/06/02
Publish in :    Applied Soft Computing
DOI :  https://doi.org/10.1016/j.asoc.2022.109015
Link :  http://dx.doi.org/10.1016/j.asoc.2022.109015
Keywords :Nonlinear system, Combined cycle power plant, PSO-GA algorithm; Intelligent control, Fuzzy control,

Abstract :

---

In this paper, a modified fuzzy gain scheduling (GS) Proportional-Integral- Derivative (PID) controller using heuristically ui,, the output of the integral term of PID, as the scheduling variable is discussed. It is shown that ui can effectively represent the inaccessible low-frequency disturbances that alter the system operating point. In addition, GS design is a time-consuming task requiring multiple run of nonlinear optimization algorithm when the nonlinear system cannot easily be linearized. In this respect, a hybrid method using PSO-GA and classical fminsearch is proposed. Interpolation of the designs, for GS buid up, is carried out using an adaptive neurofuzzy inference system (ANFIS) to provide PID continuous gain surfaces. The algorithm is applied to the speed control of a combined cycle power plant under unmeasured bounded load disturbances. The simulation outputs indicate that the design quality is effectively preserved in case of small perturbations where the achievements are also superior to the responses of some published methods. In the case of large perturbations, the fuzzy scheduling PID (FSPID) supersedes successfully two other optimally designed regulators. Therefore, it is concluded that the proposed approach renders a better transient response in facing both small and large stepwise load disturbances. The findings have been verified under various working loads through extensive simulations.