Aydoğdu, ÖmerAlkan, Özdemir2020-03-262020-03-2620131300-06321303-6203https://dx.doi.org/10.3906/elk-1201-115https://hdl.handle.net/20.500.12395/29194The constant parameters in a conventional fuzzy controller lead to a poor performance for time-varying systems. In this study, a fuzzy model reference learning controller (FMRLC) with a newly defined variable adaptation gain is designed and implemented in the adaptive fuzzy control of a time-varying rotary servo (TVRS) system. In the design of the FMRLC, a knowledge-base modification algorithm with variable adaptation gain is used instead of a fuzzy relation table. Hence, it is provided that the learning and adaptation mechanism continuously updates the knowledge base of the adaptive fuzzy controller against any parameter variations, such as changing loads. By means of the learning and adaptation mechanism, the TVRS system behaves as a defined reference model in the desired performance in time. The initial parameters of the FMRLC are easily determined by trial and error because of the variable adaptation gain. Using the designed controller, the adaptive fuzzy control of the TVRS system performs successfully in the simulation and practical implementation. The simulation of the system is executed in a MATLAB-Simulink environment and the practical application is implemented in a Quanser Q3 experimental servo module based on MATLAB-Simulink. The simulation and experimental results are given to demonstrate the effectiveness of the proposed control structure.en10.3906/elk-1201-115info:eu-repo/semantics/openAccessAdaptive fuzzy controlfuzzy model reference learning controlvariable adaptation gaintime varying servo systemArticle2121682180Q3WOS:000326514200004Q3