Dynamic Tuning of the Optimal Fuzzy LQR Controller
Miroslav Pokorný, Tomáš Dočekal · 2019
The paper is dedicated to the problems of control of non-linear continuous systems. The nonlinear fuzzy model of Takagi-Sugeno type utilizes method of local approximation in fuzzy partition spaces. Model-based fuzzy controller synthesis is based on the concept of parallel distributed compensation PDC. LQR fuzzy controller is designed using linear matrix inequalities LMIs method. In additional, GA-based tuning of LQR weighting matrices is proposed to achieve better performance and high robustness of close-loop LQR fuzzy controller. Described method is illustrated to the problem of balancing an inverted pendulum on the cart in upright position.