Classical and Fuzzy Controllers for an Optimally Tuned Model of a Strip Winding System

This paper presents a refined approach to controlling the speed of a strip winding system as a representative mechatronic system with varying parameters, specifically the variable reference and moment of inertia. The complexity inherent in conventional control design is navigated by simplifying the nonlinear mathematical model commonly used in control design by introducing a second-order benchmark-type transfer function. The parameters of this benchmark-type model are optimally tuned using six optimization algorithms. One linear and two nonlinear controllers based on the popular Proportional-Integral (PI) controller are designed and evaluated in a classical control structure aiming at improving the performance of the speed control system: a conventional PI controller, a Mamdani PI-fuzzy controller with input integration, and a hybrid Mamdani PI-neuro-fuzzy controller with input integration. Simulation results are included to support the comparative discussion.