Neuro-fuzzy design of a fuzzy PI controller with real-time implementation on a speed control system

Linguistic modelling of complex and nonlinear system constitutes to be the heart of many control and decision-making process. In this area, fuzzy logic is one of the most effective tools to build such linguistic models. Here, initially a fuzzy PI controller is designed with expert defined 49 rules to achieve desirable performance for a speed control system. Thereafter, a neuro-fuzzy controller is developed through back propagation training based on the input-output data set obtained from the previously designed fuzzy controller. Performance of the proposed neuro-fuzzy PI controller is tested through simulation study as well as real time experimentation on a DC servo speed control system. Both the simulation and experimental results substantiate the suitability of the designed neuro-fuzzy controller for closely approximating the behaviour of nonlinear fuzzy controller.