Real-time neuro-fuzzy inverse control applied to a DC motor

This paper describes the development of an inverse model for a direct current (DC) motor. The model consist of an Adaptive Network Fuzzy Inference System (ANFIS). The identification procedure includes: the experiment to collect data, ANFIS training and model validation in real-time. The obtained model is used to design a neuro-fuzzy inverse control strategy for trajectory tracking. The obtained real-time results are compared when an inverse ARX model is used for inverse control and to demonstrate that neuro-fuzzy strategy has a successful performance.