Neuro fuzzy network (NFN): based speed estimators for DC motor / Nafisah Abdullah

This report describes the design of Neurofuzzy based speed estimator for separately excited DC motor using MA TLAB/f oolbox. A comparative analysis of the DC motor drive's behavior with and without Neurofuzzy based was performed. It is shown that Neurofuzzy is a good estimator to estimate speed and enables very good quality of the drive performance over a wide range operating conditions for both open and close loop systems. For the purpose of the training, ANFIS (Adaptive-Network-based Fuzzy Inference System) was used because the problem only can be tackled by using differentiable functions in the inference system. ANFIS uses back-propagation learning to determine premise parameters (to learn the parameters related to membership functions) and least mean square estimation to determine the consequent parameters. General rule is to obtain the best performance of Neurofuzzy DC motor speed estimators with minimum training parameters. From the results obtained, it shows that Neurofuzzy is alternative controller to replace the classical method.