Rotor Flux Observer Based Neuro-Fuzzy Techniques to Speed Vector Control of a Bearingless AC Motor
This study presents the problem of rotor flux orientation control of bearingless induction motor. The key of this solution is the estimation of rotor flux. This work applied an inference system using fuzzy logic and the neural networks with the MATLAB(R). The Adaptive Neuro-Fuzzy Inference System (ANFIS) is based in an input-output model. ANFIS is used to tune the membership functions in fuzzy system and utilize the backpropagation learning algorithm. This system was applied to estimate the rotor flux and the magnetization current of the purpose of identifying bearingless induction motor angular speed. The hybrid estimator aims at compensating possible parametric variations of the machine caused by agents such as temperature or nucleus saturation. The simulated results showed good performance. The results demonstrated the feasibility of the proposed technique. The ANFIS estimator proposed will be embedded in the DSP TMS 3208F28335 in future works.
Booktitle: Proceedings of ISMB14