Search

The current study proposed a robust sensor-less sliding mode second-order based on a super twisting algorithm (STA-SMSO) approach using a new observer Model Reference Adaptive System-Adaptative Neuro-Fuzzy Inference System (MRAS-ANFIS). This model was applied to a doubly fed induction generator (DFIG) wind turbine running under variable wind speed and DFIG fed with a power voltage source without a speed sensor, while the control objective was used to regulate independently, the active and reactive power DFIG stator were decoupled by using the field-oriented control technique. Additionally, this process reduced the cost of the control scheme and the size of DFIG by eliminating the speed sensor (encoder). In order to improve the traditional MRAS, the MRAS-ANFIS observer was proposed to replace the usual PI controller in the adaptation mechanism of MRAS with an Adaptative Neuro-Fuzzy Inference System (ANFIS) controller. The estimation of rotor position was tested and discussed under varying load conditions in low, zero, and high-speed region. The results mentioned that the proposed observer (MRAS-ANFIS) presented an attractive feature, such as guarantees finite time convergence, good response on speed wind variations, high robustness against machine parameter variations, and load variations compared to the conventional MRAS observer and MRAS-Fuzzy. Hence, the estimated rotor speed converged to their actual value has the capacity for estimating position in deferent region (low/zero/high) of speed. [ABSTRACT FROM AUTHOR]