New Hybrid Five-Level Space Vector Modulation Strategy to Minimize Current Distortion and Power Ripple for a DFIG Controlled by NSOSMC Control

This article proposes a new modulation strategy that can minimize the stator current distortion and powers ripples for a doubly fed induction generator (DFIG) controlled by neuro-second order sliding mode controller (NSOSMC). The proposed five-level space vector modulation (5L-SVM) strategy is a hybrid technique which consists of 5L-SVM strategy and the intelligence technique (neural networks (NNs) and Fuzzy logic controller (FLC)). The former provides low current distortion and reduce powers ripples. On the other hand, the NSOSMC method with classical 5L-SVM strategy gives more powers ripples and harmonic distortion of stator current. For the proposed techniques, in the simulation region of neural space vector modulation (NSVM) technique, fuzzy space vector modulation (FSVM) is used to reduce the stator current distortion and powers ripples. Comparison of simulation results derived from a Matlab/Simulink software based controlled inverter is included. The work demonstrates that using the proposed hybrid 5L-SVM strategy, the three-phase current signals of the inverter can be reconstructed even under very low modulation index range while providing a low current harmonic distortion of the five-level inverter output.