High penetration of wind power generation system into grid network has satisfied enormous energy demand and has necessitated the wind power generation system to remain connected with grid network to maintain essential grid stability, even during grid fault conditions. This paper presents a novel hybrid genetic algorithm optimized Elman neural network controller for proton exchange membrane fuel cell supported customized Dynamic Voltage Restorer (DVR) to improve the fault ride-through capability of Doubly Fed Induction Generator (DFIG) based wind power generation systems. The outstanding aspect of the proposed customized DVR is that it has low heat transmission losses, therefore has a better efficiency up to 3% when compared with the conventional energy storage based DVR. The proposed shrewd DVR adequately rides through the fault by improving voltage regulation and prevents wind turbine termination during grid faults. The effectiveness of the customized DVR and proposed control strategy is analyzed for doubly fed induction generator based wind power generation system under symmetrical and unsymmetrical fault conditions. The operation of customized DVR is confirmed using time-domain simulation study carried out in MATLAB/Simulink environment. © 2018 Elsevier Ltd