Comparative evaluation of hybrid control strategies for fault ride-through enhancement in type-5 wind turbines: MPC-SMC, RL-MPC, and fractional-order SMC-MPC

This study presents a comparative evaluation of three hybrid control strategies for Fault Ride-Through (FRT) enhancement in Type-5 wind turbines: Model Predictive Control with Sliding Mode Control (MPC-SMC), Reinforcement Learning-enhanced MPC (RL-MPC), and the proposed Fractional-Order SMC-MPC (FOSMC-MPC). While MPC, SMC, fractional calculus, and RL are individually established, the incremental contributions of this work are threefold: (i) a systematic integration of fractional-order sliding dynamics within the MPC prediction horizon rather than as a cascaded add-on; (ii) the first quantitative three-way comparison under identical 70% voltage dip scenarios with standardized metrics; and (iii) implementation-oriented analysis including DSP-based computational benchmarks. In the simulations, FOSMC-MPC achieves a voltage-recovery time of 0.20 s (0.58 s for RL-MPC), a peak-to-peak torque-ripple metric of 3.2 kNm (12.4 kNm for MPC-SMC), and a rotor-speed deviation of ± 5% (± 12% for RL-MPC). These results are interpreted as comparative simulation evidence; hardware-in-the-loop validation, plant-parameter uncertainty sweeps, and a fully documented matched-input test campaign remain outside the present dataset.