Isolated power systems driven by wind energy face significant challenges when supplying non-linear loads without grid support. These systems require sophisticated control mechanisms to maintain voltage stability and power quality. Researchers have developed an optimized control strategy using evolving Takagi-Sugeno Kang artificial neuro-fuzzy inference (ETSK-ANFIS) paired with a static synchronous compensator (DSTATCOM) at the point of common coupling.
The control system addresses the nonlinearity inherent in standalone wind generation with permanent magnet synchronous generators (PMSG). Rather than relying on fixed control parameters, the ETSK-ANFIS algorithm adaptively extracts active and reactive current components in real time, enabling effective harmonic mitigation. The system was further optimized using the marine predator algorithm (MPA), a bio-inspired optimization technique that balances local refinement with global exploration to improve control performance.
Testing demonstrated strong performance under demanding conditions. The PMSG-based system generated line voltage at 235.1 volts with total harmonic distortion of 2.8%, while current THD measured 4.6%—both within IEEE-519 standards for power quality. Critically, the system maintained balanced three-phase fundamental source current at the generator terminals even when the load was severely unbalanced, a key challenge in isolated grids.
The prototype, implemented on a real-time Micro-Lab Box platform, validated both simulation and experimental responses. Variable wind speeds and unbalanced load conditions were tested to confirm robustness. This adaptive control approach offers significant value for remote microgrids, island communities, and military applications where diesel backup is expensive or impractical. The combination of fuzzy logic with marine predator optimization provides a flexible framework that could be extended to hybrid wind-solar systems or multi-source distributed generation networks.
