LSTM- GA Enhanced Power Quality Improvement in STATCOM Integrated SPV Microgrid

Abstract

This research explores the performance of various control techniques for Static Synchronous Compensators (STATCOM) and Doubly Fed Induction Generators (DFIG) in microgrid environments, with a focus on the integration of solar photovoltaic (SPV) systems. The study compares Fuzzy Proportional-Integral (Fuzzy-PI), Particle Swarm Optimization-tuned PI (PSO-PI), and Long Short-Term Memory-Genetic Algorithm optimized PI (LSTM-GA-PI) controllers, analyzing their impact on time-domain response, harmonic distortion, and power quality. Incorporating SPV systems into microgrids facilitates the use of renewable energy, reducing dependence on fossil fuels and lowering greenhouse gas emissions, thus enhancing environmental sustainability. The coordinated control between SPV and STATCOM improves power quality by ensuring stable voltage and frequency, optimizing energy use, and reducing electrical disturbances and waste. The results show that the LSTM-GA-PI controller outperforms the other methods, achieving the fastest response times and the lowest maximum overshoot, while also significantly reducing Total Harmonic Distortion (THD) in both voltage and current. This underscores the LSTM-GA-PI controller’s capability to enhance microgrid performance by maintaining stable and efficient power quality. Advanced control techniques like LSTM and GA enable microgrids to adapt to changing conditions, minimize losses, and improve energy flow. This study demonstrates that integrating these algorithms into STATCOM and SPV systems provides a robust solution for achieving high stability and power quality in diverse operational scenarios, promoting more reliable and efficient microgrid implementations.