Power Quality Enhancement in Grid-Connected Photovoltaic Systems Using Hybrid Harbor Seal Whiskers Optimization and Interpretable Generalized Additive Neural Networks

The integration of solar energy into modern power grids supports sustainability and energy efficiency but also introduces power quality (PQ) challenges such as harmonic distortion, voltage sag, swell, and fluctuations. In order to reduce PQ problems in renewable energy systems (RESs), this research proposes a novel hybrid control strategy that combines the interpretable generalized additive neural networks (IGANNs) with the harbor seal whiskers optimization algorithm (HSWOA). The unified PQ conditioner (UPQC), enhanced with a tilted integral fractional derivative with filter plus fractional derivative (TIFDNFD) controller, is employed for compensation. IGANN predicts the gain parameters for the TIFDNFD controller, while HSWOA efficiently tunes these parameters through an adaptive optimization process. The proposed HSWOA-IGANN technique is implemented in the MATLAB platform, and its performance is compared to various existing methods. The proposed method achieved a minimum total harmonic distortion (THD) of 0.42% and ensured stable load voltage at ±220 V during grid disturbances with deviations limited to 0.09–0.11 V, outperforming existing methods. The proposed method ensures efficiency, faster convergence, and accurate control of PQ, enhancing grid stability and reliability.