Power Quality Management of Inverter Based on Gradient Descent Optimization

Abstract

In isolated microgrids, the dynamic performance of the inverter output voltage is degraded due to the connection of unbalanced and nonlinear load, load switching, and significant perturbation of filter parameters. A compensation control structure based on the residual generator integrated with an optimization algorithm is proposed to improve the power quality of the inverter output voltage in this paper. First, the structure and mathematical model of the inverter are analyzed. The sensitivity results show that the residual generator–based compensating control structure improves the robustness of the inverter compared to the conventional V/f control. Second, a controller structure with lower order is derived with reduced-order theory for easier engineering implementation. Then, the compensation controller parameters are optimized online using the gradient descent optimization algorithm to improve the load disturbance’s dynamic compensation performance and reduce the power quality degradation resulting from parameter perturbation. Finally, the PEK-530 inverter platform developed by GWinstek is utilized for experimental verification. When the inverter is tested with unbalanced loads and nonlinear loads at load switching and with LC filter parameter perturbation, the three-phase unbalance, total harmonic distortion rate, d-axis voltage fluctuation, and total harmonic distortion rate of the inverter output voltage are 3.2%, 5.4%, ±2 V, and 0.5%, respectively, and are further reduced to 0.3%, 2.8%, ±0.2 V, and 0.22%, after the compensation controller is optimized by the gradient descent method. The experimental results show that the proposed improved control strategy can effectively improve the disturbance’s fast compensation performance, robustness to inverter parameter perturbation, and the power quality.