Fuzzy soft deep deterministic policy gradient for distribution-static synchronous compensation of distribution networks

To address power quality issues in distribution networks, static synchronous compensation, also known as distribution-static synchronous compensation (D-STATCOM), has been included. The D-STATCOM control effect has an impact on the electromagnetic interference of distribution networks. However, the current D-STATCOM control method is slightly inadequate. To administer D-STATCOM, a fuzzy soft deep deterministic policy gradient (FSDDPG) organization mission fuzzy controller, soft actor-critic (SAC), and deep deterministic policy gradient (DDPG) is developed. The FSDDPG improves the corrosion resistance and generalization ability of D-STATCOM. The FSDDPG applies the advantages of fuzzy controller, SAC, and DDPG to complement each other and improves control performances. The FSDDPG scheme is implemented in the 415 V environment to verify feasibility and reliability. Compared with SAC, DDPG, twin-delayed deep deterministic policy gradient, proximal policy optimization, trust Region Policy Optimization, and proportional-integral methods, the FSDDPG scheme has the best performances for reactive power compensation, direct current side voltage of D-STATCOM, and active power adjustment.