Based on Deep Reinforcement Learning Algorithm, Energy Storage Optimization and Loss Reduction Strategy for Distribution Network with High Proportion of Distributed Generation

Abstract

The integration of distributed generation (DG) at high levels exacerbates line loss in distribution networks. Improving the output power stability of DG and clarifying the impact of DG integration on line loss are critical issues in distribution network optimization. Firstly, the impact of DG integration on line loss in distribution networks is analyzed, and the line loss situation in an IEEE 14-bus system when DG deviate from their design capacity is calculated. The results show that the line loss in the distribution network will increase rapidly when DG operate outside their design capacity. Then, a deep reinforcement learning-based DG energy storage optimization strategy is proposed with the objective of improving the net output power stability of DG. Simulation results demonstrate that this energy storage control algorithm can effectively alleviate the instability of DG output power in the distribution network, ensuring that DG operate at or near their design capacity for extended periods, thereby reducing the additional line loss caused by DG integration in distribution networks.