Decentralized voltage control of PV storage integrated distribution networks considering device lifetime

To address the voltage violation problem caused by the integration of distributed photovoltaic (PV) systems into distribution networks, and considering the economic operation and device lifetime of PV inverters and energy storage systems (ESS), this paper proposes a decentralized voltage control method for PV-storage integrated distribution networks, which takes into account device lifetime. First, a lifetime degradation cost model for PV inverters and ESS is established to quantify the lifetime degradation during their operation. Second, a decentralized voltage control framework based on multi-agent deep reinforcement learning is designed, which coordinates the PV inverters and ESS through centralized training and decentralized implementation, thus eliminating the need for centralized communication. Finally, the proposed method is verified through simulation on an improved Institute of Electrical and Electronics Engineers 33-bus distribution network with PV-storage integration. The results show that the proposed control strategy can balance the lifetime degradation of devices at each bus with the overall voltage control economics. It effectively suppresses voltage violations while significantly reducing the voltage control cost in the distribution network. Compared to the baseline algorithm, this method reduces voltage control costs by 39.35%.