Optimal Planning Considering Distributed Energy Storage Full Life Cycle of Charge/Discharge Strategy

Optimizing charging/discharging strategies for distributed energy storage systems in power networks over their lifecycle is crucial for maximizing benefits and ensuring economic viability. This study maximizes investment returns over the storage lifecycle, key to economic viability. A model integrating time-of-use pricing and random power flows uses DDQN-based deep reinforcement learning to optimize storage strategies, locations, and capacities. It enhances return rates and power quality in distribution networks under variable flows. Applied to an IEEE33-node system with six candidate sites, it identifies optimal configurations, yielding 1.7 % - 2. 5 % annual returns and proving the method's effectiveness.