Abstract:

[Objective] The new type of power distribution system, namely PEDF (photovoltaics, energy storage, direct current, flexibility) can effectively utilize flexible resources and promote on-site consumption of photovoltaics. To achieve optimized operation of the PEDF system, [Methods]This article uses the difference between photovoltaic output and load demand power to adjust the DC bus voltage, and uses this voltage as a signal to guide the flexible equipment of the system to adjust power, thus constructing an optimized scheduling model for PEDF system; In addition, to achieve efficient online economic scheduling of various flexible devices within the PEDF system, it is proposed to use a pulse experience replay reinforcement learning algorithm for offline training and online optimization of the model. [Results] Finally, the superiority of the optimization scheduling model for the optical storage direct flexible system proposed in this paper and the effectiveness of the proposed optimization strategy were verified through numerical simulations. [Conclusions] The control strategy proposed in this article can effectively regulate the flexible devices of the PEDF system, reduce dependence on the higher-level power grid, and improve the economic efficiency of the system operation. Through various algorithm comparisons, it has been verified that the algorithm proposed in this article has better performance.