A Reinforcement Learning-Based Approach for Retail Electricity Pricing Strategy for Multi-Microgrid Distribution Systems

This paper presents a retail electricity pri-cing strategy for multi-microgrid distri-bution systems based on a policy-driven reinforcement learning algorithm. Deve-loped from the perspective of the distribu-tion system operator (DSO), the approach enables the practical derivation of a set of hourly electricity prices that simultaneous-ly maximize profit from energy exchanges and minimize the system’s peak-to-average load ratio, thereby flattening the aggrega-te load profile. To address the absence of a complete system model from the DSO’s viewpoint, the training process employs a Monte Carlo-based method that genera-tes synthetic data from representative base profiles, enabling extensive interaction between the DSO and its environment. Si-mulations are conducted to validate the effectiveness of the proposed method. Ad-ditionally, a sensitivity analysis is presen-ted to evaluate the influence of key para-meters on the training performance and the strategy’s effectiveness