DLMP-based Congestion Management Model for Power Distribution Network Considering Network Loss and EV Charging Demand Uncertainty

The uncoordinated charging of large-scale electric vehicles (EVs) tends to cause severe network congestion, posing a significant threat to the secure and reliable operation of power distribution networks. To address this issue, this paper proposes a distribution locational marginal price (DLMP) based congestion management method, utilizing differentiated congestion prices to incentivize EV aggregators (EVAs) to actively adjust charging patterns, thereby mitigating network congestion. In this model, an improved power flow method based on continuous implicit linearization is employed, which integrates network loss modeling and dynamically updates the linearization point, thus guaranteeing the accuracy of power flow approximation and price signal calculation. Moreover, both the uncertainty of EVA parameters and its impact on the safety constraints of the grid are characterized by robust ambiguity sets within the congestion management model, ensuring the reliability of the proposed pricing mechanism in guiding EV charging behavior in uncertain environments. Furthermore, a two-layer iterative algorithm is designed to facilitate bidirectional coordination between the distribution system operator (DSO) and EVAs. The outer layer updates network-related coefficient matrices to improve accuracy in grid state estimation, while the inner layer uses the alternating direction method of multipliers (ADMM) to iteratively adjust DLMPs while preserving the privacy of EVAs. Numerical results demonstrate that the proposed method can incentivize EV charging pattern adaptation and effectively relieve network congestion.