Enhancing open capacity in active distribution networks via GBD-based network reconfiguration considering PV integration constraints
Abstract
As the penetration rate of distributed photovoltaic (PV) systems continues to rise, the power fluctuation, reverse power flow, and node voltage constraint challenges within active distribution networks have increased significantly. This makes traditional open capacity assessment methods—based on fixed operating conditions—less effective at accurately capturing the dynamic impact of PV integration on the network's load-bearing capacity, thereby presenting new challenges for both open capacity assessment and enhancement efforts. To address this challenge, this study proposes a proactive distribution network openable capacity model that incorporates constraints arising from distributed photovoltaic (PV) integration. The model integrates multiple components, including node power flow, voltage deviation, line capacity, reverse power flow, and N–1 security constraint. Building upon this framework, the proposed approach employs the generalized Benders decomposition method (GBD) combined with the feasible cut generation method (FCG) to solve the constructed hybrid integer nonlinear optimization model. Specifically, GBD is utilized to decompose the complex optimization problem, while FCG reduces the search space by screening infeasible regions and generating effective constraints, thereby enhancing the computational efficiency of the model. Experimental results from the IEEE 33-node active distribution network test system demonstrate that the maximum open capacity reaches 7.8 MW, with an 18.2% increase after optimization. The average and peak load rates decrease from 67.2% to 43.6% and from 89.4% to 54.8%, respectively. Under extreme load scenarios, the maximum open capacity reaches 10.7 MW with a calculation time of 10.8 seconds. The proposed method improves capacity assessment and reconfiguration performance, supporting photovoltaic integration and safe distribution network operation.