Bi-Level Graph-Based Optimisation for Distribution Network Reconfiguration and Optimal Placement of TCLBS and DC Switches

Distribution network reconfiguration (DNR) has been extensively studied as a strategy to improve network performance indices such as loss reduction, voltage profile enhancement, and reliability. Despite significant progress, challenges remain—particularly concerning energy imbalances during peak load periods and the need to preserve critical load points while managing load shedding. In this study, a novel graph-based bi-level optimisation model is proposed to address these issues. At the first level, a load flow analysis is performed to determine the optimal network configuration by minimising network losses and voltage deviation. During this stage, only topologies that satisfy voltage convergence and maintain the network's radial condition are retained. In the second level, a graph theory-based search algorithm is employed to determine the optimal placement of two types of switches: disconnector switches (for reducing unsupplied energy and enhancing network reliability) and telecommunication load breaker switches (TCLBS, for shedding non-critical loads during peak demand). This two-level approach ensures that the final solution complies with all operational constraints while effectively addressing the energy imbalance issue. Simulations conducted on an IEEE 33-bus test network demonstrate that the proposed method significantly improves network performance. For instance, in one scenario, energy losses, energy not supplied, and voltage deviation were reduced by approximately 29%, 21%, and 52%, respectively, compared to the initial network conditions. Moreover, the load shedding objective improved by 20%, thereby preserving critical load points. The proposed bi-level optimisation model, which leverages advanced graph-based techniques, offers an efficient and robust solution to the distribution network reconfiguration problem. It not only addresses existing challenges but also provides a promising framework for future research aimed at further enhancing network stability and efficiency.