Reliability correlated optimal planning of distribution network with distributed generation

Integration of distributed generation (DG) in distribution network changes the topological structure of distribution network. Meanwhile, the location and size of DG as well as the evolution of the topology structure of distribution network will affect the power supply reliability, which will make the planning process of the distribution network extremely complicated. For this, this paper presents a new distribution network planning model with reliability estimation method. The several network templates are obtained with the hierarchical clustering algorithm based on tree edit distance (TED), which could reduce the complexity of the distribution network. Then, the topology characteristics of the clustered network templates will be correlated with reliability index, which is formulated as explicit formulation. Similarly, the correlation model between the reliability index and DG planning is established through BP neural network. Furthermore, the distribution network planning model under each network template is aggregated, which contains the established reliability estimation model. Investment costs of distribution networks with different network templates are analyzed for different scenarios. The results suggest that unlike existing solution techniques for reliability correlated distribution network planning, the proposed method can effectively converge to the optimal solution, and evaluate the investment performance for large-scale distribution networks.