A Practical Urban Distribution Network Planning Method with Geographic Information System

Zhen Wang,Xinwei Shen,Hongbin Sun,Qiuwei Wu
DOI: https://doi.org/10.1109/tpwrs.2024.3384534
IF: 7.326
2024-01-01
IEEE Transactions on Power Systems
Abstract:Urban distribution network planning (UDNP) faces new requirements and restrictions due to rapid urbanization. This paper proposes a framework for UDNP, in which the geographic information system (GIS) and a hybrid optimization model are combined to co-optimize investment and operation decisions. Firstly, we extract the street layout graph from the GIS data and produce the shortest path between distribution transformers/power substations. Then the distance matrix could be formed and used in the following UDNP optimization. The proposed hybrid optimization model utilizes 1) Clark and Wright's Savings (CWS) algorithm to produce a close-loop network topology, and 2) mixed integer linear programming (MILP) to produce the corresponding open-loop operation strategy. Case studies on an urban area in China demonstrate the effectiveness of the proposed method. The CWS algorithm improved the computation efficiency by over 1000 times compared with MILP, while introducing the shortest path for UDNP leads to a 30%+ decrease in investment compared with that of using Euclidean or Manhattan distance.
What problem does this paper attempt to address?