Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

GeoDistNet: An Open-Source Tool for Synthetic Distribution Network Generation

Yunqi Wang, Xinghuo Yu, Mahdi Jalili

Abstract

Distribution-level studies increasingly require feeder models that are both electrically usable and structurally representative of practical service areas. However, detailed utility feeder data are rarely accessible, while benchmark systems often fail to capture the geographic organization of real urban and suburban networks. This paper presents GeoDistNet, an open-source tool for synthetic distribution network generation from publicly available geographic information. Starting from map-derived spatial data, the proposed workflow constructs a candidate graph, synthesizes feeder-compatible radial topology through a mixed-integer formulation, assigns representative electrical parameters and loads, and exports the resulting network for power-flow analysis. A Melbourne case study shows that the generated feeder remains geographically interpretable, topologically structured, and directly usable in \texttt{pandapower} under multiple loading levels. GeoDistNet therefore provides a reproducible workflow for bridging publicly accessible GIS data and simulation-ready distribution feeder models when detailed utility networks are unavailable.

GeoDistNet: An Open-Source Tool for Synthetic Distribution Network Generation

Abstract

Distribution-level studies increasingly require feeder models that are both electrically usable and structurally representative of practical service areas. However, detailed utility feeder data are rarely accessible, while benchmark systems often fail to capture the geographic organization of real urban and suburban networks. This paper presents GeoDistNet, an open-source tool for synthetic distribution network generation from publicly available geographic information. Starting from map-derived spatial data, the proposed workflow constructs a candidate graph, synthesizes feeder-compatible radial topology through a mixed-integer formulation, assigns representative electrical parameters and loads, and exports the resulting network for power-flow analysis. A Melbourne case study shows that the generated feeder remains geographically interpretable, topologically structured, and directly usable in \texttt{pandapower} under multiple loading levels. GeoDistNet therefore provides a reproducible workflow for bridging publicly accessible GIS data and simulation-ready distribution feeder models when detailed utility networks are unavailable.

Paper Structure

This paper contains 13 sections, 23 equations, 3 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Feeder Synthesis Formulation
  3. Feeder Construction and Electrical Modeling
  4. Geographic Data Processing and Candidate Graph Extraction
  5. Mixed-Integer Feeder Synthesis
  6. Electrical Parameter Assignment
  7. Load Allocation
  8. Export and Validation Quantities
  9. Case Study and Validation
  10. Study Setup
  11. Generated Network Characteristics
  12. Power-Flow Validation
  13. Conclusion

Figures (3)

  • Figure 1: Overall workflow of GeoDistNet. Geographic data are processed into a candidate graph, from which a radial feeder is synthesized and then converted into a simulation-ready electrical model.
  • Figure 2: Study area and synthesized feeder topology. The grey layer shows the underlying street-layout geometry of the selected Melbourne study area, while the red network denotes the radial feeder synthesized by GeoDistNet and the cyan marker indicates the source bus.
  • Figure 3: Illustrative load-flow validation results for the synthesized feeder under the sanity, representative, and stressed loading scenarios.