On Sept 12, staff member of State Grid Jiaxing Power Supply Company Hu Sheng carried out a new round of rolling revisions for the planning of Jiashan county's distribution network through the automatic planning system of the distribution network's digital twin. "With the system doing it all automatically, planning only takes two weeks. It's not just fast but accurate too!"

The distribution network is the last mile of power supply. In recent years, with the advancement of the Yangtze River Delta integration strategy, the urban-rural pattern and industrial layout of Jiaxing are complex and fluid, which has an impact on the implementation of distribution network planning. At the same time, due to the large scale of distribution network facilities and the continuous access of new power consumption facilities such as source-grid-load-storage, the difficulty of scientific and accurate planning is constantly increasing.

In response to the above pain points, State Grid Jiaxing Power Supply Company developed an automatic planning system for the distribution network's digital twin back in 2022 relying on the "online power grid" platform to realize the full-process automatic generation of distribution network planning projects such as diagnosing power grids and locating substations.

The automatic system is much more efficient than planning manually. "Automatic planning effectively frees up lower-level human resources. The planning time is shortened from three or four months to one or two weeks, and the power grid planning investment is reduced by about 10 percent," Hu said that this system innovates the digital management mode of distribution network planning and spatially couples the power grid data model with the local urban construction "multiple plans in one" data model to achieve comprehensive overall planning and organic connection between power grid planning and national land space planning.

As the first application of intelligent planning in the field of distribution networks, intelligence is the crowning achievement for the system's efficiency. During the research and development period, the project team innovatively developed more than 30 algorithms such as "topology + power flow" traversal analysis and for the first time introduced intelligent location and wiring technology in the application of intelligent distribution network planning.

"When people go out, they navigate with a map. This algorithm can be regarded as the navigation of the distribution network," said Hu. He noted that relying on intelligent location and wiring technology, as long as relevant parameters were set, the distribution network vision grid could be automatically generated through the algorithm to provide the optimal interconnected path for the distribution network. After the plan is made, relevant parameters also support differentiated settings and expert-assisted adjustments to ensure the rationality and implementation rate of implementing the plan. According to statistics, after using this system, the implementation rate of planning projects has increased by about 12 percent from before, and the deviation rate of planning feasibility study has been reduced by about 16 percent.

At present, this system has been promoted and applied in seven cities in Zhejiang province. The next step is for the company to add functions to the system and gradually enhance its algorithms to make it suitable for special environments such as rural areas, islands, and mountainous areas in China.