Grid-based Spatial Load Forecasting Method Based on POI Information Mining

Spatial load forecasting (SLF) is the basis of grid planning for distribution networks. The development of the Internet makes it possible to apply big data technology to spatial load forecasting. Based on open source big data, this paper proposes a grid-based spatial load forecasting method based on point of information (POI) mining. Firstly, the application programming interface (API) of the network map platform is invoked based on Python programming language, the open source user data is obtained, and the ray method and regular matching method are used to carry out information mining, so as to obtain the basic data for grid-based spatial load forecasting. Secondly, by referring to the distribution network planning standards, the recommended power density per unit area of the planning area is decided, and the basic data are normalized to reflect the load distribution of each grid. Then, based on the specific situation of the planning area, three total load forecasting methods are integrated for total load forecasting. Finally, combined with the total load forecasting results and load distribution, grid-based spatial load forecasting was carried out based on the total load forecasting results and grid load normalized proportion, and the forecasting results were verified by fitting the load development characteristic curve. Applying the method described in this paper according to the actual situation of the planning area can provide a basis for grid planning of the distribution network. At the end of this paper, the practicality of the method is verified in combination with the actual area.