XGBoost-based power distribution network line summer peak load prediction method

Abstract

The invention relates to a power distribution network line summer peak load prediction method based on XGBoost. The method comprises the steps: firstly carrying out the summer load rate peak increase trend clustering analysis of lines in a prediction line list, and screening out a line which is likely to be subjected to heavy overload in the future as a to-be-predicted line; then selecting a typical line from the screened prediction lines to carry out XGBoost model modeling and parameter tuning, wherein sample characteristic values adopted by an XGBoost model comprise meteorological characteristic data, time characteristic data and spring load characteristic data; and finally, predicting a line load curve of a summer load peak cycle of the screened prediction line list by using a unified model parameter, and outputting a prediction result list. Aiming at the problems that the historical data volume of the line is large and the load prediction calculation amount is large, the line to be predicted is reduced through load rate peak value growth trend clustering analysis, unnecessary calculation is avoided, and the prediction precision is improved.

Description

technical field [0001] The invention relates to an XGBoost-based summer peak load prediction method for distribution network lines, belonging to the technical field of distribution network load prediction. Background technique [0002] During the peak hours of electricity consumption in summer, the power supply situation of the urban distribution network is relatively severe, and the phenomenon of heavy overload of lines may occur, which will have an impact on the distribution network. In order to meet the summer peak smoothly, before the arrival of summer each year, the distribution network management personnel usually predict the peak load of the distribution network lines 1-2 months in advance, accurately identify the lines that will experience heavy overload in summer, and predict their peak load It is very important for the power sector to formulate and implement grid planning, capacity expansion and reconstruction and expansion programs in a planned way. [0003] Ther...

AI Technical Summary

Problems solved by technology

[0003] There are usually thousands of distribution network lines. To predict the summer load peak value of distribution network lines in the next 1-2 months, it is necessary to extract a large amount of historical data of the forecast object, and at the same time, it is necessary to consider that different lines may have different levels of load influencing factors , and the forecast time span is between the short-term forecast and the medium-term forecast, which puts forward higher requirements on the generalization of the model, making load forecasting more difficult

Images