Holiday Load Forecasting Method Based on Lunar Date

Abstract

The invention relates to a lunar calendar date based holiday load predication method and belongs to the technical field of electric power system short-term load predication. The lunar calendar date based holiday load predication method comprises the steps of S1 selecting a festival in n year and m working days before the festival to form into history data, predicating a festival load curve according to the selected history data, wherein the working days does no comprise rest days and the festival is a traditional lunar calendar festival in China; S2 selecting a holiday which is relevant to the history data and is not the festival date and predicating the holiday load curve, wherein the holiday is a rest day which is adjacent to the festival and selection of the holiday in history is according to the same date interval between the to-be-predicated holiday and the festival; S3 connecting the festival load curve and the holiday load curve, smoothing an overlapping region and obtaining a holiday and festival predication result.

Description

technical field [0001] The invention relates to a holiday load forecasting method based on lunar calendar dates, and belongs to the technical field of short-term load forecasting of electric power systems. Background technique [0002] The structure of the modern power grid is becoming more and more complex, and the interconnection of the power grid is becoming more and more extensive. A large power grid characterized by long distances, heavy loads, and large-area networking has gradually formed. Load forecasting is an important research field in modern power system disciplines, and it plays an extremely important role in meeting social needs and ensuring safe and economical operation of power grids. With the continuous deepening of power system reform, the whole society has increasingly higher requirements for the safe, energy-saving, economical and efficient operation of the power grid and the entire power industry. [0003] Although Chinese festivals are only one day, th...

AI Technical Summary

Problems solved by technology

[0007] 1. The load level before holidays often requires manual selection of a specified date. When there is a sudden change in load on that day due to external factors, it will lead to errors in calculating the load drop during holidays;

[0008] 2. The traditional holiday load forecasting method makes a unified forecast for each day during the holiday, without taking into account the difference in the load change trend of the entire holiday caused by the difference in the holiday range caused by the holiday, which eventually leads to the predicted Although the overall load level during the holidays was well predicted, the changing trend of the load curves of each day was not correct, and there were even problems such as early load recovery.

Images