Daily electricity sales prediction method and system based on dual clustering

Abstract

The invention discloses a daily electricity sales prediction method and system based on dual clustering and a computer readable storage medium. The method comprises steps that historical daily electricity release data of a preset user is acquired; dual clustering of the historical daily electricity release data is carried out to acquire respective corresponding historical daily electricity releasedata of each dual clustering category; a respective corresponding daily electricity sales prediction model of each dual clustering category is utilized to acquire the daily electricity sales prediction result of the respective corresponding prediction time segment of each dual clustering category, and the respective corresponding daily electricity sales prediction results of all the dual clustering categories are added to acquire the prediction electricity sales of the prediction time segment. The method is advantaged in that the dual clustering algorithm requiring no information normalization is employed, information compression and loss of the data in different degrees are avoided, the respective corresponding daily electricity sales prediction model of each dual clustering category isutilized, a problem that a single prediction model can not excellently express each type of load characteristic is avoided, and electricity sales prediction precision is improved.

Description

technical field [0001] The present invention relates to the field of electricity sales forecasting, in particular to a double clustering-based daily electricity sales forecasting method, system and computer-readable storage medium. Background technique [0002] Electricity sales forecast refers to estimating or expressing future electricity sales by means of sorting out and analyzing historical data. It is the basis for formulating power generation plans and the basis for economic operation. Due to its significant importance, power grid companies have incorporated electricity sales forecast into the benchmarking index assessment of the same industry. At the same time, the accuracy of its forecast is of great significance to promote the development and construction of the electricity market. In the past, limited by the popularization of smart meters, there were many researches on the forecasting of system electricity sales and the whole distribution network electricity sales...

AI Technical Summary

Problems solved by technology

However, in the prior art schemes, when classifying the loads of power users, in order to remove the influence of the load on the load form identification method based on various distances in order of magnitude, the normalization method is often used, but different normalization methods This method will lead to information compression and loss of load data to varying degrees, and when making overall predictions for small-area users, a single prediction model cannot fit various curves well, which will lead to a decline in the final prediction accuracy

Images