Power consumer load prediction model selection method and device

Abstract

The invention discloses a power consumer load prediction model selection method and device, and the method comprises the steps: collecting a load prediction model and prediction dimension data of an existing user, and carrying out the preprocessing of the data, and forming a data set; scanning the data set layer by layer, setting a minimum support degree and a minimum confidence degree, determining a frequent item set of the load prediction model, and forming a user applicable model classification rule; acquiring newly-added user data, traversing load prediction models of all existing users for the newly-added user data according to user applicable model classification rules, and determining the load prediction models of the newly-added users; and predicting the power consumption load of the newly added user according to the load prediction model of the newly added user. The method achieves the quick selection of the newly-added user load prediction model, and improves the accuracy ofnewly-added user load prediction of an electricity selling company.

Description

technical field [0001] The present invention relates to the application field of data forecasting and model selection technology, and in particular to a method, device, readable medium and electronic equipment for selecting a power user load forecasting model. Background technique [0002] With the gradual opening of the electricity market, more and more users are participating in electricity market transactions. When electricity sales companies act as agents for electricity users, reducing assessment deviations is the key to their profitability. The load data of new power users is still relatively small, so how to carry out effective load forecasting for new users is very important. [0003] The current forecasting methods for user power forecasting are based on the load data and production information of existing users. For new users, due to the collection of less load data and production information data, the traditional load forecasting method needs to be iterated conti...

AI Technical Summary

Problems solved by technology

[0003] The current forecasting methods for user power forecasting are based on the load data and production information of existing users. For new users, due to the collection of less load data and production information data, the traditional load forecasting method needs to be iterated continuously to achieve better results. The forecast effect of

Images