Load prediction method, system and device based on multi-source data and hybrid neural network

Abstract

The invention provides a load prediction method, system and device based on multi-source data and a hybrid neural network. Various types of input data, including historical load data, holiday data, numerical data of weather information and image data, related to ultra-short-term load prediction are fully utilized, and a hybrid neural network is adopted to train and predict a load prediction model.Compared with a traditional prediction method and a single neural network prediction method, the method has the advantages that the prediction precision is high, and the prediction result provides support and basis for a power grid power generation plan and power grid safety check. In the collected data, a box plot method and an interpolation method are adopted to correct abnormal data, so that the stability of a database is greatly improved, and the influence of large individual deviation and incomplete information on the prediction accuracy is reduced.

Description

technical field [0001] The invention relates to the technical field of power grid scheduling control automation, in particular to a load forecasting method, system and equipment based on multi-source data and a hybrid neural network. Background technique [0002] Power system load forecasting has always been a focus of attention of scholars at home and abroad. Accurate forecasting of power system load is one of the basic prerequisites for realizing safe, high-quality and economical operation of power systems. Power system ultra-short-term load forecasting refers to load forecasting within one hour in the future. The forecast results can provide support and basis for grid power generation planning and grid security checks, and further improve the accuracy of online analysis and decision-making results of power systems. [0003] The time-series curve of power load itself has certain characteristics such as uncertainty, nonlinearity and randomness. In addition, there are many f...

AI Technical Summary

Problems solved by technology

Traditional methods such as linear extrapolation and time series methods have poor prediction accuracy when the law of load changes is not clear enough. Therefore, how to effectively predict the load and improve the prediction accuracy is an important problem to be solved in power system dispatching and production planning.

Images