Short-term electric quantity prediction method based on cross-correlation entropy gating circulation unit

Abstract

The invention discloses a short-term electric quantity prediction method based on a cross-correlation entropy gating cycle unit, and the method comprises the steps: carrying out the data preprocessing, supplementing missing data in historical electricity consumption data, constructing a training sample set, constructing the feature input of a prediction model, and finally carrying out the data standardization. Performing feature selection by using a maximum correlation minimum redundancy algorithm; selecting a gated cycle unit (GRU) model to predict the power consumption per hour, and using a maximum correlation entropy criterion MCC corresponding to the cross-correlation entropy to replace a mean square error criterion in the gated cycle unit as a cost function of the prediction model; optimizing a key parameter p and a kernel width theta of the cross-correlation entropy gating cycle unit model through a K-fold cross validation and grid optimization method; and predicting the electricity sale quantity of the hour time scale by using the cross-correlation entropy gating cycle unit prediction model to obtain a prediction result. According to the method, two indexes including a root-mean-square error and an average absolute percentage error are used as evaluation indexes of the model.

Description

technical field [0001] The invention belongs to the technical field of electricity forecasting in electric power systems, and relates to a short-term electricity forecasting method based on cross-correlation entropy gating cycle units. [0002] technical background [0003] From the perspective of power system planning, electricity consumption forecasting is used as the basis for power system planning, national economic operation and intelligent power regulation system, and also as a necessary condition for safe and reliable power supply; from the perspective of power market, under the new power reform The trading forms of electricity retail companies participating in the electricity market mainly include mid-to-long-term transactions and spot transactions. Because when the electricity retail companies report in advance that the company's estimated user energy consumption is significantly different from the user's actual energy consumption, the electricity retail companies nee...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to propose a short-term power prediction method based on cross-correlation entropy gating cyclic unit, which solves the problem that the existing prediction technology has low prediction accuracy for non-Gaussian nonlinear electricity sales data, and it is difficult to meet the needs of electricity sales companies when they conduct electricity transactions. The problem of demand for power forecasting accuracy

Images