Electricity consumption prediction method based on energy development and Shapley value weighting

Abstract

The invention relates to an electricity consumption prediction method based on energy development and Shapley value weighting, and the method comprises the following steps: predicting the economic development condition based on main economic factors influencing the energy and electricity consumption development; predicting the energy development condition by adopting a multivariate regression model; according to historical electricity consumption data and energy data, on the basis of the economic development condition and the prediction result of the energy development condition, adopting multiple electricity consumption prediction models to predict regional electricity consumption development, and obtaining multiple electricity consumption prediction values; and carrying out weighting processing on the plurality of electricity consumption prediction values through a Shapley value theory to obtain a final electricity consumption prediction result. Compared with the prior art, the electricity consumption prediction result obtained finally can better consider the medium and long term electricity consumption increase change condition, a basis can be provided for a long-term planning target of the power grid, the electricity consumption demand condition is accurately analyzed from multiple angles of economy and energy, and an effective reference is provided for power grid planning.

Description

technical field

[0001] The invention relates to the technical field of power system load forecasting, in particular to a power consumption forecasting method based on energy development and Shapley value weighting. Background technique

[0002] As a branch of power load forecasting, power consumption forecasting is an important preliminary work for power system planning. The scale of electricity consumption is affected by many factors, including population, GDP, urbanization rate, etc. Power consumption forecasting can provide the basis for the long-term planning goals of the power grid, and provide an effective reference for power grid planning.

[0003] The typical autocorrelation model for saturated load forecasting is the Logistic curve model. Some studies select the Logistic curve model for saturated load forecasting, and set different future development scenarios for forecasting according to the traditional S-curve characteristics of load growth; The correlation mode...

Images