Power load prediction method based on EEMD secondary decomposition

Abstract

The invention discloses a power load prediction method based on EEMD secondary decomposition. The method comprises the following steps: constructing a load time sequence; preprocessing the data; carrying out primary signal decomposition; carrying out secondary signal decomposition on the high-frequency signal; performing time sequence combination prediction; and outputting a load prediction result. According to the invention, the power load time series data is mined by using the multi-layer long short-term memory network. The non-stationary nonlinear original time sequence is converted into aplurality of sub-sequences through a signal decomposition mode, and the decomposed high-frequency sub-sequences are subjected to secondary decomposition, thereby obtaining the implicit deep features of the data, and effectively improving the load prediction precision.

Description

Technical field [0001] The present invention relates to the technical field of power load forecasting, in particular to a power load forecasting method based on EEMD secondary decomposition. Background technique [0002] Power load forecasting is the basis for implementing power system supply and demand balance and operation optimization. Using power market policies, power users can implement adjustment measures such as "peak cutting and filling valleys" on power consumption patterns based on load forecasts, saving production costs and realizing load control strategies with different levels of urgency. In fact, short-term or ultra-short-term load forecasting is of great significance to the optimization of power system operation. It is the basis of system power distribution, preventive control, and emergency handling. [0003] For some concentrated large power users, such as high-energy-consuming industrial users, their energy structure is complex and demand is affected by many fac...

AI Technical Summary

Problems solved by technology

[0004] The sub-sequences decomposed by EEMD are divided into high-frequency sub-sequences, low-frequency sub-sequences and residual sequences. The low-frequency sub-sequences have a long and regular period, which is easy to train and learn, while the high-frequency sub-sequences have large fluctuations and irregularities. It is also difficult for the neural network to efficiently extract and learn the feature patterns in it.

Images