GRU-NN power load level prediction method based on EMD-SVR-MLR and attention mechanism

The invention discloses a GRU-NN power load level prediction method based on EMD-SVR-MLR and an attention mechanism. The method comprises the steps: carrying out the decomposition processing of original time series data of a user load through employing an empirical mode decomposition signal processing algorithm, and carrying out the reconstruction of a mode function; and establishing a support vector machine multiple linear regression primary prediction model for the reconstruction component so as to obtain a prediction value. The prediction value of the primary prediction model and original time sequence data are fused to form a new time sequence, the new time sequence serves as the input of a gating circulation unit neural network, modeling learning is conducted on the internal dynamic change rule of features, an attention mechanism is introduced to give different weights to the implicit state of the GRU, and finally short-term load prediction is completed; according to the method, the feature extraction ca