The invention provides a long-
time sequence power load prediction method based on an improved Informer model, and particularly relates to the technical field of
power load prediction, and the method comprises the steps: carrying out the
feature coding and embedding of meteorological features, scalar information, position information,
time information and trend information through a feature embedding layer and an SSI model, and carrying out the
feature coding and embedding of the meteorological features, scalar information, position information,
time information and trend information; the method comprises the following steps of: providing richer input feature representation for a model, carrying out secondary sparsity on probability sparse attention by adopting a method for carrying out sparsity evaluation on a key value matrix, reducing the
time complexity of an attention mechanism from O (LlnL) to O ((lnL) 2), and finally outputting a final predicted value through a decoder and an
encoder; according to the method, the problems that when an original Informer model is applied to
power load prediction, feature richness is insufficient and a probability sparse attention module still has insufficient sparsity are solved, so that the effect of improving the precision and efficiency of long-
time sequence power load prediction is achieved.