Short-time air quality prediction method and system based on data fusion

The invention relates to the technical field of artificial intelligence, and provides a short-time air quality prediction method based on data fusion, which comprises the following steps: extracting a plurality of characteristic factors in historical air quality data, carrying out data preprocessing, processing a pollutant data sample X by adopting a wavelet denoising method, and outputting a first data set XW, the method comprises the following steps: decomposing a pollutant data sample X into a multi-mode component set XV by adopting a variational mode decomposition method, fusing the multi-mode component set XV with meteorological data X * to generate a fused data set D, carrying out dimensionality reduction and feature extraction on the fused data set D by adopting a principal component analysis method, and carrying out PM2.5 concentration prediction by adopting a trained BiLSTM model, so as to reduce cross interference of feature factors. Effective characteristics of nonlinear and unsteady air quality da