Short-term photovoltaic output prediction method based on improved Informer model

The invention is suitable for the field of photovoltaic prediction, and discloses a short-term photovoltaic output prediction method based on an improved Informer model, and the method comprises the steps: collecting historical data of a photovoltaic power generation system, and carrying out the preprocessing of the historical data; clustering the preprocessed meteorological environment data by using a twin network model to obtain a plurality of meteorological clusters; extracting feature vectors of the meteorological environment data samples through the ResNet initial model, and encoding the feature vectors of the meteorological environment data samples by using a PSO (Particle Swarm Optimization) algorithm; classifying the encoded feature vectors through an SVM classifier, and generating a plurality of training samples according to a classification result; inputting the plurality of training samples into an improved Informer model for training to obtain a short-term photovoltaic output prediction model; obt