Power equipment state monitoring and model training method, electronic equipment and storage medium

The embodiment of the invention discloses a power equipment state monitoring and model training method, electronic equipment and a storage medium. The method comprises the following steps: inputting a patrol image of power equipment into a pre-trained image reconstruction model; performing shallow layer feature extraction on the patrol image through a feature extraction module of the model to obtain shallow layer image features, and obtaining deep layer image features based on the shallow layer image features through a nonlinear mapping module of the model; through an up-sampling module of the model, an up-sampling operation is adopted based on the shallow image features and the deep image features, and a reconstructed high-resolution patrol image is obtained; and determining the operation state of the power equipment based on the reconstructed high-resolution patrol image. According to the embodiment of the invention, the accurate operation state of the power equipment can be obtained based on patrol image m