Non-intrusive load identification method based on feature visualization

The invention relates to a non-intrusive load identification method based on feature visualization. Aiming at the problem of low identification accuracy caused by large network model parameters and high-power load and harmonic rich load when multiple loads work simultaneously in a traditional load identification algorithm based on one-dimensional sequence characteristic quantity, harmonic analysis is carried out on original current data, and then, through a Gramian Angle Field (GAF), a recurrence plot (Recurrence Plot), a high-power load and a high-power load when multiple loads work at the same time, a high-power load and a high-power load are used for carrying out harmonic analysis on the original current data, so that the load identification accuracy of the load identification algorithm based on the one-dimensional sequence characteristic quantity is improved. And the one-dimensional harmonic feature sequence is converted into a two-dimensional image through a CNN-based image classification method, and the